Journal of Agricultural Science in Finland Maataloustieteellinen Aikakauskirja Vol 63:143—254 DETERMINATION OF THE CAPITALIZED AND MARKET VALUES OF SUPPLEMENTARY ARABLE LAND IN SOUTHERN FINLAND, 1972—1986 Selostus: Lisäpellon tuotto- ja kauppa-arvon määrittämisen perusteet ja soveltuvuus pellon arvon osoittamiseen Etelä-Suomessa vuosina 1972—1986 MATTI YLÄTALO Pellervo Economic Research Institute Revontulentie 8 A SF-02100 Espoo, Finland Academic Dissertation To BE PRESENTED, WITH THE PERMISSION OF THE Faculty of Agriculture and Forestry of the University of Helsinki, for public criticism in Auditorium Porthania hi on October 25th, 1991, at 12 o’clock. SUOMEN MAATALOUSTIETEELLINEN SEURA • HELSINKI https://www.c-info.fi/en/info/?token=ETl60UmLFH5nHjyM.aS4E3bVOrEnvDZhlafZUtg.u18Kgt8s3PBCpFFJ-WS7AsYXySRZ4YbW7-o5JpexsotpM7fRlnB6oRa8Phqni8ipC8kaYsHVHLhtuOMescIbAcoXT36OzOb5BuRTrvIK1Py22hzT5ZmwXpamvlLPdN06RH_4nwMQKUO2jNg9HClyOgBaODV5HLTZ3M80jztjT88 145 PREFACE With the completion of this thesis, I wish to acknowledge all those individuals from whom I have received encouragement and support throughout the various phases of my work. The study was begun in theDepartment of Agricultural Economics at the Univer- sity of Helsinki, where I held the post of Junior Researcher, financed by the Acade- my of Finland. In 1984, I moved to Pellervo Economic Research Institute (PTT), where the theoretical part of my study was completed. I was able to concentrate on the empirical part of the study while serving as an Assistant in the Department of Agricultural Policy at the University of Helsinki, in 1989. First, I wish to extend my warmest thanks to Professor, the Department Head Viljo Ryynänen, in the Department of Agricultural Economics, for his guidance throughout the study. My interest in the appraisal of real estate, especially concerning agricultural land, is largely due to his influence. The Managing Director of PTT, Professor Pertti Kukkonen, as my main academic advisor, has provided valuable comments which have assisted me in improving the empirical analyses, as well as sharpening the English terminology. To him, and to my second academic advisor, Professor Matias Torvela, Director of the Agricultural Economics Research Insti- tute, I extend my gratitude for their cooperation. I also extend my thanks to the Vice Rector of the University of Helsinki, Profes- sor Risto Ihamuotila for his continuing interest in my work. Similarly, Professor Kau- ko Hahtola has offered critical advice concerning the theoretical aspects of my work. Assistant Professor Ossi Heiskanen and the Department Head in theAssociation of Agricultural Advisory Centers, Mr. Mikko Siitonen, have also helped me with their constructive discussions. Mr. Arto Latukka and Mr. Timo Räsänen have helped me in the collection and preparation of the empirical data from bookkeeping farms. They, together with the Agricultural Profitability Office Directors, Mr. Heikki Järvelä and Mr. Juhani Iko- nen, earn my deepest gratitude. I also thank my colleagues at PTT, especially Re- search Directors, Mr. Erik Haggren and Mr. Seppo Aaltonen, Secretary, Mrs. Ritva Tuutti, and Mr. Perttu Pyykkönen for their help and support at various stages of this work. The original Finnish text was translated by Miss Ulla Palomäki and checked by Dr. Ashley Selby and Ms. Terese Forster, to whom I extend my thanks. My research received financial support from the Kyösti Haataja Foundation and the Finnish Cultural Foundation (Veikko Ihamuotila scholarship). Further financial support was received from the Academy of Finland and PTT. I am most grateful for the support of these organizations. I also thank the Scientific Agricultural Socie- ty of Finland for permission to publish this study in its journal. Finally, I thank my wife Marja for her support, understanding, and tolerance with respect to my thesis. Helsinki, August 1991 Matti Ylätalo 147 Determination of the capitalized and market values of supplementary arable land in southern Finland, 1972—1986 Contents 1. INTRODUCTION 151 1.1. Land and its special properties 151 1.2. Essential concepts of value and price in the appraisal of agricultural land 152 1.3. Aim of the study 154 2. FACTORS INFLUENCING THE VALUE AND PRICE OF SUPPLEMENTARY ARABLE LAND 156 2.1. The relative value of arable land 156 2.2. Demand for and supply of land 159 2.3. Technological change 161 2.4. Other factors 164 3. PREVIOUS STUDIES BASED ON THE CAPITALIZED AND MARKET VALUE AP- PROACHES 167 3.1. Studies based on the capitalized value approach 167 3.1.1. Basic principles for the determination of capitalized value 167 3.1.2. Land value based on rent 168 3.1.3. Determination of the value of supplementary arable land by the residual approach 170 3.1.4. The value of supplementary arable land based on its marginal productivity ... 172 3.2. Studies based on market value 174 3.2.1. A general survey of price investigations concerning agricultural land 174 3.2.2. Price investigations concerning supplementary arable land 177 4. DETERMINATION OF THE CAPITALIZED VALUE OF SUPPLEMENTARY ARABLE LAND 180 4.1. The research region 180 4.2. Agricultural conditions during the research period 181 4.3. Choice and extent of data 182 4.3.1. Private economic reasons for acquiring supplementary arable land 184 4.3.2. Measuring the return from supplementary arable land 187 4.4. Production function analysis 188 4.4.1. General grounds for the use of production functions 188 4.4.2. Selection of variables 189 4.4.3. The functional form of the production function 191 4.4.4. Results of the regression analysis 192 4.5 Returns from supplementary arable land 196 4.5.1. Returns from supplementary arable land by farm types 196 4.5.2. Return from supplementary arable land as a function of the size of farm .... 199 4.6. Assessment of the capitalized value of supplementary arable land 202 4.6.1. The selection of the capitalization rate 202 4.6.2. Capitalized value of supplementary arable land 204 148 5 MARKET VALUE OF SUPPLEMENTARY ARABLE LAND AND ITS DEVELOPMENT IN THE INVESTIGATED REGION 207 5.1. The collection and extent of data 207 5.2. Price level of supplementary arable land and its development 210 5.2.1. Questions concerning the use and assessment of data on market prices 210 5.2.2. Price of supplementary arable land and its development according to the market price register 212 5.2.3. The price of supplementary arable land and its developmentaccording to purchases of arable land by the National Board of Agriculture 216 5.2.4. Price differences between transactions of supplementary arable land in the market price register and the purchases of arable land by the National Board of Agriculture 218 6, COMPARISON BETWEEN CAPITALIZED VALUE AND MARKET PRICES OF SUP- PLEMENTARY ARABLE LAND 221 7. RESULTS AND CONCLUSIONS 224 7.1. Return from supplementary arable land 224 7.2. Market prices of supplementary arable land 226 7.3. Correspondence between the capitalized and market values 227 7.4. Conclusions 230 8. SUMiMARY 232 REFERENCES 236 APPENDICES 241 SELOSTUS; Lisäpellon tuotto- ja kauppa-arvon määrittämisen perusteet ja soveltuvuus pellon arvon osoittamiseen Etelä-Suomessa vuosina 1972—1986 252 JOURNAL OF AGRICULTURAL SCIENCE IN FINLAND Maataloustieteellinen A ikakauskirja Vol. 63: 149—254, 1991 Determination of the Capitalized and Market Values of Supplementary Arable Land in Southern Finland, 1972—1986 MATTI YLÄTALO Pellervo Economic Research Institute Revontulentie 8 A, 02100 Espoo, Finland Abstract. Rapid technological change in agriculture has permitted the management of larger farms with an existing labour input. Following this the demand for supplementary arable land has grown considerbly, while the supply has remained low. This imbalance between demand and supply has been reflected in rising land prices. For that reason, the main aim of this study was to determine the capitalized and market value of supplementary arable land and the corre- lations between them. The determination of the capitalized value of supplementary arable land is based on the use of production function analysis. The marginalvalue product was estimated from a Cobb-Douglas function, using as independent variables, the production inputs, and a dummy variable for the quality of land. The present value of arable land was calculated by capitalizing the net marginal operating margin at interest rates of 3 %, 5 Vo, and 7 % for durations of return of 5, 10, and 15 years. The benefit obtained from supplementary arable land proved highest on cattle farms with a small arable area, while on grain farms the marginal value product and the differential re- turn still remained high as the size of the farm increased. The real price of arable land in the annual purchases of land by the National Board ofAgricul- ture rose very slightly, on an average of 1.7 % per year, during the period 1972—1986. The capitalized value on the cattle farms reached the prices paid for arable land by the Na- tional Board of Agriculture, when a duration of return of 15 years and a capitalization rate of 3 % was used. In contrast, the capitalized value of the grain farms did not reach, at any time during the study period, the price level of the National Board of Agriculture. Also, the market prices of arable land, according to the data collected by the National Board of Land Surveying, clearly exceeded the capitalized value of both cattle and grain farms, during the period 1982—1986. Index words: Supplementary arable land, marginal value product, capitalization rate, capitalized value, market value 149 1. INTRODUCTION 1.1. Land and its special properties Land is a natural factor of production that is the basis of human life and activity. Land is destined for production, housing and re- creation, while the raw materials used by the mining industry and for crop cultivation are derived from it. These raw materials provide the base for the production of more sophisti- cated goods. For agricultural practices, land is the most important and qualitatively the most stable factor of production. Land is a necessary ba- sic element, along with water, air, light and heat, for the satisfaction of different human needs. Land is also an object of possession and has certain special characteristics that most other goods lack. These characteristics are: 1. Land area is limited and it cannot be in- creased. 2. Land cannot be relocated. 3. Land cannot be entirely destroyed. 4. The productive capacity of land is limited. Land’s special attributes contain important implications. Since the land area is limited and cannot be relocated, a lack of land is created, especially around large population centers. On the other hand, the area of agricultural land is elastic, to a certain extent, because it can be increased by land clearing or by draining wetlands. In some cases it is possible to move land blocks from one place to another. Yet, this is expensive and often impossible. The ground always remains where it is. In the same way, land can be reclaimed from lake or sea beds, but this method of land acquisition is of mar- ginal significance, at least in Finland. Accord- ing to Aereboe (1923, p. 21), land destruc- tion extends, in the first place, to sand, clay, humus, lime, and large and small stones com- posing the superstratum. Natural conditions and cultivation methods can change the proportions of the above- mentioned substances, and yet the ground is permanent and cannot be destroyed. For this reason, Goltz (1905, p. 18) separates from land all factors of production that change ei- ther immediately or when being used. These are food-stuffs, firewood and stocks. The ex- ploitation of buildings also ends after a cer- tain time. By the limited productivity of the soil, Goltz (1905, p. 19—20) means the reduction of the productivity of the soil if it is perpetually used for crop production with- out adequate fertilization. According to Siebert (1969, p. 40), there are no production costs related to land be- cause of the indestructibility of the soil. Wen- trup (1978, p. 8) considers this view quite precarious, and argues that it is uncultivated land that implies no cost. Putting land to eco- nomic use, that is to say converting it to ara- ble from the uncultivated state, almost always produces costs. These costs are so closely related to land that they must be considered as “belonging to the land”. The economic measures applied to land are scarcely arith- metically discernable and, thus, difficult to separate from the price of the land. Although land’s special characteristics are not absolute, they tend to make its supply in- elastic. As land cannot be increased by pro- duction, the determination of its price is not the same as with other mass production goods. 151 Consequently, according to Cassel (1938, p. 276), land price is a secondary product of its own yield. The productivity of land cannot be infinitely increased by intensive methods, and yet, because of population growth, more and more submarginal land must be cultivated. As a result, the decrease of available land can lead to overpopulation. This is why some re- searchers in agricultural economics consider that agricultural land possesses monopolistic characteristics, ifonly partially, compared to other capital (e.g. Renne 1947, p. 9). On the other hand the more recent literature (e.g. Feuerstein 1970, p. 4) rejects the concept that the inelasticity of supply results in a situ- ation in which only demand determines land price (compare Nositschka 1973, p. 11). The concept of land is very extensive. A narrow definitionincludes only the solid part of the land surface. Barlowe’s (1958, p. 7) economic definition sums the capital jointly created by nature and man upon the land sur- face. This economic concept is comparable to Gryst’s and Timmons’s (1961) definition in which land resources are defined by the fol- lowing characteristics: 1) natural characteris- tics, e.g. soil and climate; 2) characteristics created by society, e.g. location and civil en- gineering: road networks, drainage and flood protection; and 3) investments closely related to the soil, such as terracing embankments and measures to improve fertility. Timmons (ref. Reynolds 1966, p. 5) includes in the concept of land resources both the underground re- sources and surface resources. From the ju- ridical point of view, land signifies any sur- face part of the land which can be an object of proprietary rights (Barlowe 1958, p. 7). There are different ownership concepts con- cerning this kind of good. One of the most familiar is the socialistic view claiming that land, as a factor of production, must belong to the state. Opposed to this is the capitalistic view which allows and protects private prop- erty. Between the extremes there are types of joint ownership, in which a family, village, or small community may possess land areas for common use. On the other hand, the content of the property concept has changed over time, along with social development (Hyvö- nen 1982, p. 58—69). In most cases, bringing arable land into cul- tivation involves considerable costs. In agricultural economics, there has been con- siderable debate on whether the cost of land clearing accrue to the actual value of agricul- tural land or whether it belongs to land im- provements (e.g. Laur and Howald 1957, p. 21 —22). Ihamuotila (1983, p. 10) claims that clearing costs should be included in the land value, because the value added clearing creates is stable, and thus comparable to the soil value. Due to its special attributes, land differs from other means of production in agricul- ture, since land is able to produce crops and services continuously. If managed properly, its value is retained and no depreciations are made from the land (Laur 1928, p. 85). There are no maintenance costs related to land. On the other hand, agricultural land is closely related to land improvements, to build- ings and to expected harvest. The separation of agricultural land and property has not al- ways been apparent and incontestable (e.g. Thaer 1880, p. 19, Aereboe 1919, p. 51 —52, Laur 1928, p. 17). With the development of agricultural economics terminology, confu- sion between land and other means of produc- tion is disappearing. At present, land improve- ments include only such land-related works that have to be renovated from time to time and, therefore, their value must be depreciated (compare Mäki 1964, p. 114). 1.2. Essential concepts of value and price in the appraisal of agricultural land In the economic literature, the concept of value is often divided into objective and sub- jective values. The classical school determined the objective value of goods by reference to the production cost. According to this theory, the value of a commodity, for example land, is defined by the highest production costs. That is to say, by those production costs creat- 152 ed under the most unfavourable conditions (e.g. Ricardo ref. Conner 1937, p. 37, Thu- nen 1921, p. 106, Marshall 1920, Clark 1923 and Cassel 1938). The objective value is often identified by the price of the com- modity, even if there is no direct exchange of commodities. Accordingly, Virolainen (1950, p. 21), in his extensive investigation of the economic concepts of the value and price of agricultural land, formulates the objective value as a price relation that indicates the re- lation between the price of a commodity and the prices of other commodities. Similarly, Vartiainen (1963, p. 135) argues that the value of a commodity can be signified by an- other commodity as a unit of measure. Thus, value is defined as “the quantity of the sec- ond commodity in exchange to one unit of the first commodity”. As a result, the objective value is the economic value measured in a generally accepted way. The Association of Finnish Real Estate Valuers (Anon. 1986, p. 15) defines objective value as “the value de- fined at the moment of appraisal, based on the generally accepted characteristics of the object”. Many publications of agricultural economics have dealt with objective values and their applicability to real estate econom- ics (e.g. Mäki 1945, p. 45, Ryynänen 1967, p. 40). The theory of subjective value, on the other hand, aims to determine the significance of commodities for the satisfaction of human needs (compare Virtanen 1966, p. 18). Such values can be based on sentimental motives (Mäki 1964, p. 102). Barlowe (1958, p. 83) states that political, mental, social, and spiritual values are strongly subjective. Sub- jective values are therefore not measurable in a generally accepted way. Consequently, their study is difficult and the results often ar- bitrary. According to Kantola (1977, p. 7—B), there are also static and dynamic values. The former is defined as the capacity of a com- modity to satisfy human needs, as well as being concerned with the realization of benefits, while the latter is concerned with time scale. When conditions at the moment of appraisal are of greatest importance, pri- ority is given to the static value. On the other hand, when the succeeding conditions already have an impact at the moment of appraisal, the condition must be considered to be dy- namic. According to Kantola, these probable value effects must often be anticipated. Vartiainen (1963, p. 135) claims that the market price of goods corresponds to their ex- change or value relations. Though value and price are different concepts, they have a cer- tain similarity. Accordingly, Wiiala (1976, p. 4) suggests that the value of different proper- ty items intends to express their significance to the farmer as a means of production. On the other hand, price signifies payments al- ready made, or to be made, for the transfer of a real estate, or monetarized compensation for it. Value and price and their relations have been examined more closely by Ahonen (1970, p. 10—14). In appraisals of the farm economy, the con- cept of agricultural land is very common. Mäki (1964, p. 110) includes in agricultural land all plots that are used for crop cultiva- tion and livestock husbandry. Agricultural land is divided into the following types of land use or farm types: garden, arable land, meadow, and pasture. Except for the soil, different rights and usages belong to it, such as rights to use common landand water areas. Also considered to belong to agricultural land are agricultural building sites and farm tracks, as well as clay, mud and peat extraction sites. The practice followed in the property taxation system in agriculture (Anon. 1967, 34 b §) and in the Finnish official profitability survey differs partly from the concept of agricultural land defined by MÄKI. For this reason, the practice of profitability accounting will be dis- cussed more in detail, together with the presentation of the empirical methods. In the appraisal of agricultural land, the de- termination of the relative value must be dis- cerned from the determination of the absolute or the monetary value. The relative value sig- nifies land value compared to the average or 153 the best land. On the other hand, the mone- tary value is based on demand and supply. It can also be based on the impact of a part of property on production activities. On this basis, exchange, capitalized and cost values are determined (e.g. Elstrand and Sonju 1978, p. 20). Gustafsson et ai. (1978, p. 8) consider the capitalized value and the market value, evalu- ated by using the selling prices of land, as the principal concept of value. Additional con- cepts of value included in appraisals are: ex- pectation value, sale value and taxable value (Virtanen 1967, p. 18, Wiiala 1976, p. 3—9). Agricultural appraisals also consider the concepts of use value and cost value (e.g. Elstrand and Sonju 1978, Elstrand 1980, Rasmussen 1981). The capitalized value of land signifies the present value of future returns based on the use of the land (Found 1971, p. 23—24). Ac- cording to the Association ofFinnish Real Es- tate Valuers (Anon. 1986, p. 16) capitalized value means “present value defined as capi- tal value of the return corresponding to the present use or possibilities of use”. According to Gustafsson et ai. (1978, p. 9), capitalized value distinguishes between the individual and the mean value. The individual value is based on the profit produced by a real estate which benefits an individual. The lat- ter concept defines the return gained under average circumstances. In addition, the capi- talized value can be analyzed according to whether the effects of loans and taxes are taken into consideration in the estimation. Use value is closely connected with capitalized value. Following Elstrand (1980, p. 17), use value can be said to have a broad and a nar- row meaning. In the usual narrow meaning, it is considered to be a subjective value re- sulting from the agricultural use of the real estate or a part of it (compare Wiiala 1976, p. 5—6). Cost value corresponds to the cost occasioned by the production or acquisition of the property or by another action, Wiia- la (1976, p. 40). This method can determine the value of agricultural land, when the clear- ing costs of the land area are known. Market price is the indicator of the general exchange value. Market value is defined by the Association of Finnish Real Estate Valuers as current value, or current price, fixed in refer- ence to market prices (Anon. 1986, p. 14). Ryynänen (1967, p. 41) says that, in practice, the average price of land and farms tends to be determined by choosing the largest possi- ble sample of farms for sale or unconstructed areas that have been and will be in agricultural use, and which are typical of the area. Using the prices of such farms and lots, the average unit prices of different land use types can be calculated. Other types of exchange values can also be calculated, for example, market prices paid for land in cases where the real estate is sold for special purposes or where marketing occurs under special conditions (e.g. Gus- tafsson et ai. 1978, p. 9). The main concepts in this study are the capitalized value and the market value. The definition of the contents follows the above- mentioned definitions used in the vocabulary of the Association of Finnish Real Estate Valuers. Supplementary arable land means arable land that has already been acquired or will be acquired and that is used or will be used for agricultural production. Thus, the capi- talized value of supplementary arable land is the capitalized value of its additional return, determined arithmetically. The determination and calculation of return and its capitalized value will be clarified in more detail later. Moreover, the market prices of supplementary arable land indicate their market value. 1.3. Aim of the study Land, together with labour and capital, are factors of production that are essential to crop cultivation. From the standpoint of the pri- vate economy, agricultural land is often regarded as capital (e.g. Virolainen 1950, p. 25, Mäki 1964 p. 68), because creating ara- ble land necessitates investment-like financial sacrifices. Rapid technological change has permitted 154 the management of larger farms with an un- changed labour input (e.g. Johnston and Bischoff 1971, p. 124, Torvela and Mäki 1974, p. 70). Simultaneously, technological change in agriculture has allowed the substi- tution of arable land by other factors of pro- duction, so that the relationships between agricultural land and other means of produc- tion can change considerably. Technological change is considered to increase the economic use of different means of production, even if the impacts of the change are not well known (compare Griesbach 1966, p. 114—116). Very often in Finnish agriculture, where small farms are predominant, an increase of arable land is regarded as an essential pre- requisite for improving the profitability of farming. The economic results will be im- proved if the return from supplementary ara- ble land is higher than the costs occasioned by its acquisition and cultivation. The acqui- sition of supplementary arable land can be ex- pected to improve the profitability increasing- ly within the limiting effect of the arable land area with respect to other factors of produc- tion on the farm (machines, buildings, labor). Yet, it is not profitable to acquire supplemen- tary arable land at all costs. The principal aim of this study is to deter- mine the capitalized and market values of sup- plementary arable land in the bookkeeping re- gion of southern Finland during the period 1972—1986. The development of capitalized and market values and their correlations will also be addressed. The study consists of a theoretical and an empirical part. In the first part, factors in- fluencing the value and price of supplemen- tary arable land are examined. Previous studies dealing with capitalized and market values, as well as their principal methods, will be examined too. In the empirical part of the investigation, the capitalized value of supplementary arable land is determined by using information from the bookkeeping farms of southern Finland that participate in the official profitability sur- vey in agriculture. The evaluation of the capitalized value is based partly on cross sec- tion and partly on time series analysis. Farms specializing in cattle and grain production are evaluated. The agricultural practices of these farms are based, to a large extent, on the ex- ploitation of agricultural land and not on an intensive use of purchased production inputs, as in the case of pig and poultry farming. The study aims to determine the dependence be- tween capitalized value and farm size. Infor- mation on the market price of supplementary arable land has been collected from the sale price register of the National Board of Land Surveying and from the statistics of the Na- tional Board of Agriculture. On the basis of these price statistics, a general picture is given of the market price development of supple- mentary arable land, as well as the level to which the capitalized values determined from the bookkeeping farms are compared. 155 2. FACTORS INFLUENCING THE VALUE AND PRICE OF SUPPLEMENTARY ARABLE LAND 2.1. The relative value of arable land Climate and soil parameters are of central importance to the growth of all plants, in- cluding crop cultivation. In this study, factors connected with the soil are of primary con- sideration, even though the impact of climatic factors are fully recognized. The classification of arable land under cul- tivation can be made by assessing the land ac- cording to the physical characteristics of its soil (a point value or index system). Moreover, topographic and climatic factors can be in- cluded in the land classification. Other fac- tors, such as the proximity of built up areas, availability of irrigation water, etc., can also have an impact on this classification. In any appraisal, the return already realized can be employed, instead of using only the produc- tion potential. According to Busch (1969, p. 96), many different criteria serve to classify land. Among them are: average yields natural vegetation characteristics derived from the soil texture and its condition different combinations of the above-men- tioned criteria. Appraisals based on the characteristics of the soil are, in practice, preferred to the meth- od based on crop yields, because the intensi- ty ofcultivation does not directly correspond to the production potential of the land (Busch 1969, p. 97). In Finland the relative value of arable land has been used in agricultural bookkeeping taxation and acts of surveying. The relative value can also be used to fix the monetary value of arable land, because relative figures indicate differences between plots. According- ly, agricultural bookkeeping has used the transformed quantity of hectares of arable land to measure farm size. It is obtained by adding the sum of the arable land and garden areas to the area of meadow and pasture, which have been transformed to correspond to the mean value of the arable land (Anon. 1986, p. 12). Since 1976, in the bookkeeping farms, the arable land area in use is considered the farm area. The determinationof the relative value for taxation and acts of surveying is called grad- ing. Grading is considered by Wiiala (1958, p. 19) to be a relative appraisal of farm lots, by which means the distinct objects of ap- praisal are ranked according to their value by a roughly agreed ratio or grading scale. In cer- tain cases and with certain grading classes, grading corresponds to the land classification. Kulmia (1943, p. 163) claims that grading must be based on the constant capitalized val- ue of the farm lot. This is achieved by taking into consideration the basic condition of the land and its utilization. Moreover, Aalto (1951, p. 104) argues that when grading ara- ble land according to properties which in- fluence its production potential, the physical and chemical properties, i.e. the foundations of the production potential, must be consid- ered. A summary of the significance of soil textures on land classificationand its value re- lations is shown in table 1. 156 Table 1, The relationship between soil textures and different properties (Vuorinen 1952, p. 422 —423). Soil texture Moisture Air Technical content capacity Fertility properties I Med. coarse sand 0 10 2 3 Very fine sand 3 113 8 Silt 1 0 2 0 3 Heavy clay 2 2 3 1 8 According to the table 1, very fine sand and heavy clay obtain the highest points, and medium coarse sand and silt the lowest. A humus content is essential to site quality, and its increase improves the aeration of silt, im- proves water retention by sand and facilitates the tillage of mineral soils, as well as reducing the risk of desiccation. Thus, the humus con- tent reduces the differences between soil tex- tures. The soil structure greatly affects the vitali- ty of plants. Because soil structure is depen- dent upon both the chemical and biological properties of the soil, the relative value of the soil can be determined by these properties. The decrease in the natural fertility of the soil emphasizes the importance of the physical properties in the measurement of the cultiva- tion value of arable land. The grade of arable land under cultivation is a reliable expression of its relative quality. Land possessing a high grading value produces higher yields than land of lower quality. This is shown by the interdependence between the quality of land and the average arable land (Ryynänen 1962, p. 123—124 and 131). The quality (quality index) of the investigated ara- ble land and the feed-unit yield determined in a given way, serve to calculate the regression line y = a+ bx, that gives the production ca- pacity of arable land of each quality class (fig- ure 1). Figure I. Quality index of arable land and feed unit yield of the investigated farms 157 The estimated regression equation y = 930 + 7.85 x means that for the quality index inter- val (60 —133), there was a yield increase of 7.8 fu/ha for each one point rise in the quality index of arable land. In this study, the coeffi- cient of regression was statistically significant. Moreover, the choice of crops can have an impact on yield and, consequently, on the eco- nomic return produced by supplementary ara- ble land (Blohm 1966). This depends, in the first place, on the operating margin of the crop on different qualities of arable land. Then, the fixed costs determined by the culti- vation of different crops play a central role. The higher the fixed costs of a crop, the sooner will its cultivation become unprofita- ble on poor soil (figure 2). In figure 2 root crops are profitable on the best soils. The fixed costs in grain cultivation are lower than in root cultivation. Thus, grain cultivation is possible even on rather poor land. These costs are lowest in extensive pas- ture cultivation. For this reason, poor soils are the most suitable for pasture cultivation. According to Nositschka (1973, p. 81 82), the price of supplementary arable land in West Germany was dependent on its suitabil- ity for the cultivation of various crops. If the price of the arable land suitable for pasture cultivation was set at 100, the ratio of arable land unsuitable for wheat cultivation was 160, that suitable for wheat cultivation was 171, and that suitable for root crop cultivation was 234. Similar conclusions were also made in Finland on the basis of local land price inves- tigations made by the Department of Agricul- ture Economics of Helsinki University. They revealed a clear correlation between arable land price and land quality. For example, prices for supplementary arable land in the communes of Elimäki, litti and Valkeala, in southeastern Finland, indicated that arable landprices under 100 tax points were less than 70 °7o of a mean price of the best arable land (> 100 points) (Mäkelä 1977, p. 36—38). In some communes in southwestern Finland the price of the poorest arable land (P y To obtain a state of equilibrium, more labour and capital could be used in the short run. This results in a decrease of the mar- ginal value product and in an increase of the marginal cost, until the following situation (point A’): MPPm MPP, MPPk 1 1M ivirrL ivirrK F, MC y PyP M Pl The total output increases while prices re- main constant. In the long run, the firm has an incentive to acquire more land, until the marginal value product of the land decreases or land price increases. If several firms expect the same situation and seek to acquire land, land prices will rise. Thus, the individual firm, as well as other firms, will drift into a situa- tion (A”) where there is no profit. Herdt and Cochrane (1966, p. 245—248) claim that such a modelpresupposes some re- quirements essential to agriculture: land price rise, rapid and wide-spread technological changes, reduction of the quantity of total labour, continued increase in output, expand- ing activities of firms, and increase in farm income despite the fact that the income per production unit derived from farming remains constant or falls. Technological change in Finnish agriculture manifests itself not only as a general intensifi- cation of production, but also as a growth in the use of factors and inputs of production acquired from outside the farm. Consequent- ly, the capital stock per farm has considera- bly increased (Ihamuotila 1983, p. 97). At the same time, the contribution of agricultural land to capital stock in agriculture has con- tinuously increased during the 1960’s and 1970’5, to reach 41.2% in 1980, at current prices (Ihamuotila 1983, p. 100). The sig- nificance of agricultural land is, thus, central to agricultural practices in Finland, in spite of the utilization of machines as a substitute for labour, and chemical and biological technol- ogy as substitutes for land. 2.4. Other factors The price of supplementary arable land de- pends on many factors, but its determination and quantification is problematic because the annual sales of land represent only a minor part of the total agricultural land area. De- mand applies, in the first place, to lots of arable land and forest. The changes in farm ownership usually occur as transfers between 164 parents and children, in which case the prices of farms are considerably lower than the prices of supplementary land. The price of additional land, as well as that of agricultural land, as parts of a whole farm, are influenced by all those factors that arise from agricultural practices, its organization and profitability. Mäki (1964, p. 143) has classified these factors as follows: 1. General factors having an effect on land values in certain regions. 2. Factors having an effect on the valueof an individual farm. 3. Factors having an effect on the value of an individual plot. To determine the conditions of appraisal of a certain area, the average price level of agricultural land in that area must be assessed. Several different factors are implied, among which the most important are, according to Ryynänen and Pölkki (1982, p. 71): 1. Climate 2. Land fertility 3. Location and the economic and technical development of the agricultural environ- ment 4. Agricultural produce marketing and price relations 5. Availability of labour 6. Aesthetic quality of the area 7. Possibilities for education and recreation 8. Other factors With most of these factors their effect on land value cannot be clearly identified. Their effect appears as a price level, to which parts of a farm can be compared. At the same time, it is necessary to determine the average price level of different land use types in the evalu- ated area, before an evaluation concerning an individual farm or plot can be accomplished. Virtanen (1979, p. 58) divides the factors influencing land value according to their na- ture into physical, juridical and market fac- tors. Among the physical factors are soil and topography. Juridical factors include plans or other regulations that indicate how a real- estate must be employed and what buildings are permitted. The most typical market fac- tors are demand and supply and the local price level. Jorgensen and Jorgensen (1971, p. 9—15) divide the factors influencing land price into three groups: general factors, quality factors, and individual factors. The main features of this grouping are shown in table 3. Table 3. Factors influencing the price of agricultural land. General factors Quality factors Individual factors Inflation Quality of the soil Buyer’s conception of the location of the Possibilities of Size of the farm farm and its making a living suitability for in agriculture Relations between different usages plots Possibilities of Location of the Buyer’s and seller’s making a living farm conception of the elsewhere possibilities ofmaking a living General interest Wealth of the buyer rate and infla- tion and gene- Willingness to retire ral economic or continue farming activity Number of farms Seller’s age and family situation Measures of pub- lic authorities The relation between the variables affect- ing land price can be very complex, especial- ly between the general factors. The distinct ef- fect of each of the contributing factors is dif- ficult to define. Jointly, the factors contrib- ute to the conditions in each area and to the possibilities of making a living in agriculture and, thus, to the price level of land. Moreo- ver, there are interrelations between the main groups which affect the individual factors in- fluencing the ability to make a living from agriculture. According to Wiiala (1976, p. 90), the most important factors related to the value of 165 an individual lot of arable land are the fol- lowing: 1. Natural fertility 2. Local position 3. Size of the field 4. Shape and other factors related to value These factors have an impact on the yield derived from arable land and, consequently, on land price. In particular, roads and popu- lation centers contribute to fluctuations in land price. Crowley (1974 a, p. 7) has argued that the proximity of highways does not have an impact on land price. Similarly, he argued that the location of population centers in re- lation to the sold land does not have an ef- fect on land price. Virolainen (1950, p. 160—177), on the other hand, has presented evidence to suggest that accessibility has an effect on the relative value of arable land. The further the arable land was located from population centers (e.g. Helsinki, Turku, Tampere, Lahti, and Mik- keli), the lower was its relative value. Nosit- schka (1973, p. 91 —95) has shown that the location of the land with respect to the road network has an impact on its price. The price rise was explained as a consequence of an ex- pected increase of land value in anticipation of its use for construction purposes near popu- lation centers. Hovi and Jokinen (1974, appendix 8/1 4), following the methods developed by Kan- tee, have calculated transportation costs per hectare in different cultivation tasks, arising from the distance between the field and the farmstead. The following figures show the transportation costs per hectare arising from the cultivation of different sizes of field, in relation to their distances. Costs are repre- sented by an interval between a minimum and a maximum cost, calculated according to the quality of the road: Distance from the Field size Transportation farmstead km ha costs FIM/ha 1 1 35 50 1 3 20— 39 1 10 18— 35 3 3 40—137 8 3 114—316 8 10 101—284 There is a tendency to acquire additional land at the shortest distance possible from the farmstead, following which the additional area to be bought is often bordering the ara- ble land of the purchasing farm. Because of the increase in arable area, cultivation costs per unit area decrease. Thus, with respect to additional land, the distance of the fields from the farmstead does not influence the transpor- tation costs as significantly as Hovi and Joki- nen claimed. This contention is supported by Nositschka (1973, p. 103—105) who ob- served that such costs did not raise prices in transactions which contributed to improve- ment of the organization of plots. The effect of the distance between the farmstead and fields on the organization and profitability of agriculture and, consequently, on the prices paid for land, are demonstrated by the fol- lowing figures. When purchased lots were divided in groups according to the distance be- tween the farmstead and fields, the average price paid per hectare of arable land decreased with increasing distance from the farmstead: Average price DM/haDistance from the farmstead km 18 620 16 350 II 400 under I I—s over 5 Purchasers were willing to pay 60 % more for arable land near the farmstead than for arable land at a distance of more than 5 km. 166 3. PREVIOUS STUDIES BASED ON THE CAPITALIZED AND MARKET VALUE APPROACHES 3.1. Studies based on the capitalized value approach 3.1.1. Basic principles for the determination of capitalized value The value of arable land is based on returns derived from it. For the appraisal of land value several models have been constructed which show possible variations or stability of returns annually derived from land (e.g. Nie- hans 1966, Reynolds and Timmons 1969, Gustafsson et ai. 1978, Aalstad et ai. 1979). Common to these models is that land value is based on the capitalization of returns de- rived from land in the future. In their simplest form the models suppose that returns are ex- pected to remain unchanged indefinitely. Thus, land value V is determined as follows: V = R/r, in which R =return from land and r = interest rate. Land value increases with both increasing return and decreasing capitalization rate. With low interest, even a small capitalized return indicates a high land value, and vice versa. Ac- cording to Mäki (1964, p. 107), the difficul- ties related to the assessment of expected returns result in a more significant error than the method of appraisal itself. Feuerstein (1970, p. 75—77) criticizes the above model because of its fixed return and unchanged in- terest. He tried to demonstrate, using various appraisal methods, the unsuitability of the model for expressing the average exchange value of real estates in agriculture and for- estry. Crowley (1974 b, p. 54) admits that the model can only be used in situations where the investment period is either infinite or of very long duration. The capitalized value has been severely criti- cized when used for the determinationof the value of arable land or of a farm as a whole. Aereboe (1919, p. 244—256) already doubted whether the gross return and costs can be as- sessed correctly and accurately in practice. Of- ten the determination of commodity prices re- mains uncertain, as does the interest rate to be used in the appraisal. Hjelm (1952, p. 2) emphasizes the theo- retical character of capitalized value which is, in principle, an important factor in the deter- mination of the market value of a real estate, but is difficult to express numerically. El- strand and Sonju (1978, p. 32) show that the capitalized value of arable land depends on the following factors: a) Yield level b) Production efficiency c) Type of farming d) Personal characteristics of the farmer. Elstrand and Sonju (1978, p. 42—56) stress the importance of inflation and the in- terest derived from alternative investments, in the choice of the interest rate to be used in ap- praisals of capitalized value. Kaarlehto (1954, p. 67), on the other hand, considers that the determination of commodity price levels is one of the most difficult tasks in the income approach; the factors having most in- fluence on the price development are general economic conditions, official agriculture poli- cies, plant breeding, weather, production 167 quantities, marketing conditions, and special- ized commodities. The capitalized value approach has been lit- tle used even for the determination of the val- ue of lots (Kanerva 1978, p. 9—12). The rea- son for this has been the low real interest on loans and the rise of the real prices of lots. The real interest, i.e. the difference between the average interest on deposits and the de- preciation of the valueof money, was approx- imately 2.2% during the period 1963— 1977. If the loan capital is invested in schemes that conserve their real value, the borrower can benefit. Concerning land sales, capital gain is a commonly used concept (e.g. Rey- nolds and Timmons 1969, p. 331, Plaxico and Kletke 1979, p. 327—330, Castle and Hoch 1982, p. 8—18). Thus, the rising value of arable land is a capital gain rather than a direct income to the landowner. During a pe- riod of rising land values, it seems reasonable to hypothesize that the anticipated apprecia- tion of land values (expected capital gains) has had an impact on land value. Consequently, a person acquiring land might have been more interested in the financial organization of the agreement than in the appraisal of the capital- ized value. Capitalized value depends essentially on the choice of the interest rate used in capitaliza- tion. Several researchers emphasize the impor- tance of the choice of the right interest rate when assessing land value. Virolainen (1950, p. 33) claims, on the basis of work done in the U.S.A., that the taxable net return has to be capitalized at an even higher interest rate than the industrial dividends given to invested capital. Niehans (1966, p. 195) emphasizes the importance of the characteristics of invest- ment projects in the determination of the in- terest rate. Then, the degree of liquidity (a risk related to the investment project), the alter- native use of capital, and the loan interest must be taken into consideration. Scofield proposes that “the rate should represent the prevailing opportunity cost of capital as de- termined by the rate of return, after taxes, that could be realized from other investments having the same liquidity and risk characteris- tics as farmland”. According to him, “non- farm income-producing real estate such as apartments and office buildings and common stocks are most clearly comparable to farm- land in an investment sense” (Scofield 1965 a, p. 101). Elstrand and Sonju (1978, p. 43—56) present five differentcases for the determina- tion of the interest rate. As well as the previ- ously presented views on this subject, they em- phasize the usefulness of both the long term and real interests for land property when as- sessing agricultural investments. According- ly, the choice of the interest rate for assess- ment affects the assessment’s applicability. 3.1.2. Land value based on rent In economics, land used for cultivation has been given considerable attention. The part of the total gross return from the soil paid to the landowner for his use of this original and in- destructible force, has been termed economic rent by the English economist Ricardo (ref. Conner 1891). However, Ricardo does not regard the return from land due to soil im- provement as economic rent. According to Ricardo, economic rent is a residual income that is left over for the landowner, after the costs of the other production factors have been substracted from the total gross return. The formation of economic rent on soils of different qualities is based on the decrease of soil fertility along with population growth, be- cause the supply of arable land is limited while, at the same time, demand for it grows. The best soils are cultivated first, then the poorer soils. The last soils put to agricultural use give a return that is equal to compensa- tion for labour and capital, and so produce no economic rent at all. The teachings of Ricardo have been shar- ply criticized. Carey (ref. Abel 1958, p. 314) tried to prove that the best soils are not put to use first. Rather, soils that are easy to bring into cultivation, but which do not necessarily produce the highest yield compared to the best 168 soils, are employed first. Equipping labour with new machines results in an increase in the labour contribution to the total gross return and, correspondingly, in a decrease in the eco- nomic rent. This is contrary to the thoughts of Ricardo. Carey thus gave a new dimension to the theory of economic rent, namely the im- pact of technological development. The diver- gence of opinion between Ricardo and Carey was due to the difference in their conceptions of “better soil”. von Thunen (1921, p. 106) also brought a new concept to the theory of economic rent, namely the interest derived from the location of agricultural production with respect to the markets. According to von Thunen, the loca- tion of the commodity market has an effect on economic rent. Farmers practicing near a market obtain the same price for their prod- ucts as farmers practicing further away, de- spite the penalty of extra transportation costs. The benefit, measured in monetary terms or in crops, is economic rent. Several researchers, e.g. Marshall (1920), Clark (1923), and Cassel (1938) were favourable, like Ricardo and von Thiinen, to “the theory of differential rent”. This kind of rent was supposedly generated when soils of different fertility were put to use. The in- tensity and location of production, in relation to the market place, were considered to have an effect on differential rent. The determinationof economic rent is not simple, because economic rent has a double significance (Cassel 1938). The costs gener- ated by land use (e.g. rent on land, interest on capital) are often included in the cost cal- culations of the production unit and are re- flected in the prices of produced goods. On the other hand, the landowner’s return from land (e.g. rent on land) is economic rent. Referring to figure 6, Vartiainen (1963, p. 149) states that when elucidating the theory of price formation and the compensation for production factor costs, if only a limited quantity of a production factor is available, for example the soil, its supply curve is verti- cal. Price has no effect on supply. The price to be determined for the produc- tion input is then called interest or economic rent, because its market price depends solely on demand which, in turn, depends on the prices of the final products produced by the input. Economic rent is then price-deter- mined, or theprice is determined by the price of the final product. On the other hand, if eco- nomic rent is included in production costs, it is price-determining. As well as in economics, rent has also been considered in the theory of agricultural ap- praisals (e.g. Aereboe 1919, p. 117—199, Laur 1930, Neukomm 1947, p. 64—65). Ac- cording to Virolainen (1950, p. 25), eco- nomic rent is the part of the taxable net re- turn that can be used as interest on land, ir- respective of whether the capital is initial or produced (real). According to Virolainen, the national economic theory of value, as such, does not apply to agricultural appraisals. In a private economy, the division of the taxable net return between economic rent and real capital is irrelevant. Ylätalo (1978) determined the capitalized value for whole farms, employing the book- keeping farms of southern Finland. Two types of farms were considered: cattle farms and grain farms. The farm family income was divided into return from labour and return from capital according to the ratio of labour and capital inputs used in production. The re- turn from capital was further divided into re- Figure 6. Effect of demand and supply on the price and quantity of a production input. 169 turn derived from land, and return derived from capital other than land, according to the ratio of the parts in question. The return from land determined in this way was called eco- nomic rent. Land value was assessed by capi- talizing the economic rent for an unlimited period, using an interest rate of 3 %. The land value of farms on which arable farming pre- dominated was 2—3 times higher during the investigation period (1968—1972) than the land value of farms on which livestock farm- ing predominated. Among recent studies in Finland, Heiska- nen (1983, 1987) studied the determination of the value based on the grading of agricultural and forest land. He aimed at determining the capitalized value of agricultural land and for- est, because according to the legislation con- cerning division surveys, agricultural lands and woodlandsmust be graded, based on their yield. The calculation of the capitalized value of agricultural land in Heiskanen’s investiga- tion was accomplished using the results of bookkeeping farms in 1983. Even if the results of one year were not highly conclusive, it was evident that the grading value of the average agricultural land in southern and southwestern Finland was 80 points, and in the “blueberry type” forest it was approximately 45—50 points (Heiskanen 1987, p. 66). Thus, ac- cording to the grading, one hectare of agricul- tural land corresponded to about 1.6—1.8 hectares of forest. Laurila (1988) has also examined theories of economic rent and their suitability for ex- plaining price formation. The aim of the study was to construct a model for the determina- tion of the market price, in which price was explained by factors both exogenous and en- dogenous to agriculture. The linear regression model Laurila used in the study explained 67 % of the variation of the market price of arable land. Factors within agriculture, such as market activity, profitability in agriculture, quality of the arable land and crop yields ex- plained 63 % of the variation, while 4 % was due to exogenous factors. 3.1.3. Determination of the value of supplementary arable land by the residual approach The valueof supplementary arable land has traditionally been determined in agricultural economics by the residual approach, where the return or income from additional land is ex- amined as a residual from land. When the costs of other factors and inputs of produc- tion are subtracted from the grossreturn, the residual indicates the economic compensation for additional land. Typical to the residual approach are certain concepts of economic results, used for the determination of the value of supplementary arable land which will be ex- amined later. According to Ryynänen (1978, p. 15—17), the profitability of acquiring supplementary arable land should be considered ultimately on the basis of surplus of taxable net return. The taxable net return shows that part of the gross return which remains as interest on invested capital. By reference to this, a farmer acquir- ing additional land can determine whether he will obtain sufficient interest on the capital in- vested in land acquisition and whether he can still make a profit. With an increase of the arable land area, the gross return and oper- ating costs rise. The difference between the gross return and operating costs provides an indicator of the surplus of return derived from the increase of arable land area. This differ- ence is also the interest on the capital invested in land acquisition. The capitalization of this difference, according to the current interest rate, gives the capitalized valueof the increase of the acquired arable land. The increase of return can be determined for a farm as a whole or for the additional area. The net return of the additional area can be determined by subtracting the increased cultivation costs from its gross return. In this case, changes in other production activities on the farm are not taken into consideration. In- stead, only a part of the increase of the net return of the farm as a whole can be attrib- uted to return derived from land. The re- 170 Table 4. Taxable net return (FIM/ha) on the profitability survey farms of southern Finland, by size of farm. Aver- age figures during 1972—1986. Year Taxable net return3 5O ha Average 1972 -651 -24 188 301 379 204 1973 -926 -266 9 180 284 57 1974 -1134 -94 124 403 522 253 1975 -1178 15 480 608 813 504 1976 -1741 -220 482 748 981 542 1977 -2262 -854 -102 116 53 -114 1978 -2176 -839 -93 194 330 -38 1979 -2469 -827 -67 488 394 103 1980 -2856 -733 408 975 1482 754 1981 -3472 -1015 -69 501 634 193 1982 -2267 -321 812 1869 2129 1472 1983 -738 737 1271 2359 2957 2152 1984 -1453 -78 916 1635 2173 1474 1985 -2022 -339 635 1719 1798 1284 1986 -1674 -398 921 1815 2316 1573 Source: ‘ Current Topics in Agricultural Economics. Results of the bookkeeping farms, business years 1972—1986. mainder is due to a more intensive use of pro- duction factors. However, the division of the net return between land and other parts of the farm is difficult in practice (e.g. Ylätalo 1978, p. 27—35). Moreover, according to the profitability survey in agriculture, the average taxable net return in agriculture remains nega- tive in small farms (table 4). From table 4 it can be concluded that the taxable net return depends on the size of the farm. Yet, other factors have an impact on income, such as the farmer’s managerial abil- ity, differences in labour efficiency, different uses of production factors, and possible other activities. It is also probable that farmers’ per- ceived incomes are quite different. For exam- ple, the farmer of a smaller farm practicing intensive livestock husbandry may consider his farm to be at most a working place, and his income is perceived principally as income from work, despite a considerable amount of capital he may have tied-up in the enterprise. In larger farms, farmers pay more attention to interest on capital than to income from work. Accordingly, taxable net return would be more suitable as a criterium for assessing the profitability of acquiring additional land on large farms than on small ones. Locken et al. (1978, p. 55—57) also ad- dressed the main problem of applying the residual approach, namely the pricing of la- bour. They pointed out, with empirical exam- ples, that the residual income and, thus, the capitalized value varied according to whether or not the farmer’s work was priced: 1. Based on the average labour costs of ex- ternal labour. 2. According to 1, increased by the entre- preneur’s management costs. 3. Leaving labour totally unpriced. The annual average capitalized value is nat- urally the highest in case 3, the second highest in case 1, and the lowest in case 2 (Locken et al. 1978, p. 56). Other studies on the use of the residual approach have been examined more closely by e.g. Clark (1973) and Doll et al. (1983). In assessing the return from land, difficul- ties arising from the pricing of labour can be avoided by leaving labour unpriced, in which case the residual signifies the gross margin on labour and capital cut off from the gross re- turn. The calculation of these assessments with respect to the operating margin is based on the hypothesis that the farm possesses rele- 171 vant production factors (land, buildings, labour) for agricultural practices. Some of these factors have been planned over decades and so their costs are of long duration. Since the costs generated by these production fac- tors do not change considerably, when the ex- tensiveness of production maintains a parity with the capacity of the factors of production, they are called fixed costs (e.g. Westermarck 1967, p. 9—ll, Elstrand 1980, p. 114). If more arable land is acquired for the farm, the fixed costs induced by buildings, equipment and labour remain almost unchanged, in which case the unit costs per hectare decrease when the area increases. Variable costs are those that increase or de- crease along with the extensiveness of produc- tion. These costs consist of, e.g. costs for ex- tended crop cultivation (purchased seed, fer- tilizers, plant protection, operating costs of equipment, veterinary costs, feed purchases, and energy costs). The subtraction of the vari- able costs from the gross return gives the residual for covering the fixed costs. This residual, or operating margin, includes the possible profit derived from production activi- ties. When assessing the growth of the operat- ing margin resultant upon the increase of a farm’s arable area, the operating margin of the coming years must be determined.The as- sessments have to be based on past returns. According to Ryynänen (1967, p. 49), return must be assessed as a mean value of several years, because weather conditions result in an- nual variations of the agricultural yield. 5.7.4. The value of supplementary arable landbased on its marginalproductivity Real estate as a wholepossesses a definable value. If supplementary arable land is ac- quired, the value of the real estate will change. When supplementary land is acquired by an existing farm, there will be a need to assess its significance to the farm complex as a whole, i.e. the farm value must be appraised with and without the additional land. The value obtained in this way is called the differential value (Ryynänen 1967, p. 25). The aim is not to determine the value of ara- ble land based on the average return of the farm, but rather to apply the concepts of mar- ginal value product and marginal productivi- ty 1 . Larsson (1954, p. 68) and Wiiala (1960, p. 86) have used the concept of marginal value to mean land value based on marginal value product or marginal productivity. According to Barlowe (1958, p. 159— 160), compensation for land can be examined using the economic rent. Economic rent was defined by Barlowe (1958, p. 150) as the residual obtained after the minimum compen- sation to attract the factors into production was subtracted from the return. It can be de- termined by using a production function if marginal and average unit costs induced by production activities are known. Barlowe (ibid.) showed that it is profitable for the entrepreneur to increase production to the point where the marginal value product and marginal cost are equal. The economic rent, obtained by the entrepreneur, is then equal to the average unit return (price of the product) when the average costs of these units have been subtracted and multiplied by the quan- tity of units. This can be presented by cost curves derived from the production function (figure 7). Barlowe (1958, p. 160—163) stated that the analysis of marginal productivity is able to show differences in economic rent arising from differences in land quality, location in relation to the market place, and transporta- tion costs. The operating margin on fixed costs of the enterprise may be considered as economic rent. When the costs of fixed pro- 1 In a state of economic equilibrium the production factors are allocated a compensation corresponding to their marginal value product, i.e. the compensation for the factor of production X| is: in which 't ' l =the marginal product of AX; AX, X|, and PYj the price of the final product (Kehrberc and Reisch 1969, p. 61). 172 duction factors, except land, are subtracted from this operating margin, we can obtain the return from land. The difficulty is how to de- termine the demand for compensation of other production factors (Barlowe 1958, p. 164): . The theory of marginal productivity has generally been used to determine the demand for land for different purposes (Gabr 1972, p. 14—17). The presuppositions for the use of the theory of marginal productivity are the validity of the law of diminishing productivi- ty and the divisibility of production factors, which is not always possible to realize in agriculture (Griesbach 1966, p. 71—72). A marginal productivity approach based on marginal value product has been used in sev- eral studies in the United States. The main ob- jective of Strobehn’s (1966, p. 11) study was an examination of the relation between the market value of land at the moment of inves- tigation and its estimated value based on re- turn. The data were collected from farms 2 In the method described above, return from land has been assessed as a residual. The same method also applies to other production factors, as explained previ- ously. specializing in crop cultivation and beef, pork, and milk production. The marginal value product of different groups was estimated by using production function analysis. The results (p. 21) indicated that marginal value product of land on pig and beef farms de- creased by c. 3 % during 1949—1959 and on dairy farms by c. 1 % from 1954 to 1959. The sharpest decrease was observed on grain farms, in which the marginal value product of land decreased from 12 % to 4 % during the investigated period. The decrease of the marginal value product of land on all farms indicated that land market value increased faster than the value of the other input groups. Schuh and Scharlach (1966, p. 6—7) also demonstrated that the phenomenon of eco- nomics of scale could result in higher land values, as long as not all farms are able to ac- quire more land. Technological change can lead to intensification of production and en- largement of farm size which results in a rise of land value compared with other means of production. Locken et ai. (1978, p. 50—63) used an ap- proach based on marginal productivity when they estimated the use values of agricultural Figure 7. The determination of the economic rent by the marginal productivity approach (Barlowe 1958) 173 land in the State of New York. The produc- tion function of the basic model was exponen- tial: Y = X, Bi X2 B; . . .X 7 8 ’ U, in which Y = dependent variable, agricultural production, X, —X 7 = production inputs, B, —B 7 = coefficients of regression, and U = error term. The present value of land was assessed by capitalizing the marginal value product. For the consideration of quality factors, arable land was divided into three quality classes and the basic model was completed with dummy variables. The suitability of different models was tested on the basis of empirical data from 26 farms. The values of agricultural land de- termined in the study proved to be lower than the market prices paid. Similar studies on land use have been made in the Nordic countries, e.g. Hjelm 1963, Petrini 1964, Pihkala 1965 and 1975, and Elstrand 1980. Moreover, by using the pro- duction function approach, several Finnish studies have examined the profitability of the acquisition of supplementary arable land. (e.g. Torvela 1966, Ylätalo 1978, Heinonen 1980, Juvonen 1983). Pihkala (1975, p. 291 —306) addressed the applicability of methods describing the profit- ability of land use when studying the profit- ability of clearing forest land for arable land. Because the profitability of land clearing varied from farm to farm, marginal assess- ments were used to show the profitability of the additional agricultural land. The effect of an increase of one unit of invested capital and labour was also examined. In the study, a linear function and its logarithmic version were employed. The calculations showed that the extension of arable land on existing farms, within given limits, is more profitable than es- tablishing new small farms by clearance ac- tivities. The marginal value product of additional land can also be assessed by using linear programming. The main principles of this ap- proach and its suitability for agricultural plan- ning have been addressed in several Finnish studies (e.g. Weckman 1970, Pihkala and Lasola 1973, Pihkala 1975). With regard to an individual farm, linear programming pro- vides information on the effect of the exten- sion or reduction of different production lines on marginal value product. Studies based on linear programming which deal with the mar- ginal productivity of land have been discussed by, e.g. Clark (1973, p. 40—57). The determination of the value of addi- tional land by marginal analysis is relatively simple when there is unused production capac- ity on the farm and a few hectares of addi- tional land have been acquired. This is prob- ably the case in most small farms practicing traditional arable and/or livestock farming. On the other hand, if a large area of additional land has been acquired, so that the capacity of the existing production factors is no longer adequate, new investments will be required. Then, the determinationof the costs generated by investments may be difficult. 3.2. Studies based on market value 3.2.1. A general survey ofprice investigations concerning agricultural land Among the appraisal methods based on ex- change, the most typical examines prices paid for properties on the free market. This ap- proach is often applied to appraisals of in- dividual property lots, such as arable lands, but is also used in the appraisals of farms as a whole. According to Wiiala (1976, p. 4), market value signifies the value of a certain real estate that has been defined according to the current price of similar real estates. It reflects the price of the lot on the free market at the moment of appraisal. By its nature, it is an objective value representing the current value deriving from the law of supply and de- mand (Wiiala 1966, p. 10). An appraisal based on market prices has some important preconditions, such as, a 174 2 sufficient quantity of practicable objects of of departure was the value of agricultural land transaction (compare Mäki 1945, p. 47). Also, the largest possible sample of market prices of farms near the assessed land, com- parable to it in quality, location, and size, must be examined. Thus, the appraisal has to be made according to the price majority or the mean value. Gustafsson et ai. (1978, p. 52 53) summarized the conditions to be consid- ered when applying the market price method to the determinationof the value of agricul- tural real estates. There is insufficient information in Finland on the prices paid for agricultural land. There are no regular statistical series or indices to indicate changes in the value of agricultural land. Yet, market price studies have been made concerning urban land in Finland (Vir- tanen 1967, Myhrberg 1969, Kanerva 1974 and 1978 etc.). Moreover, many towns keep statistics on real estate transactions in their region. On the other hand, before the Na- tional Board of Land Surveying started the registration of market prices of real estates in 1981, price information covering the whole territory or concerning agricultural real estates was available only for brief periods. In addi- tion, there may be reservations concerning former market prices, for reasons to be dis- cussed later. Information concerning the value and price covering the whole country is available from bookkeeping farms belonging to the official agricultural profitability survey. The value of arable land in bookkeeping farms is estimated according to the local price level. The book- keeping advisor of the local agricultural ad- visory center determines, with the agricultur- al advisor and the farmer, the land price. One of the inadequacies of land value used in bookkeeping is that it is rarely revised (com- pare Torvela 1966, p. 58, Ihamuotila 1968, p. 86). Nevertheless, the bookkeeping values of the 1950’s are probably sufficient to de- scribe the price level of land during that peri- od. When Ihamuotila and Stanton (1970) tried to determine the quantity of capital in- vested in agriculture in the 1960’5, their point in 1951, based on the bookkeeping farms. Many of the transactions concerning all or parts of agricultural real estates are such that their prices are not comparable. This was also the case when results were compared in price investigations of Leponiemi and Lammi (1968), which covered the whole country in 1961, 1962, and 1966, as well as in other in- vestigations on land price. Leponiemi and Lammi (ibid) examined farms as whole enti- ties, complete with their buildings and forests, while most other land price investigations only examined the price of arable land. Moreover, the qualitative differences in arable land area may vary considerably (Leponiemi and Lind- berg 1966, p. 141). When the market price of a whole agricul- tural real estate is known, the price of an in- dividual lot can be determined by subtracting from the total valueof the transaction the ap- propriate values of other properties. Accord- ing to Mäki (1964, p. 105), this procedure is followed when the share of the appraised lot is considerable with respect to the total price of the transaction. Thus the price of arable land can be appraised separately from the prices of farms that the state has purchased in voluntary transactions. The purchased farms have mostly been used as supplemen- tary arable land for farms having development potential. In the Department of Agricultural Eco- nomics of the University of Helsinki, aca- demic theses have examined arable land prices using these data. The results of these studies have been compiled by Ryynänen (1978 a, p. 78) who has also developed a time series describing the development of arable land prices (figure 8). Though the information on arable land prices presented in figure 8 is somewhat in- adequate, it provides a general picture of the development of land prices. Since the begin- ning of the 1960’5, increase in arable land prices in southern and central Finland has been very rapid. Consequently, the real price of arable land doubled in these regions dur- 175 ing the period 1960—1976. At the same time, the real price of arable land in northern Fin- land remained almost unchanged. A parallel price development has been presented by Valkama (1979, p. 9—21), who examined the prices paid between 1960 and 1977 by the National Board of Agriculture for arable land in purchases financed within the framework of the Farm Act. According to this investigation, the real price increase of arable land, deflated by the wholesale index, was relatively greater in southern Finland during the 1960’5, than during the 1970’5. On the other hand, according to the ap- proach adopted by Ihamuotila and Stanton (1970), see also Ihamuotila (1983, p. 16), in an investigation concerning capital stock in agriculture, the estimated land price also in- creased sharply at the end of the 1970’s (ta- ble 5). In this investigation, the change in the price of agricultural land is one and a half times greater than the change in the producer price index for agricultural products, in per- centage points. Table 5. Price of agricultural land (FIM/ha) according to different statistical sources during 1977— 1980. Year Ihamuotila Purchases of Bookkeeping (1983) the National farms, Board of weighted Agriculture average (NBA) 1977 6566 5100 6681 1978 7118 6022 7245 1979 7758 6837 7619 1980 9163 6662 8278 Ihamuotila (1983, p. 18) points out that some reservations must be made concerning figures describing the price level of agricul- tural land presented in table 5, because of in- adequacies in statistics and price series. The Figure 8. The relative price of arable land in southern, central and northern Finland, and the stock price and whole- sale price index for the period 1960—1976. 176 figures reflect mean values for the whole coun- try. Regional price changes may be consider- able, as can be observed in figure 8. In a preliminary study for the National Board of Land Surveying concerning the province of Mikkeli, Heiskanen (1977) inves- tigated how market price statistics concerning scattered settlements could be determined and how they could be used to estimate the value of agricultural and forestry real estates com- posed of various types of lots. Some 818 trans- actions were studied, of which lots of sup- plementary arable land formed a relatively small proportion of the area for sale. Accord- ing to Heiskanen (1977, p. 34), the informa- tion concerning the quantity and quality of real estate lots under agricultural and forest management proved inadequate for the deter- mination of the price formation of agricultur- al and forestry land. The newest studies in Finland addressing price and its impact on the development of arable land are investigations by Holmsten and Myhrberg (1986), and Myhrberg and Väänänen (1988). The first study examined, through the use of price models, how different factors affect the price of arable land. The sec- ond study developed a special land-price in- dex to show the price development of arable land in Finland, as a whole, and in different regions. Both studies used information from the market price register collected by the Na- tional Board of Land Surveying. As a result of the studies, it is now possible to update an- nually the price index and publish it in the market price register. 3.2.2. Price investigations concerning supplementary arable land The considerable development of arable land prices has been partly influenced by the change in the nature of real estate transactions since the beginning of the 1960’s (Ryynänen 1978, p. 77—78). Earlier, a typical farm trans- action was followed by the purchaser moving to the property and continuing cultivation. The farm on sale was an independent eco- nomic entity, managed as such by the pur- chaser. Today, transactions of agricultural real estates increasingly concern the purchase of supplementary arable land. Purchases of entire farms, to be cultivated as independent economic units, are now insignificant. Sup- plementary arable land is acquired by pur- chasing farms in part or in their entirety, as well as by purchases of separate lots. To give a general picture concerning the price of additional land, the results of academ- ic studies made in the Department of Agricul- tural Economics of the University of Helsinki are represented in table 6. The results indicate that the price of supplementary arable land was approximately 9 900 FIM/ha in Kymen- laakso, in southeasternFinland, in 1975. This price was clearly higher than in the agricul- tural district of Seinäjoki in western Finland, where the price was 6 500 FIM in the same Table 6. Price of supplementary arable land (FIM/ha) and its development in different regionsB . Year Varsinais- Ind. 1975 Seinäjoki Ind. 1975 Kymen- Ind. 1975 Kainuu Ind. 1975 Wholesale Suomi = 100 = 100 laakso = 100 = 100 price index 1972 1910 65 60 6550 66 2150 73 71 7980 81 2475 84 88 9890 100 2935 100 100 3161 108 111 123 129 1973 1974 1975 9361 100 6546 100 11840 126 7262 111 15384 164 9165 140 17310 185 11277 172 1976 1977 1978 Sources: Ala-Kantti 1981, p. 72, Hannila 1980, p. 36, Mäkelä 1977, p. 31, Heikkinen 1978, p.BO. 177 year. The difference occurred because the agricultural land in Seinäjoki was not of the best quality. On the other hand, in the area of Varsinais-Suomi (southwestern Finland), the price paid was similar to the level paid in Kymenlaakso, namely 9 360 FIM. However, in the Kainuu district of northeasternFinland, the price of supplementary arable land was on- ly 2 900 FIM/ha. The results indicate that the price of sup- plementary arable land in southwestern Fin- land increased more rapidly than in the agri- cultural district of agriculture in Seinäjoki. In both regions, the price rise of supplementary arable land was 72 °70—85 % in 1978, com- pared to 1975, while, at the same time, the wholesaleprice index rose roughly 30 °7o . Even though the studies examined here are local studies, their conclusions show that the price of supplementary arable land continued to grow rapidly even at the end of the 1970’5. Besides the price formation of agricultural and forestry land, Kantola (1979) has also studied theirprice development in the area of some municipalities surrounding the City of Hämeenlinna. Information on market prices and the physical characteristics of the lots was collected from the card index of district regis- trars, and later completed with supplementary information. Regression analysis and the regula falsi method were used as research methods for the analysis of land prices (Kan- tola 1978, p. 4—13). The model used in the study, based on an exponential function, was as follows (Kantola 1979, p. 46—47): Hp =lp X 497.362 x S H -° 30 x A"017 X K° 33 x B° 56 , in which HP = price of arable land, FIM/ha, Ip = factor of price development of arable land (determined from data), SH = location factor, in relation to the central built up area (Hämeenlinna), km, A = size factor, measured as arable land area, ha, K = centrality factor, B = fertility factor, measured as tax classifi- cation points. The empirical data concerning arable land consisted of all land transactions in the munic- ipalities of Hattula, Renko, andKalvola. The degree of determination of the model was ap- proximately 85 %, while the residual standard deviation was approximately 21 °7o. Referring to the model, Kantola (1979, p. 98) con- structed a table for arable land estimation that showed the determination of the price of in- dividual lots. The increase in thereal price of arable land was approximately 3.5 % a year during the investigation period. Other similar models of the price of arable land, as well as the estimation tables based on them, have been constructed by, e.g. Matikainen (1980) and Kanerva (1980). However, because the sources concern the prices of farms as a whole, the results do not give a sufficiently accurate description of the price formation of sup- plementary arable land, and for this reason they will not be discussed further in this chapter. According to Gulbrandsen and Lindbeck (1969, p. 66), the development of land prices cannot be explained by prevailing circum- stances in an average agricultural enterprise. They claim that it is probable that the high price level of land is considerably influenced by the demand for supplementary arable land by large farms. Gulbrandsen and Lindbeck (p. 66) explained the rise of land prices with a simplified model in which the price of agricultural real estates was assumed to rise along with the price of agricultural products (productivity was assumed to be constant and the prices of production factors unchanged). Land prices estimated according to the model in question were compared to the real prices paid for agricultural real estates. During the research period 1952—1966, the prices of real estates increased by approximately 5 % a year, while the rise of the consumer price index was over 3 %. The land price rise corresponded mainly to the price paid for supplementary arable land by large farms. However, one should be cautious when evaluating the ac- curacy of the results. Further, it should be em- phasized that the demand for supplementary 178 Table 7. Prices of arable land and their variation in different research regions. Land Price Handein Vechta Biberach Heilbronn Total Number of transactions 134 103 127 143 507 Mean value, DM/ha 8 437 17 867 19 006 30 798 19 184 Stand, dev., DM/ha 3 435 11 800 14 609 40 768 24 915 arable land by large farms, together with the entrepreneurial income of small farms, deter- mines land prices (compare Jaggi 1945, p. 48). In a West German investigation, Nosit- schka (1973) determined factors influencing price formation of agricultural land. Market prices of lots of land were collected in four areas, 507 samples were examined altogether, which accounted for 15 % of the land trans- actions in the regions concerned. Market prices in different regions had a high degree of variation, as shown in table 7 (Nositschka 1973, p. 55). Nositschka (1973) and Wentrup (1978) es- timated the effect of a number of variables influencing the variation of land prices, al- though the effect of any single variable on land price proved mostly to be weak. The most reliable variables explaining variations in land price were (Nositschka 1973, p. 55— 103): 1) density of population 2) share of farms of s—lo ha of all farms 3) average size of arable land lots in the area 4) soil quality of arable land lots 5) land suitability for construction purposes. Even if good market price statistics con- cerning only supplementary arable land pur- chases were available by region and time, it would nevertheless be impossible to employ them as such in the appraisal of supplemen- tary arable land. According to Ryynänen (1967, p. 42), market prices can serve only to determine an abstract mean value, around which individualcases are dispersed. Ryynä- nen (1967, p. 42) argues that each individual case has to be appraised separately by taking into consideration its deviations from the characteristics of the farm representing the mean value. In this way, the value of each unique lot can be estimated. A purchaser cannot refer only to market prices when evaluating the advantages brought to his farm by the acquisition of supplemen- tary arable land, even though he/she possesses much informationconcerning transactions of supplementary arable land. This is because in- fluences exogenous to agriculture may be con- tained in the price information. The market price approach is nevertheless adequate as a means of completing and revising the capi- talized value approach (Locken et al. 1979, p. 409—410). 179 4. DETERMINATION OF THE CAPITALIZED VALUE OF SUPPLEMENTARY ARABLE LAND 4.1. The research region The natural conditions for agriculture dif- fer so much throughout Finland that the de- termination of the value of supplementary arable land has been considered necessary only in the case of one specific region. The south- ern and southwesternparts of the country are presented as one region in the official agricul- tural profitability survey, and they form an integrated region with respect to agricultural practices in the country as a whole. Thus, the bookkeeping region of southern Finland has been chosen as the research region (fig- ure 9). In southernFinland crop farming represents an essential part of agricultural production. Climatic conditions enable the cultivation of demanding crops. The cultivation of bread grains and root crops is concentrated in this region. According to Ulvinen (1980, p. 28), 99 % of all seed inspections of spring wheat, and 82% of barley and oats in 1979, were made in the research region. As well as cli- matic conditions, soil conditions also have a direct bearing on crop cultivation. Clay soils prevail in the arable farming districts of Uusi- maa, Varsinais-Suomi and Häme. Soils of fine sand, coarse sand, silt, and mull also occur in the area. Peat lands are rare (Kurki 1982, p. 95—99) in the research area. Table 8 presents a description of the quali- ty and quantity of arable land in the research 1 The research region includes the following agricul- tural advisory centers and Swedish-speakingagricultural societies: Uusimaa (1), Nylands Svenska (2), Varsinais- Suomi (3), Finska Hushällningsällskapet (4), Satakunta (5), Pirkanmaa (6), Province of Häme (7), Itä-Häme (8), Kymenlaakso (9), and Etelä-Karjala (10). Kymenlaakso and Etelä-Karjala, until the end of 1985, constituted the agricultural center of the Province of Kymi. Figure 9. Research region. 180 Table 8. Parameters describing arable land in the research region. Agricultural Class IAa Subsurface Ave. arable Arable land Arable land advisory of arable drainage 6, area area ha of the region center land, % % ha' in the region % Uusimaa 78.3 54.6 17.5 135 315 37.0 Nylands Svenska 69.4 51.0 17.9 75 142 33.2 Varsinais-Suomi 80.0 64,0 16.1 240 371 40.5 Finska Hushällninss. 54.4 43.9 11.5 29 226 17.7 Satakunta 51.5 48.0 10.8 178 809 26.0 Pirkanmaa 37.9 26.9 10.2 109 104 22.2 Häme 80.057.5 15.9 155 146 35.1 Itä-Häme 62.040.8 11.8 68 007 20.9 Kymi 53.530.7 11.4 152 272 24.1 Research region as a whole 65.348.7 13.4 1 114 166 29.6 Notes: • Anon. 1980 a b Anon. 1980 b c Anon. 1983 c. Figures are related to the arable land under cultivation. region. The best arable land is found in the agricultural advisory centers of Uusimaa, Var- sinais-Suomi, and Häme. The arable land of the eastern regions of northern Häme, north- ern Satakunta, and Kymi is of poorer quali- ty, as demonstrated by the results of the ara- ble land inventory study. On the whole, the quality of arable land in the research region can be considered to be clearly better than in the other bookkeeping regions. This is natu- rally reflected by the crop yield (information to be presented later). Moreover, technologi- cal innovations in agriculture are assumed to have been more rapidly accepted in the re- search region than elsewhere in Finland. 4.2. Agricultural conditions during the research period The research period covers the years 1972 to 1986. Because southern Finland can be con- sidered rather homogenous with respect to its soil conditions, climatic conditions are impor- tant, especially concerning the impact of weather on crop yields and their yearly varia- tions. The years 1977—1979, as well as 1981, were poor because of high rainfall, while 1982 was warmer than normal. The years 1974 and 1976 were more favourable than the average weather conditions during the 1970’5. The feed unit is used for the description of crop yields, but its application to pasture and straw yields in the profitability survey are not taken into consideration. The tops of root crops are considered only when they are har- vested. The average feed unit yield per hec- tare and the milk production per cow in the bookkeeping region of southern Finland are presented in table 9. Table9. Average feed unit yield per hectare and the milk production per cow from 1972 to 1986. Feed unit/ha* Milk production,Year kg/cow 502836991972 494431671973 502736311974 509630331975 550832621976 557526621977 560027651978 571628111979 593431221980 57381981 2384 586932311982 593434061983 594831801984 594330411985 62491986 3285 Average for 5607years 1972—1986 3112 Source; a Current Topics in Agricultural Economics. Results of bookkeping farms, business years 1972—1986. 181 182 The effects of weather conditions appear, above all, as yield variations. Unfavourable weather conditions resulted in a sharp de- crease in yield levels during 1977—1979. Ex- cept in 1985, yield levels have been higher dur- ing the 1980’s, than the average yield level for the period 1972—1986 as a whole. Milk pro- duction per cow has mostly increased during the investigated period. There was a sharp in- crease in milk production during the middle of the 1970’5. Besides the quality and quanti- ty of feed crop, improvements in cattle breed- ing and the feeding of cows have also had an impact in the increase of the average milk production per cow. On the other hand, dairy- ing is a long term economic activity, and so production level can be assumed to depend on a wide range of factors, not just those related to feeding and breeding. Besides yields, the development of produc- tion costs, as well as price developments of agricultural products, have an impact on the farm economy. Some central indicators are presented in table 10. The economic conditions prevailing during the research period changed considerably. This was especially the case with respect to the labour cost index, which is based on yearly wage statistics of average hourly wages of agricultural workers, which rose more sharp- ly than the other factors. Even if the labour input in agriculture decreased during the re- search period, the simultaneous rise in labour costs contributed to labour being the second largest individual cost item, after the cost of purchased supplies (Anon. 1986 b). Thus, the greater increase in the price of supplies and labour, compared to producer prices and the cost-of-living index, had an effect on the profitability of agricultural production. On the other hand, producer prices and the cost- of-living index have evolved nearly in the same way. 4.3. Choice and extent of data Data have been collected from farms that had continuously takenpart in the agricultural profitability survey since 1968; 186 farms were studied in 1985. Cattle I and II farms, as well as farms practicing pig husbandry or grain cul- tivation, were chosen as research material in Table 10. Indexes of producer prices and costs in agriculture during 1972—1986 (1972= 100). Year Producer Price index Labour costb Cost-of- price index8 of agri. ~ “ living .. , Men Women . , ,suppliesB index 0 1972 100 100 100 100 100 1973 113 114 125 136 112 1974 131 144 151 158 131 1975 164 175 200 206 154 1976 186 210 251 251 177 1977 199 249 312 311 199 1978 211 255 335 311 214 1979 224 263 349 347 230 1980 251 296 386 393 256 1981 282 358 433 438 287 1982 322 394 458 472 314 1983 343 430 488 514 341 1984 365 469 535 565 364 1985 390 495 558 593 386 1986 396 468 581 621 400 Sources: 8 Indexes assessed by the Agricultural Economics Research Institute. b Based on hourly wages in the determination of the value of its own labour by the farm family in the agricultural profitability survey. c Index of the Central Statistical Office. 183 1985 (see table 11 for more information). In addition, the material included four farms that were classified as other crop cultivation farms, although they were considered to be grain farms in 1968. Thus, 155 farms re- mained in the investigation. When the 1986 data were later collected, three farms were observed to have left the profitability survey. Because of the need to preserve the time series nature of the data, these three farms were also excluded from previous years. The final research data there- fore consisted of 152 farms whose participa- tion in the profitability survey had been unin- terrupted throughout the research period, 1968—1986. The information relating to par- ticular farms was collected both manually and from tapes in the Agricultural Economics Re- search Institute. The information consists principally of return, cost, and property value statistics. When classifying farms by functions, they were first divided into two main groups, or farming systems: animal husbandry farms and grain cultivation farms; according to whether the return from livestock husbandry or from arable farming was greater in the gross return. The subdivision between these two types of farming was accomplished by examining the proportion of the main product in the gross return. The division into production lines* put into effect in 1983 in bookkeeping farms, is described by table 11. Table 11. Types of farm production lines. Production line Proportion of gross return Cattle farms I share of cattle production > 80 % (of milk >5O %) Cattle farms II share of cattle production 60— 80 % Pig farms share of pig production >5O % Other livestock farms Grain farms share of bread and feed grain >5O % Other crop culti- vation farms This division into production lines has been under revision since 1968, although the changes have been relatively slight. These changes have been explained in detail in “Cur- rent Topics in Agricultural Economics”, pub- lished by the Agricultural Economics Research Institute. It is worth noting that before 1983, the bread grain return on grain farms had to be over 30 °7o of the gross return in order to be classified as grain farms. Since 1983, the gross return from bread and feed grains on grain farms has to be over 50 %. There have been small changes in the farms from one production line to another during the research period. The changes are partial- ly explained by variations in gross returns on farms, i.e. the basis of the division into pro- duction lines. Some of the changes in the farms’ functional classifications stem from changes in their production structure and in their specialization. Latukka (1989) has ex- amined the stability of farm structure in detail. Because the bookkeeping data is principal- ly designed to determine the capitalized value of supplementary arable land, only farms strongly oriented towards utilization of ara- ble land under cultivation will be considered. Cattle farms include farms on which the gross return from cattle represents at least 80 %, and the gross return from milk produc- tion at least 50 % of the gross return from agriculture. In these farms livestock husband- ry rests basically upon feed produced on the farm and not on an abundant use of pur- chased feed, as in the case of pig farms, for example. Cattle I farms will simply be called cattle farms in subsequent text. Grain farms are the best representation of the dependence of production activities on crops, because the returns from grain sales represent the major part of the gross return from agriculture. Table 12 indicates the development in the number of farms chosen for investigation when the division into production lines has been made according to calculations from the annual bookkeeping. Table 12. Development of the number of investigated farms, 1968—1986. Year Cattle farms Grain farms 1968 14 36 1969 22 32 1970 22 26 1971 29 23 1972 39 21 1973 45 16 1974 40 24 1975 43 44 1976 44 50 1977 57 41 1978 60 42 1979 52 41 1980 51 44 1981 54 49 1982 44 51 1983 44 54 1984 42 58 1985 43 58 1986 41 50 The figures show that the number of farms in each class are rather small. Consequently, the final material only includes data since 1972. At the end of the research period the number of cattle farms increased to over 40 and that of grain farms to 50. For statistical analysis the number of farms is sufficient be- cause the data in question are “panel data”, where time series of farm dataare employed. The size of a farm has been traditionally measured by its arable land, but this is not the sole indicator of the size of a farm, because even large farms can be managed without ara- ble land. This is why turnover and labour in- put have also been used to measure the size of a farm. Nevertheless, arable land is an in- separable part of agriculture and its superi- ority in measuring the size of the farm can be based on the increasing use of factors of pro- duction following the extension of arable area. Economic results with respect to the arable area of farm were declared in the profitabili- ty survey as converted arable hectare until 1975. Since 1976, results have been fixed ac- cording to the total area of arable land in use, as well as the leased arable area. In this study, the arable area is defined in a different way than in the profitability study. Arable area Table 13. Development of arable land area (ha/farm) during 1972—1986. Year Cattle farms Grain farms 1972 19.47 37.56 1973 20.27 44.53 1974 19.89 45.39 1975 21.52 44.05 1976 21.03 40.07 1977 22.79 41.53 1978 24.30 37.93 1979 23.03 39.04 1980 23.92 36.15 1981 25.47 37.41 1982 25.05 37.00 1983 24.68 36.26 1984 26.07 37.10 1985 26.28 37.00 1986 27.85 39.45 Average 23.44 39.36 (table 13) is considered to be the arable land area under cultivation and the fallow area was excluded in such cases where fallow agree- ments are in force. The arable land area of cattle farms has in- creased from approximately 20 hectares to al- most 28 hectares while the variationof the ara- ble land area on grain farms has been clearly less. On the other hand, grain farms are gener- ally almost twice as large as cattle farms. By the end of the research period the size differ- ences between the farm types were decreasing. 4.3.1. Private economic reasons for acquiring supplementary arable land The main objective of the professional farmer is to secure the best possible return from labour performed by himself and his family, as well as to obtain a sufficient interest on capital invested in the farm enterprise (compare Mäki 1964, p. 69). For the realiza- tion of these objectives, the farmer has to adapt his production activities from time to time to respond to changing circumstances. Then, in theory, he may continue with the production structure unchanged, but more likely, he will attempt to modify it to meet the new circumstances. The modifications to the production structure can be made by reducing 184 the use of factors of production. The agricul- tural policy measures prevailing in the second half of the 1970’5, aimed at a reduction in pro- duction volumes based on voluntary schemes, in which farmers were eligible for state grants to reduce the use of factors of production. De- spite these efforts, individual farmers have strived to increase their production and returns. In practice, this latter effort has manifested itself as an increase in the volume of use of production inputs, the implementa- tion of new technology, or as efforts to in- crease the arable land area of the farm. The increase in production at the farm level can be economically justified to the point where the marginal value product and the marginal cost of different factors of produc- tion are equal. At this point, the optimum size of the enterprise in the short term is attained (Renborg and Karlsson 1969, p. 15). If pro- duction activities occur in the area of di- minishing productivity2 , the optimum is ob- tained when: MVP X MVPX MVPX, MVP ,±L *2 - *3 ±O. = 1# P}<2 Px„ In the formula, MVPXj signifies the mar- ginal value product of each input (x,...x„) and Px , the marginal cost of the use of the input x,. Sometimes, inadequate resources (K) can set a limit to the expansion of produc- tion to such an extent that the above-men- tioned formula is not valid. In this case, the optimum result is attained when the marginal value products are equal, or MVP, = MVP 2 = ... MVPn , when x, + x2 + ... l. The first case is governed by the law of diminishing returns, the second one by the law of constant returns and the third one by the law of increasing returns to scale (e.g. Niitamo 1969, p. 12, Pekkarinen and Sutela 1979, p. 37—38). For estimation, a Cobb-Douglas function can be expressed in the following form: log Q = log a + Ejbjlog X; (i = 1,2,3...n). The marginal value product will be deter- mined from: MVPx-b, As a second function, the following tran- scendental function will be tried: Q = aX, b 'ec'x>X 2V2X2 .. .Xn b "ec"x". This function is, in fact, a combination of the linear function and the Cobb-Douglas function. The merit of this function, compared with a Cobb-Douglas function, is its flexibility. It obtains very variable forms according to the values and signs of regression coefficients (e.g. Kettunen and Torvela 1969, p. 54). For es- timation, the function obtains the following form: log Q = log a + Eibjlog X; + IjCjX, (i= 1,2,3.. .n). 191 The marginal value product MVPx = + Ci )Q. 4.4.4. Results of the regression analysis The differences in the production structure of individual farms have a significant effect on the economic result in agriculture and, thus, on returns derived from supplementary arable land. The assessment of returns from supplementary arable land were therefore made separately for cattle and grain farms. An intensive pasture economy is typical in cattle farms, while grain farms mainly specialize in the cultivation of marketable crops. Conse- quently, the quantities and shares of differ- ent production inputs differ considerably ac- cording to the type of farm. The main focus of the study is the determi- nation of returns from supplementary arable land during thefinal years of the research pe- riod. The returns from supplementary arable land are to be determined most thoroughly for the period 1982—1986, for which price infor- mation from the market price register is also available. Even though the research data concern farms that have continually practiced book- keeping since 1972, a cross section analysis is used for certain years before 1982. Estima- tions were eventually made for the following years: 1972, 1976, 1980, and 1982—1986. As to the yield, the years 1972 and 1976 repre- sent at least the average level of the research period (see table 9). The same applies to the year 1980. The extension of the cross section analysis back to the year 1972 enables the examination of the development of the capitalized valueof arable land as a function of time, i.e. the need for supplementary arable land has most ob- viously increased during the research period along with the implementation of new labour- saving technologies. Coefficients of the functions are estimated with regression analysis in which the validity of the models is considered primarily accord- ing to the degree of determination (R 2), the significance of regression coefficients and their signs. The dependence of gross return on different factors is explained with variables chosen on thebasis of theories of production and costs, correlation analysis and experimen- tation. Accordingly, the following indepen- dent variables considered to explain the gross return in cattle and grain farms were selected for estimation: Cattle Farms Grain Farms arable land area arable land area cost of purchased feed cost of purchased seed cost of purchased fertilizers revised cost of equipment livestock cost agricultural worksagricultural works revised cost of equipment agricultural works First, production function analyses were es- timated for cattle farms using the linear and the Cobb-Douglas functions. The results of the preliminary regression models showed quite high degrees of determination. The degrees of determination indicated that the linear functions explained 90 % of the varia- tion of gross return. In addition, regression coefficients of almost all the variables were statistically significant at 5 % risk level. The degrees of determination of the Cobb-Douglas functions were, in general, a little lower than for the linear models, even though their de- grees of determination still exceeded 90%. The difference in results between the linear and the Cobb-Douglas function derives from the measuring units of variables employed. Thus, both functions explained the variation in gross return very well. The regression coefficients of the models also proved logical with respect to signs. The results of the models will not be presented in more detail because the main task of the investigation concerns the marginal value product of arable land and its consequences (table 14). For the same reason, the estimates of the parameters of other pro- duction inputs included in the models are not of great interest here. The marginal value product of arable land on cattle farms, at fixed 1986 prices, rose from 192 Table 14. Marginal value product of arable land (FIM/ha) on cattle farms, according to the linear and the Cobb-Douglas functions*, at 1986 prices'l . Linear function Cobb-DouglasYear function 1972 2008 2370 1976 (952) 3381 1980 3019 3182 1982—86 2109 2505 Notes: a The assessment of the marginal value product of ara- ble land is based on the use of the formula: MVPx =b.-2- X. b As a deflator, the 1986 cost-of-living index was used. a level c. 2 000 FIM in 1972 to over 3 000 FIM until 1980, while during 1982—1986 the mar- ginal product in a linear function fell to 2 109 FIM and to 2 505 FIM according to the Cobb- Douglas estimation. These differences in levels (c. 700 —900 FIM) compared to the results of 1980, seem large and unbelievable. The fall in the marginal value product of arable land may partly be due to annual changes in the clas- sification of farms into cattle farms and crop farms, and because of the variation in the crop level. This will be examined more closely later. The results of table 14 permit the general ob- servation that there is a rather good correla- tion between results of the linear and the Cobb-Douglas functions. For 1976, the regres- sion coefficient of the linear function con- cerning the marginal value product of arable land did not prove to be statistically signifi- cant (identified by brackets). Farms which specialize in grain farming show considerable variations in the economic results, principally due to weather conditions during the growing seasons. As a result, the estimation of production functions on grain farms proved far more difficult than for cat- tle farms. As criteria of choice for variables for grain farms, production and cost theories, as well as correlation analyses, were used. On this basis, the following variables were final- ly chosen: arable landarea, cost of purchased feed, revised cost of equipment, and agricul- tural work cost (farm family plus hired labour). As the fertilizer input has a strong correlation with the arable land area variable (see appendix 4, table 4.2) it was excluded from the model because of the multicollinear- ity problem. The effect of the fertilizer input in the assessment of the returns from sup- plementary arable land has been taken into consideration in a later stage (see section 4.5.1). The explanatory power of the selected vari- ables with respect to the variation of gross re- turn proved rather poor, except for the period 1982—1986. As the annual crop on grain farms may vary considerably, a cross section analysis of three years was chosen for ex- perimentation. In this case, the applicability of the linear function turned out to be poorer than that of the Cobb-Douglas function when the significance of regression coefficients was used as a criterium. Similarly, the applicabil- ity of the transcendental function for the es- timation of parameters was also tried. Its de- gree of determination remained lower than that of the Cobb-Douglas function, and the coefficient of the linear term had low or even negative values. In addition, the t-values of the linear terms did not regularly prove sta- tistically significant. Thus, it was decided to abandon the use of this function. Consequent- ly, for the estimation of parameters for grain farms, it was decided that only the marginal value product of arable land given by the Cobb-Douglas function would be used when the estimates of the coefficients of the models were normally statistically significant at 5 % risk (table 15). Figures in table 15 show that the marginal value product of arable land has remained al- most unchanged, except at the turn of the de- cade. When assessing the results, it has to be noted that in variables, the model, except for arable land, during the two first periods of ex- amination did not prove statistically signifi- cant at 5 % risk. Consequently, the marginal value product is in parenthesis. During other 193 Table 15. Marginal value product of arable land (FIM/ha) on grain farms according to the Cobb-Douglas function at 1986 prices B . Marginal value product of arable land Years6 1972—74 1975—77 1979—81 1982—86 (5656) (3303) 2890 4607 Notes: a The assessment of the marginal value product of ara- ble land is based on the use of the formula: MVP=b-2- X, b Farms with the same production structure during the cross section period. periods under examination, except the final one in 1986, some variables were not statisti- cally significant. The marginal value product of arable land, determined at the average level of production input utilization, remained low- er in comparison to other periods under ex- amination during the period 1979—1981, re- sulting principally from the poor crops of 1979 and 1981. The preliminary regression models pre- sented above, the linear and the Cobb-Doug- las functions for cattle farms and the Cobb- Douglas function for grain farms, seem to be well adapted to explain the variation in gross return. The benefit of the Cobb-Douglas func- tion, compared to a linear function, is that it can be useful for the study of the returns to scale in production, as estimated by the sum of the parameters of the production inputs. The Cobb-Douglas function is based on the assumption that elasticities are constant for all input and output levels. The constant elasticity assumption has, nevertheless, been criticized. Despite this, the use of this func»; on is justi- fied, according to Heady and .billon (1972, p. 97), by stressing that the Cobb-Douglas production function is a useful tool for esti- mation of the productivity of inputs at their mean levels, even though its functional form does not correctly predict the production sur- face at the extremes. In the regression models examined above, differences in the quality of arable land have not been considered. The estimate of the regression parameters shows only one class, i.e. the average land value. For those farms on which the production potential of land is greater or lower than the average, land valuation will be inaccurately estimated. The investigation will therefore now examine whether, on the basis of the research data, it is possible to determine differences due to the quality of agricultural land and take them into consideration in the specification of the regres- sion model. As was shown in the theoretical part of the study, information on the tax grading and classification of arable land has not been re- vised since 1967. Even the bookkeeping data do not contain this information. Consequent- ly, it was decided that the average feed unit yield of arable land (fu/ha) should be used. Yields can be used to indicate the long term production potential of arable land under cul- tivation. In the profitability survey, the feed unit yield does not include the pasture or straw yield, while only those tops of root crops which have been harvested are included in the yield. The feed unit yield consists mainly of grains, cultivated grass, oil seed crops, and sugarbeet. Only the areas of these crops have been employed in calculating the feed unit yields per hectare. The change in the regression model caused by the inclusion of the variable describing the quality of arable land will now be considered in more detail. The improvement of the qual- ity of arable land is also expected to increase its marginal value product. The inclusion of the feed unit quantity variable describing the production capacity of arable land in the regression models is accomplished with dum- my variables. The Cobb-Douglas function then has the following form: Y = AX, B| + Bl ’D +B ‘” D” X2 b*. . ,X n B", in which dummy variables are noted with D’ and D”. 194 The marginal value product from arable land can be assessed with the formula: t(B, + B,’D’ +B,”D”)/X,] Y Dummy variables and their limits are deter- mined in the following way: if yield (fu/ha) is <2500 fu, D’ =O, D” =O, the marginal value product from arable land will be (B.Y/X,), if yield >2500 fu, but <3300 fu, D’ = l, D” =0 and the marginal value product from arable land will be [(B, + B/J/X,] Y, and if yield >3300 fu, D’ =O, D” = l, then the marginal value product of arable land will be [(B, + B|”)/X,] Y. For estimation, the function is expressed in the form: log Y =Log A + B,log X, + B,”D”log X, + B2 log X 2 + .. + B„log Xn . The results that have been estimated for cat- tle and grain farms are presented in appen- dices 2 and 3. In addition, correlations be- tween gross return and independent variables for the period 1982—1986 are shown in ap- pendix 4. Quite strong correlations occur be- tween some variablesand arable land area, but they are not considered to cause a significant bias on theresults, because the degrees of de- terminationof models are clearly higher than the correlationbetween independent variables. As a general comment on the results of the regression estimations, it can be noted that the integration of dummy variables for the qual- ity of land has somewhat improved the degrees of determination of models compared to those previously estimated. On the other hand, the dummies did not prove statistically significant in all cases, especially in the case of cattle farms. Yet, the estimates of parameters for the dummies of cattle farms were regularly higher than parameters describing their dispersion ( = standard deviation). On the other hand, dummy variablesfor grain farms were statisti- cally significant at a minimum of 5 % risk during all investigated periods. This demon- strates that the dependence of gross return on yield level (measured mostly as the feed unit yield of cash crops) is clearly higher for grain farms than for cattle farms. In addition, for both types of farms under examination, the estimates of parameters for the dummies, de- scribing the capacity of arable lands to pro- duce yields, increased along with the yield level (fu/ha). Thus, the results can be considered to be logical in this respect. Table 16, Marginal value product from arable land (FIM/ha) on cattle and grain farms in diTerent years, at 1986 prices. Year Yield, fu/ha 1972 1976 1980 1982—1986 Cattle farms <2500 2187 3017 3316 4013 2500—3300 2777 3542 4147 >3300 2900 3530 3768 4294 Grain farms» <2500 1660 2833 4089 2500—3300 2925 3390 4919 >3300 3457 3769 5512 Notes; The investigation of grain farms covered the years 1972—1974, 1975—1977, 1979—1981, and concerned only those farms which have maintained the same production structure throughout the period. The same procedure has been applied to cattle and grain farms for the years 1982—1986. 195 According to table 16 the marginal value product of arable land at fixed prices on cat- tle farms increased from less than 3 000 FIM in 1972 to over 4 000 FIM at the end of the research period. The integration of dummy variables seemed to increase the marginal value product of arable land for higher qual- ity land. Marginal value product figures un- derlined in table 16indicate the class of yield level to which the investigated farms belong according to the average yield of arable land during the investigated years. Even though the dummy variables for cattle farms did not prove in all cases to be statistically significant, their results were logical, except for 1976 when the dummy parameter was attributed a nega- tive sign. To replace missing information, the marginal value product of 3 274 FIM per hec- tare was used ( =average value of the marginal value product of arable land between the lower and the higher quality class in 1976). The average feed unit yield on cattle farms fluctuated between 2500 and 3300 fu/ha dur- ing all investigated years. On grain farms, the degrees of determina- tion improved in the same way as on cattle farms, along with the inclusion of dummy variables. Also worth noting was the greater increase in the marginal value product of ara- ble farms than for cattle farms, when yield levels increased. The results indicate that the variation of marginal value product is clearly related to the gross return, on which weather conditions also have a considerable effect. In order to even out the weather influences, the production functions for grain farms were es- timated on the basis of data for three or five years. The results of the first period of exami- nation had to be abandoned because only 14 farms had maintained the same production structure during the period 1972—1974. in general, it can be stated that the varia- tion of marginal value product of arable land has been more considerable on grain farms than on cattle farms. The marginal value product from arable land fluctuated, depend- ing on yield level, from c. 1 700 FIM to c. 3 500 FIM in 1976. In 1980, when the research period included two poor crop years, 1979and 1981, the average marginal value of arable land decreased slightly in comparison to the previous period. During 1982—1986, the mar- ginal value product of arable land on grain farms exceeded 4 000 FIM in the lowest qual- ity class, just under 5 000 FIM in the middle case, and slightly over 5 500 FIM on farms with a yield level of over 3300 fu/ha. The marginal value product of arable land seems to have continuously increased on cat- tle farms during the period of investigation. On grain farms, the marginal value product appeared to be more dependent on yield level (fu/ha), as may be observed by the marginal value product in 1980. On cattle farms, pro- duction is largely based on the processing of pasture feed to milk or meat, in which case the variation in crop quantity and quality is not so likely to be reflected as variations in gross return as in the case of grain farms. 4.5. Returns from supplementary arable land 4.5.1. Returns from supplementary arable land by farm types The marginal value product of arable land presented in table 16 has been determined with the estimates of parameters of the production function of grossreturn in agriculture, at the average level of input utilization. The mar- ginal value product of arable land obtained indicates the growth in gross return generated by one supplementary hectare of arable land, when the use of other production inputs in- cluded in the model remains unchanged. The additional return estimated in this way is not entirely available for the acquisition of sup- plementary arable land, because inputs not in- cluded in the model must be taken into ac- count. Therefore, those variable costs which are necessary for production, but were not in- cluded as explanatory variables in the regres- sion models, must be subtracted from the marginal value product. For the assessment of variable costs, the structure of production costs ofboth types of farms in question must be known. The re- 196 quired costs may then be determined. When the variable costs are subtracted from the gross return, the operating margin attributed to the fixed means of production of the en- terprise is obtained. In principle, this corre- sponds to the actual concept of value added3 used in the system of national accounts for agriculture proper and applied to the condi- tions of individual farms. In this study, this correspondence does not exist because the cost of equipment on cattle and grain farms also includes the depreciation of these items. Operating margin was not initially used as an explanatory variable because some of the variables describing labour input and property were strongly correlated and resulted in mul- ticollinearity problems (see appendix 4, tables 4.1 and 4.2). For this reason, the solution ap- plied above was preferred. In this approach, the part of variable costs not included in the model is only deducted after the marginal value product of arable land has been deter- mined on thebasis of the production function. Only variable costs that have been excluded from regression models will be considered as variable costs. These are: cost of purchased supplies cost of reparation and maintenance ( = buildings and land improvements) other costs. The costs of commercial supplies include not only purchased expenses, but also changes in stocks of supplies. The increment in stock has been deducted from the purchasing cost of the corresponding supplies and the deduc- tion has been added to them. Supplies con- sist of, e.g. purchased feed, seeds and fer- tilizers, plant protection and grain drying, as well as fuels and lubricants. The variable costs of purchased supplies on cattle farms do not include the costs of purchased feed and fer- tilizers, while on grain farms they do not in- clude the cost of purchased seeds. The costs of reparation and maintenance 3 Gross output in agriculture proper (livestock pro- duction plus crop production) minus intermediate con- sumption. arising from buildings and land improvements include the expenses generated by the prop- erty items in question. Costs due to buildings and land improvements, except depreciations, have been identified as variable costs on cat- tle and grain farms. Other costs consist of, e.g. costs of maintenance of drains, fences and farm tracks. For a more detailed exami- nation of the costs concerned see, e.g. Anon. (1986 b). The assessment of variable costs, at 1986 prices, on cattle and grain farms during the period 1982—1986 was made for farms that had maintained the same production structure during the whole period. The average variable costs (FIM/farm) not included in the regres- sion models for cattle and grain farms are shown in appendix 5 in table 1. On cattle farms, the proportion of variable costs in gross return reached 15 %, while on grain farms it was 32 °/o. The large percentage on grain farms resulted from the inclusion of pur- chased fertilizers in the variable costs. The operating margin available for land acquisi- tion is obtained after the percentage of vari- able costs in gross return has been subtracted from the marginal value product of arable land. Then, the marginal productivity of vari- able inputs that are not included in the model is assumed to equal one, at the point cor- responding to the average arable area of cat- tle and grain farms. The reality of this assumption is examined in figures 1 and 2 of appendix 6. The figures depict the variable costs per farm and per hec- tare as a function of the arable area costs which were not included in regression models. They show a linear trend. The variable costs per hectare, in both farm types, appear to de- crease as the size of the farm increases. This partially results from thefact that large farms have had the possibility to pay less for their materials than the small ones. In particular, this is the case in purchases of fertilizers and feed until 1985. Until then, taxable standard retail prices were used for fertilizer prices, while standard retail prices from the price lists of feed suppliers were used for mixed feed 197 prices (Siltanen and Ala-Mantila 1989, p. 36—37). In practice, considerable discounts from these standard retail prices were avail- able, depending upon the purchased quanti- ty. Consequently, in the agricultural total cal- culations basic prices of fertilizers are used for fertilizers and wholesale prices are used for feed. Even though the investigated farm types benefit from the advantage of scale, due to the size of farm, it is considered necessary to set the marginal productivity of these variables to one at the point of the average arable area of farm groups. For cattle farms this is 24.09 ha and on grain farms 30.85 ha. Determined in this way, the operating margin remaining for land acquisition is called the marginal operating margin (table 17). eluding supplementary incomes, was clearly higher on farms in receipt of state grants than the average value on farms in the farm econ- omy statistics used as comparative material. Thus, Siirola’s results correspond, to a large extent, to those for the bookkeeping farms. The results indicated that the average marginal tax rate for additional income in agriculture was 34 % on cattle farms in 1980 and 39 l/o in 1986. On grain farms the corresponding figures at the same points were 39 ®7o and 45 % (Siirola 1988, p. 55—58). On grain farms the average marginal tax was clearly higher than on cattle farms. The supplementary income on grain farms main- ly increased the taxable income of the farmer, which because of the unequal distribution of Table 17. Marginal operating margin of arable land (FIM/ha) on cattle and grain farms in different years, at 1986 prices. Year Yield, fu/ha 1972 1976 1980 1982—1986 2819 3411 3011 3525 3203 3650 1926 2781 2305 3345 2563 3748 Cattle farms 1859 2564<2500 2500—3300 >3300 2361 2783 30012465 Grain farms» <2500 1129 2500—3300 >3300 1989 2412 Notes: a See note to table 16. The marginal operating margin in table 17 indicates the pre-tax value, in Finnish marks (FIM), that is available for land acquisition. Because the bookkeeping data used in this study do not permit the determinationof the marginal tax rate to cattle and grain farms, the conclusions presented by Siirola (1988) were used to get after tax values. Based on the panel data of 341 farms from the period 1980—1984, Siirola examined the effect of an increase in income in agriculture on the marginal tax level, as well as the dis- tribution of the returns from labour between spouses. The average taxable net return, in- income, was already characterized by a high- ly progressive tax rate. In this study, the relatively low rates of mar- ginal taxation of 30 %, 40 ®7o, and 50 %, were employed, on the basis of which effect of tax- ation was approximately evaluated. When the marginal operating margin of arable land is multiplied by a coefficient (1 minus the mar- ginal tax rate) the net marginal operating mar- gin will be obtained. This can be regarded as the return to arable land under cultivation. Table 18 presents the net marginal operating margin of arable land determined in this way. Results of table 18 employ the same yield 198 Table 18. Net marginal operating margin of arable land (FIM/ha) on cattle and grain farms, at 1986 prices. Year Yield, fu/ha 1972 1976 1980 1982—1986 Cattle farms 1. 1301 1795 1973 2388 2. <2500 1115 1538 1691 2047 3. 930 1282 1410 1706 1. 1653 1948 2108 2468 2. 2500—3300 1417 1M 2U5 3. 1181 1392 1506 1763 1. 1726 2101 2242 2555 2.>3300 1479 1801 1922 2190 3. 1233 1501 1602 1825 Grain farms 1. 790 1348 1947 2. <2500 677 1156 1669 3. 565 963 1391 1. 1392 1614 2342 2. 2500—3300 1193 1383 2007 3. 995 1153 1673 1. 1688 1794 2624 2. >3300 1447 1538 2249 3. 1206 1282 1874 1. Marginal tax rate 30 Vo. 2. Marginal tax rate 40 Vo. 3. Marginal tax rate 50 Vo. levels as used in table 17. The marginal tax rate is assumed to be 40 % on cattle farms, and 50 % on grain farms. The corresponding net marginal operating margins are underlined in table 18. The capitalized value assessed from these figures will be compared later to the market prices of supplementary arable land. The net marginal operating margins pre- sented here have been carefully determined be- cause in the assessment of taxable net returns, the taxable depreciations of different proper- ty items are normally deducted from the re- turns of the tax year. In this study, only depreciations of equipment have been taken into consideration in the assessment of the marginal operating margin and, thus, in the assessment of the net marginal operating mar- gin of arable land. In addition, the rates of marginal tax used in the study are relatively low. 4.5.2. Return from supplementary arable land as a function of the size offarm The profitability gained from an addition of arable area to an existing farm is often justified by the return derived from intensi- fied use of the production inputs. The produc- tion inputs of rather small farms, in particu- lar, can be considered to be used at below full capacity. Consequently, additional arable land area should be most profitable on these farms. The increase in returns from the increase of arable land would gradually decrease, accord- ing to the theory, along with the growth of the size of farm (figure 11). Initially, the dependence of returns on the size of farm was examined with a model ex- plaining the gross return of agriculture in the same way for both cattle and grain farms, as shown above. The farm types concerned were divided in two size classes so that there would be a sufficient number of farms in each group 199 Table 19. Marginal value product of arable land (FIM/ha) on cattle and grain farms, at 1986 prices*. Cattle farms Grain farms <25 ha >25 ha <3O ha >3O ha No. of farms 74 81 90 105 Yield fu/ha <2500 4775 1361 3344 4319 2500—3300 4877 1572 4106 4883 >3300 5183 1701 4481 5382 Notes: a The assessments are from 1982—1986. for statistical analyses. The investigation cov- ered the years 1982—1986. The results in table 19 indicate that the mar- ginal value product of arable land on cattle farms of less than 25 ha is clearly higher than cattle farms on an average (compare table 16). On the other hand, in the larger farm group (over 25 ha), the marginal valueproduct from arable land had decreased by a third compared to the small farm group. The results show that the increase of marginal value product gener- ated by the extension of arable area is clearly highest on rather small cattle farms. Along with the extension of arable area, the marginal value product starts to decrease. Therefore, it seems that the economic result of large cat- tle farms is more dependent on other produc- tion inputs and their intensity of use than on the increase of arable area. Contrary to cattle farms, on grain farms the marginal value product of arable land on farms of less than 30 hectares was at a lower level than the average marginal value product of arable farms as a whole (compare table 16), while for rather large (>3O ha) farms, the marginal value product in different yield classes was approximately the same size as for arable farms on average. The marginal value product of grain farms does not seem to de- crease with increasing size class as it does on cattle farms. Another way of analyzing the dependence of the economic result upon farm size is to ex- amine the average taxable net return and its importance in different farm size classes. In other words, the economic result reveals the part of grossreturn remaining for interest on capital invested in agriculture. In its assess- ment, depreciable property items have been re- vised by the cost-of-living index, and depreci- ations determined on the same basis. The results are shown in table 20. Table 20. Average taxable net return in agriculture (FIM/ha) on the investigated farms by size classes, at 1986 prices*. Farm size Cattle farms Grain farms daSS ’ ha FIM/ha No. of FIM/ha No. of farms farms <2O -2 796 43 771 40 20—30 416 69 1 129 50 30—50 1 318 38 2 228 74 >5O (952) (5)b 2 498 31 Notes: a The assessments are from 1982—1986. b Only one farm in this class. 200 In both farm types the results indicate an increase in theaverage taxable net return with increasing farm size. The single exception is the largest class of cattle farm, in which the returns per hectare decreased. The difference in the results for cattle farms in relation to grain farms can be largely explained by the different use of labour. Rather small cattle farms are labour-intensive, which is the rea- son why uniform pricing of labour input in bookkeeping farms, without consideration of differences in labour efficiency, leads to a negative interest on capital on farms of less than 20 hectares. The average interest on capital on grain farms is positive in classes of size of farm of less than 20 ha, and exceeds 1 000 FIM in farms of 20—30 ha. The net return almost doubles when moving to the 30—50 ha class, and is about 2 500 FIM at its highest level, on farms of over 50 hectares. The average taxable net return can be used to examine theadditional return derived from an increase in arable area. The approach fol- lows that presented by Elstrand (1980), ac- cording to which the net return of two size classes of farm is first multiplied by the ara- ble area, after which the return of the smaller class is deducted from the return of the larger and the differenceobtained is divided by the area differences of the farms concerned4 . Ta- ble 21 presents the returns determined in this way. The results show that the differential re- turns for the investigated farms are clearly higher than theaverage taxable net return. In addition, the differential return of small cat- tle farms was higher than for larger ones. It may therefore be assumed that arable land is not a scarce production factor on larger cat- tle farms, and that other means of production, together with matters related to management and organization, have a more pronounced ef- fect on economic results than arable area. On grain farms, the growth in differential return with an increase in arable land con- tinues to be strong, as does the marginal val- ue product. Nevertheless, on rather large (> 50 ha) farms, the growth in differential re- turn has been a little slower than between the smaller size classes of 20—30 ha and 30—50 ha. According to the results, cattle farms are labour-intensive enterprises, and farms of less than 30 hectares benefit the most from sup- plementary arable land. However, on grain farms, which are capital-intensive, it is still economically justified to increase the arable area on farms of over 50 hectares. These differences in economic results between the farming types are partly explained by their differences in labour-intensity and capital- intensity. 4 (37 : 5 x 1318) —(25.2x416) =3166 FIM/ha i1.5-25.2 Table 21. Return of arable land (FIM/ha) on the investigated farms, based on the difference approach, at 1986 prices*. Farm size Cattle farms Grain farms Average Differential Average Differential arable return arable return area, ha FIM/ha area, ha FIM/ha <2O 145 140 ™ , A 4769 .. e 1606 3S_so 37 5 3166 39 4023 >5O 52A 95A 2689 Notes; a4b See notes at table 20. 201 4.6. Assessment of the capitalized value of supplementary arable land 4.6.1. The selection of the capitalization rate In the determinationof the capitalized value of supplementary arable land, the capitaliza- tion rate, as well as the size and duration of return, has a considerable effect on the final result. In principle, the more information available on the returns and market prices of real estates, the more reliable the determina- tion of capitalization rate. According to Ryy- nänen (1989, p. 46), the evaluation of return and capitalization rate for the lifetime of a permanent real estate proves problematic in a rapidly changing economy. This also con- cerns arable land under cultivation. Murray (1969, p. 183—195) has ap- proached the selection of capitalization rate in two ways. First, land is considered as an ordinary investment, in which case interest is determined according to the returns on alter- native investments. Because land is considered to be a safe investment, it is compared to bank deposits, bonds, or stable shares. Alternative- ly, the interest rate can be determined accord- ing to the relationship between the return and market value of arable land. The interest rate determined in this way usually remains lower than in the former case. Suter (1976, p. 261 —262) considers many economic factors which have an effect on capitalization rate. Among them are the gen- eral interest rate, tightness of the money mar- ket, risks related to different assets, and an- nual variations in return. Furthermore, Suter (1976, p. 273—275) divides the factors influencing the interest rate into two groups; A. Factors increasing the interest rate: 1. The owner has acquired land for invest- ment. 2. The land has been acquired for intensive production. 3. The interest rate is high either because of a strong demand or a weak supply. 4. The alternative investment possibilities provide high returns and seems to have a good future. 5. Whole farms or supplementary arable land can be acquired quite easily. 6. Investors are not interested in acquiring land for protection against inflation. 7. Land is intended for a specialized produc- tion and the return is known to have a con- siderable annual variation, and there is a high level of risk. B. Factors decreasing the interest rate: 1. The farm is a family farm, or the land is appropriate for supplementary arable land. 2. Factors other than pure business are related to agriculture. 3. When traditional production, appropriate for the region, is practiced there is a low level of related risk. 4. The farm is located in a good farming re- gion and the land quality is good. 5. The general interest rate is low. 6. Changes in land use methods are possible. 7. The return from alternative investments is low. 8. The acquisition of whole farms and/or supplementary arable land is difficult. 9. Inflation is high and investors are willing to invest in land. If capitalized value is assessed for a sepa- rate farm property item, the choice of capitali- zation rate depends, in the first place, on the generally accepted interest rate and on the risk related to the use of theproperty item for pro- duction activities and the return to be derived from it. In fact, agricultural cost assessment practice has used different interest rates on capital invested in different property items. A preliminary consideration has been the risk of loss inherent in investments in different prop- erty items in agriculture (Laur and Howald 1957, p. 110, Skomroch 1962, p. 819). In these studies, the risk of loss is related to the importance of the interest requirement. Also, the interest rates on different property items 202 discussed by Meimberg (1956, p. 227) are based on the same grounds. He proposes an interest rate of 3.5 % for long-term capital, 7 % for medium term capital, and 9 % for short term capital. Laur and Howald (1957, p. 110—111) recommend a rate of 3 % for land, to which a 1/2% is added when build- ings or land improvements are taken into ac- count. For stocks a 5 % interest rate has been proposed and for forest stands a 31/2 % in- terest rate. In more recent studies in agricul- tural economics (e.g. Elstrand and Sonju 1978, Ylätalo 1978), an interest of 3 °7o— 4 % was used for the determination of the capitalized value of arable land. In practice, it may be assumed that every farmer subjectively determines the level of the interest used in the capitalization of the return of supplementary arable land. If the farmer uses an 8 % interest rate, he is not willing to pay as much for supplementary arable land as a person who uses a 3 % interest rate. On the other hand, the farmer who cultivates the land he owns and who values his profession more than alternative ones, may pay a high price for land and may be satisfied with a low labour income if the acquisition of supplemen- tary arable land guarantees the continuation of his professional practices. Economic results may also remain low when, after the acquisi- tion of supplementary arable land, the means of production or inputs are deficient; these being essential prerequisites for the efficient use of supplementary arable land. A high price paid for supplementary ara- ble land may also result in a change in pro- duction structure, in which case production is intensified or extensified according to the cir- cumstances. Supplementary arable land is real property which keeps its value. Therefore, regarding as- sessments of the capitalized value, Kanerva (1982, p. Ill) proposes the use of a real in- terest, i.e. one that corresponds to thereal in- terest on long term investments. He argues that the capitalization rate should be chosen according to the real interest, as derived from alternative investment opportunities. The in- vestments considered are banking deposits, savings via the equity market, or state bonds. The safety and liquidity of investments also have an effect on interest. According to Kanerva (1982, p. 113), the liquidity of banking deposits is quite good, but the real interest on capital has remained low or even negative. Conversely, returns from shares have varied quite a lot, in addition to which risk factors related to their acquisition are greater than those related to real estate and banking deposits. The return from quoted shares in Finland has probably followed a similar development to those in Sweden, where Gustafsson et ai. (1978, p. 62—63) report the return from quoted shares to be about 5 ®/o during the pe- riod 1950—1974, but less than 2 °/o in the 1970’5. Taking into consideration the esti- mated development of return, they propose an interest rate of about 3 % for land. The above discussion supports the selection of a relatively low interest rate for the capitali- zation of the return from supplementary ara- ble land, although in practice, the situation is much more complex. In this study the capi- talized value of supplementary arable land is being compared to market prices paid for ara- ble land under cultivation. Therefore, three different capitalization rates were finally selected for use in the calculations. First, the real interest rate is set at 3 %, in which case the acquisition of supplementary arable land is assumed to be self-financed. The second alternative sets the interest at 5 %. Supplementary arable land would then be ac- quired by self-financing and loans. In the third alternative for the capitalization of thereturn of supplementary arable land, a 7 % real in- terest rate is used, and the acquisition of sup- plementary arable land is then considered to be entirely financed with external capital, i.e. via ordinary bank financing. Though a real interest of 7 % seems rather high, it can be justifiedby thepresent high interest rates, and by the ever-increasing dependence of agricul- ture on loans at market rates of interest. Naturally, for individual farms, such sche- 203 matic examples are not valid, but the choice of the interest rate used in assessments de- pends on whether the farmer is granted a farm loan or an interest-subsidized loan. Their rate of interest in the financing instructions of the Farm Act (Anon. 1989) for zone IV, which corresponds approximately to southern Fin- land, is principally 6 %, and the financing per- centage is at most 50 % of the market price. The capitalization rate can also be deter- mined by comparing the return from arable land to the capital that has been instrumental to the realization of the return during the equivalent period. This approach is hindered by the small annual number of purchases of arable land. According to the Farm Register, an average 14 900 hectares of arable land a year were purchased in the country as a whole during the period 1982—1986, which corre- sponds to little more than 0.5 % of the total arable area. The market prices indicating the return and value of supplementary arable land are there- fore based on the use of marginal informa- tion on returns and prices. Consequently, they are much more sensitive to changes, e.g. in the money market situation, than the average returns and prices of real estates. Land rent, which can be compared to re- turn from supplementary arable land, is con- sidered in the system of national accounting as income transfer, in the same way as divi- dends, interests, and income taxes of enter- prises (Björk 1984, p. 460). If the landowner does not lease his land to an outsider, but uses it for his own production activities, it may be considered as a production factor from which a return will be implicitly required, cor- responding to the rent that would have been obtained in an alternative use. Laurila (1988, p. 35), who compared rents with mar- ket prices of arable land under cultivation, also followed this method. In this case, rents paid for arable land represented approximate- ly 5.6 % of the market price of arable land in southern Finland. The market prices of ara- ble land were determined from the loan statis- tics of the National Board of Agriculture in 1986. As a result, the real capitalization rate of 3 %-7 °7o, chosen for the study, was justi- fied. The use of a relatively low real interest is also justified in that the taxable values of the depreciable property items of the inves- tigated farms were transformed to better cor- respond to their real market values (appendix 1). 4.6.2. Capitalized value of supplementary arable land According to Found (1971, p. 23—24), the value of land is derived in one of two ways: (i) return derived from productive activities, or (ii) the price obtained or expected from selling the land. For a farmer, land is an in- dispensable production factor, the absence of which makes farming impossible. In this re- spect, the determination of land value is based on the value assessed from the return from supplementary arable land. Land value is therefore considered as the present value of all future returns. Even though the method of capitalized value discussed above is fundamentally sim- ple, its applicability has been controversial, primarily because of difficulties related to the selection of an appropriate interest rate (see section 3.1.1.). The interest rate used in capi- talization may vary considerably in different periods of time, Scofield (1965, p. 46). Thus, in periods of relatively stable return from land, the effective interest on capitali- zation tends to decrease, thereby increasing land values. The opposite situation prevails when there is uncertainty connected to the re- turn from land. The duration of the capitalization of return has a significant effect on the present value. Because fixed costs vary over time, the oper- ating margin has to be dividedamong the fac- tors that participate in its formation (Ryynä- nen 1967, p. 61). This is because the benefit from supplementary arable land equals the operating margin only at the beginning of the appraisal moment. Operating margin does not, however, continue to be as high, because 204 the adequacy of existing farm buildings and equipment with respect to cultivation at land to be purchased, as well as fixed costs derived from them, remain at their initial level only as long as they are effective. With respect to the capitalization of the operating margin, the proportions of the fixed costs of each factor contributing to the for- mation of the operating margin should first be determined, as well as the number of years required by each factor to adapt to the new situation, Mäki (1963, p. 3). Mäki (1963, p. 4) also proposes that the interaction of dif- ferent factors should be assessed as a weighted average value, giving, as a result, a coefficient by which the operating margin will be multi- plied for the determinationof the capitalized value of return. According to Ryynänen (1967, p. 136), the coefficient is approximately 10—13 (according to a 4 % interest rate) at the conveyance of the land. The coefficient of present value, used for capitalization of return from supplementary arable land, is influencedby the period oftime during which the acquisition of supplementary arable land should be financed. On the other hand, the age of the farmer at the moment of acquisition also has an effect on the duration of return from supplementary arable land. If the farmer acquires supplementary arable land at the age of 25, he theoretically has 30 work- ing years left before he reaches retirement age, according to the present statutes of the Farm- ers’ Pension Act. According to an investigation by Huhta- mäki (1985, p. 19), concerning pensions re- lated to generation transfers on farms under- taken by the Farmers’ Social Insurance Insti- tution, the age of those taking over farming was, on an average, 29 years. It is probable that those taking over farming will have ex- pensive loans for at least ten years in connec- tion with the acquisition of their farms. The economic burden of the farm transaction can be estimated to ease only at about the age of 40, after which some 15 years would be left for the financing of the acquisition of sup- plementary arable land before retirement. Tolonen (1985, p. 17), examining middle aged farmers who had acquired supplemen- tary arable land via the Farm Closure Pension system, came to similar conclusions. On the other hand, improvements and replacements of buildings and equipment during the years following the farm transaction increase the fixed costs. Thus, the shortening of the dura- tion of fixed costs together with the brevity of the remaining “active life” of the farmer, demands a relatively short duration for the de- termination of coefficients of present value. Therefore, durations of 5, 10, and 15 years for different capitalization rates were chosen for use in this study (table 22). Table 22. Coefficients of capitalization as functions of the duration of return and interest rate. Duration of return Interest rate % 35 7 5 4.580 4.329 4.100 10 8.530 7.722 7.024 15 11.938 10.380 9.108 The lengthening of the duration of return beyond 10 years considerably increases the coefficient of capitalization. Compared with a ten year duration of return, at 15 years, the coefficient of capitalization will be about 40 % greater when using an interest rate of 3 %, about 35 % greater when using an interest rate of 5 %, and about 30 % greater when using an interest rate of 7 %. In addition, the sig- nificance of the interest rate is emphasized with the lengthening of the duration of return. Table 23 presents the annual variation of capitalized value for cattle and grain farms in more detail. The capitalized value of arable land on cat- tle farms has steadily increased with time (ta- ble 23). In real terms, the value of arable land increased almost one and a half times during the years investigated. During the last period investigated, the capitalized value of arable land varied, according to the ten year dura- tion, from approximately 18 000 FIM to lit- tle less than 15 000 FIM, depending on the in- 205 Table 23. Average capitalized value of arable land (FIM/ha) for cattle and grain farms for different durations of return: a = 5 years, b= 10 years, c= 15 years8 . Duration of return Interest rate % 3 5 7 Year(s) Cattle farms 6490 6134 5810a b 12087 10942 9953 1972 16916 14708 12906c 7649 7229 6847a b 14245 12896 11730 1976 19936 17335 15210c 8276 7823 7409a b 15414 13954 12692 1980 21572 18757 16458c 9687 9156 8672a b 18041 16332 14856 1982 198625249 21954 19263c Grain farms 4557 8487 4307 4080a b 7683 6989 1975 197711878 10328 9062c 4411 4170 3948a b 8214 7436 6764 1979 198111496 9996 8771c 7662 7242 6859a b 14271 12919 11751 1982 198619972 17366 15238c Notes: 8 Capitalized values have been calculated by choosing from table 18 the net marginal operating margin that cor- respond to a 40 % tax rate for cattle farms and a 50 % tax rate for grain farms, which have been capitalized. terest rate. On the other hand, on grain farms the increase of the capitalized value of arable land was not stable, and was a little lower dur- ing the period 1979—1981 than it was during the period 1975—1977, primarily due to poor yields. The capitalized value ofboth grain and cattle farms was highest during the most re- cently investigated period, and yet the capital- ized value of grain farms remained somewhat lower than that of cattle farms, as during previous periods. The real increase in the capitalized value of grain farms was 1.7-fold, which was clearly more rapid than for cattle farms. 206 3 5. MARKET VALUE OF SUPPLEMENTARY ARABLE LAND AND ITS DEVELOPMENT IN THE INVESTIGATED REGION 5.1. The collection and extent of data Price information on arable land under cul- tivation is available from three different statistical sources: the land acquisition statis- tics of the National Board of Agriculture; pur- chases of additional land in accordance with the Farm Act; and the market price register of the National Board of Land Surveying. Each of these statistics has its particular ad- vantages and shortcomings (see e.g. Laurila 1986, p. 15—16). In this study, the market price register was chosen for price informa- tion on arable lands. In principle, it includes land acquisitions of the National Board of Agriculture and transactions of supplemen- tary arable lands financed in accordance with the Farm Act. The data concern transactions of over 2.0 hectares, made during 1982—1986, consisting principally of cultivated land. Another set of data on market prices con- sists of price statistics on land purchased for the state by the National Board of Agricul- ture. Because the purchasing activities of the National Board of Agriculture have continued for so long, information from the market price statistics is used for the description of the price development of arable land during 1972—1986. The National Board of Land Surveying’s real estate market price register contains the following information on all arable land trans- actions: time of conveyance location of the object of transaction the area of land involved total market price and unit price buildings on the area classification of buyers and sellers. In addition, information on the distance be- tween the lot and the additional real estate is also available for some transactions. Yet, the market price register does not include infor- mation on soil quality, the workability of the land, drainage, or other related activities. Transactions between relatives are also omit- ted from the market price register. In theresearch region of southern Finland, the number of arable land transactions has continuously increased. In 1982, 197 arable land transactions were specifically of culti- vated land 1 . In 1986, the number of these transactions reached 271, when land divisions and obviously underpriced transactions were discarded from the material. In 1982, prices of under 1 500 FIM/ha were considered to be underpriced. The limit of underpricing was raised in proportion to the rise of the cost-of- living index. Some of the transactions also in- cluded buildings, the values of which were not declared separately. The number of transac- tions of supplementary arable land and the areas involved are given in table 24. The table indicates a slight increase in the mean area of transactions concerning solely supplementary arable land. On the other hand, there has been a slight decrease in the mean area of transactions including buildings. During the investigated period, the area of transactions of supplementary arable lands 1 In this study these transactions are called sup- plementary purchases, although the information wheth- er the purchase concerned supplementary area was not available for each case. 207 Table 24. Transactions of supplementary arable land by number and size, 1982—1986. Year No buildings With buildings Total No. Mean No. Mean No. Mean Total area area area area ha ha ha ha 1982 156 4.91 41 7.61 197 5.48 1 078.6 1983 176 5.07 32 7.90 208 5.50 1 144.5 1984 192 4.87 25 5.39 217 4.93 1 070,4 1985 220 5.44 36 6.57 256 5.60 1 432.7 1986 222 5.09 49 6.28 271 5.30 I 437.1 has increased from c. 1 100 hectares to 1 440 hectares a year in southern Finland. The aver- age arable land area entering the market dur- ing theperiod 1982—1986 was 1 233 ha/year, which represents slightly over 0.1 % of the total arable land area in southern Finland (1.1 mill. ha). Almost 6 200 hectares of land were exchanged in arable land transactions. Even though most transactions concern only sup- plementary arable lands, a considerable num- ber of transactions included buildings. Their part in the exchanged total area increased from 13 % to 29 %. The annual area transac- tions of supplementary arable land including buildings is shown as follows: Year Total 1982 312 ha 1983 253 ha 1984 135 ha 1985 237 ha 1986 308 ha. Table 25 presents the number and average area of transactions of supplementary arable land by Agricultural Advisory Centers. It can been seen that most transactions were made Table 25. Number of transactions of supplementary arable land and average area (ha) by Agricultural Advisory Centers in southern Finland, 1982—1986. Agric. No buildings With buildings All transactions Adv. _ . No. ha Total No. ha Total No. ha TotalCenter , , ,ha ha ha Uusimaa 101 5.31 536.49 26 7.66 199.26 127 5.79 735.75 Nylands Svenska 55 6.91 380.22 7 6.42 44.97 62 6.86 425.90 Vars.-Suomi 215 5.27 1132.26 26 8.21 213.51 241 5.58 1345.77 Finska Hus- hällningss. 5 6.88 34.40 2 14.75 29.50 7 9.13 63.90 Satakunta 172 4.45 765.42 46 5.81 267.32 218 4.74 1032.74 Pirkanmaa 81 4.67 373.20 8 7.26 58.07 89 4.90 345.27 Province of Häme 126 5.42 682.96 39 5.89 229.56 165 5.53 912.52 Itä-Häme 45 6.29 283.23 9 10.40 93.63 54 6.98 376.86 Kymenlaakso 97 4.82467.37 10 6.3263.18 107 4.96530.55 Etelä-Karjala 69 3.76259.10 10 4.4644.61 79 3.84303.71 Southern Finland Total 966 5.094919.65 183 6.801243.61 1149 5.366163.26 208 in the Agricultural Advisory Center areas of Varsinais-Suomi and Satakunta. The quanti- ty of sold land in these areas exceeded 1 000 hectares during the five years investigated. The smallest number of transactions reported by the market price register was observed in the area of Finska Hushällningssällskapet. In the Agricultural Advisory Center areas of Itä- Häme and Etelä-Karjala the number of trans- actions also remained small; nevertheless, the total area of transactions of supplementary arable lands exceeded 300 hectares. Thus, 966 transactions of solely supplemen- tary arable land were concluded during the period 1982—1986. The average area was 5.09 ha, while in transactions of supplementary arable land including buildings the area was considerably larger, 6.8 ha. During the inves- tigated period, there were 183 transactions of supplementary arable lands including build- ings and 1 244 ha of land were exchanged, which represents about a fifth of the total area of supplementary arable land. The activities of land acquisition by the Na- tional Board of Agriculture are principally controlled by the Farm Act, implemented on April 1, 1977. The National Board of Agricul- ture is able to acquire land areas destined for the expansion of farms by making voluntary transactions or by exchanging land. The cost of transactions, which must not exceed the current price, is paid from the funds of the Agricultural Development Fund. Land which is not able to be used immediately for the ex- pansion of farms can also be purchased. Nevertheless, the main objective is to acquire land that is suitable as additional land or for improving the structure of holdings (Anon. 1977). When making purchases, the priorities are for farms whose arable land area is con- siderable and for land related to the farm clo- sure pension scheme. Table 26 shows how the purchases of arable land by the National Board of Agriculture in the bookkeeping re- gion of southern Finland developed during the period 1972—1986. Table 26. Quantity of arable land (ha) purchased by the National Board of Agriculture in southern Finland, 1972—1986. Year Total I cl.* II cl.* 11l cl. a trans- (> 120 pts) (70—120 pts) (<7O pts) actions ha ha ha ha 1972 37 299.40 118.15 140.55 40.70 1973 20 209.27 97.38 98.37 13.70 1974 65 657.23 412.87 215.96 28.40 1975 70 529.49 247.56 241.14 40.79 1976 50 384.21 147.62 190.49 46.10 1977 55 638.27 263.00 303.15 72.12 Subsurface drainage Open drainage 1978 77 849.40 200.90 648.50 1979 80 903.90 264.40 639.50 1980 73 667.85 84.81 583.04 1981 57 648.52 244.24 404.28 1982 40 448.43 167.14 281.29 1983 32 249.52 88.82 161.30 1984 35 407.32 172.70 234.62 1985 34 212.67 30.50 182.17 1986 37 469.45 310.36 159.09 Period 1972 1986 762 7 574.90 Notes: These points are from tax grading during the period 1972—1977. 209 Table 26 shows an increase in land pur- chases until 1979 and a decrease thereafter. The number of land purchases was particu- larly small in 1983 and 1985 when the state purchased less than 250 ha annually. The an- nual average quantity of arable land pur- chased by the state was 505 ha during the in- vestigated period, i.e. less than half of the an- nual area of supplementary arable land pur- chased according to the market price register. In the land acquisitions of the National Board of Agriculture, the quality of arable land has been given as an index according to the tax grading. Since 1978, transactions have been classified according to whether arable land has open or subsurface drainage. Arable land is considered to be subsurface drained, if over half of the area is so drained. The arable land purchased by the National Board of Agriculture has mostly been open drained, except in 1986. The area of land pur- chased with subsurface drainage amounted to 1 563 hectares during the period 1978—1986, which represents 32 % of the total area of pur- chased arable land. The relation between open and subsurface drained arable land varies an- nually. Naturally, this has an effect on prices weighted by surface areas. Class I included 47.3 % of the arable land purchased during 1972—1977, class II 43.8%, and class 111 8.9%. 5.2. Price level of supplementary arable land and its development 5.2.1. Questions concerning the use and assessment of data on market prices In the market data approach, the value of real estate is defined, in most cases, accord- ing to market price data. This approach re- quires that enough observations on free mar- ket transactions are available and that they are sufficiently homogenous. In addition to price comparisons, the market valueof arable land under cultivation can be determined by using expert appraisals of the price level. This is the approach applied by the National Board of Agriculture in its land purchases; experts in agricultural administration appraise the prop- erty items of the real estates. The value of ara- ble land is then based on the subjective view of the assessor concerning local current price. The market value of arable land depends on whether it is purchased together with a farm as a whole, or as supplementary arable land. The value of supplementary arable land is de- termined according to the additional benefit it brings, which may be considerable in in- dividual cases. When connected with a farm as a whole, the value of arable land is deter- mined according to an average level of bene- fit. In the market price data, the price of ara- ble land purchased in connection with a farm as a whole is determined by subtracting the value of other property parts from the total market price. If the total value of property parts is higher than the market price actually paid, the value of property parts is reduced by the same proportion (Mäki 1964, p. 105). This situation is common in appraisals of farms as a whole based on the summation of separately assessed property parts. After a wholesale discount on the total valueof prop- erty parts, the current price can be appraised. The discount is lower, the greater thepropor- tion of arable land of the total area. Cor- respondingly, the discount percentage is in- creased when there is a large proportion of forest land and buildings in the total area. Transactions of arable land alone contain lit- tle or no discount, depending on the scale of the transaction (compare Ryynänen 1978 b, p. 11). In appraisals based on market prices, some compromises are made concerning the repre- sentativeness of the data, otherwise the num- ber of observations would not be sufficient. According to Wiiala (1976, p. 15—21), a representative real estate transaction meets the following requirements: similarity, equivalence of the type of real estate, consistency of potential use, similarity of size, 210 same location, congruence of transaction time, and freedom of transaction. Compulsory auctions, family transactions, and transactions including so-called “back- handers” are not included in the market price data. Even though the non-representative transactions could be discarded on the above- mentioned grounds, the material would still contain variations in unit prices that derive from the heterogeneity of the objects in trans- actions and from the individual decisions of the parties involved. The variations in mar- ket prices are represented by average figures, as well as by parameters for dispersion and for skewness in distribution. The probable market price, concerning comparable trans- actions, can be indicated as either arithmeti- cal average price, area-weighted average price, mode price, or median price. In fact, the probable market price corre- sponds to the mode price, which signifies the class containing most observations. Mode is adapted for use with classified materials, and the choice of the interval of the class has an effect on its numerical value. The arithmetical average value, the weighted average price, and median are adapted for continuous data. The median price divides the data in half, thus, it is not as sensitive to ex- treme values as are average values. Often just low extreme values are non-representative (containing hidden relationship, black market price, etc.). Wiiala (1976, p. 21) recom- mends the use of the median price as the mar- ket value. However, matters related to appraisals sup- port the use of the arithmetical average value. The average value is required for the calcula- tion of several statistical parameters. The area- weighted average value takes into considera- tion the size of the transaction. Consequent- ly, high unit prices paid for small areas do not raise the average price as much as the use of the arithmetical average value would do. On the other hand, a few large transactions re- ceive considerable weight when a small mate- rial is weighted. One prerequisite for the application of the market data approach is the equivalence of the size of observations and the objects appraised. In this case, the market value of additional areas should be determined on the basis of transactions already made in the district con- cerning supplementary arable land of similar areas. In practice, all transactions of sup- plementary arable land have to be considered for the appraisal, because the totalnumber of transactions is low. In this study, an approach was selected in which the total market price of an arable land area is divided by its area. The distribution of the unit price (FIM/ha), determined in this way, is described by averages and dispersion measures. The average figure aims to repre- sent the average size or quality of the variable, or the position of the distribution in the con- tinuum. The dispersion describes the scatter around the average figure. Average prices, determined in the follow- ing way, are employed in this investigation: 1. arithmetical average price xr = L _ . , . . Ep,xi2. area-weighted average price x-—IL, 1L ,Zpj in which Xj = price per hectare transacted Pj = area of the additional land N =number of transactions. 3. median price which divides the market price data in half. If the number of obser- vations is even, the median price is the mean value of the two middle observa- tions. Dispersion figures 1. standard deviation 5 =\l’L(x j -x) 2/(n- 1) 2. interval of variation w =(xf"\ xjt'ax) 211 5.2.2. Price of supplementary arable land and its development according to the market price register The prices of supplementary arable land, appraised on the basis of data from the mar- ket price register, and expressed with different average figures for the bookkeeping region of southern Finland, are presented in table 27. A strong variation in the price of supplemen- tary arable land in districts belonging to dif- ferent Agricultural Advisory Centers is ap- parent. The year 1982 has been chosen for closer examination, because it is the first year from which data from the market price regis- ter concerning the whole year are available. Appraisals are made on the same basis as those for subsequent years, even though their conclusions will not be presented in detail. The median price remained lower than the average prices. Thus, there is skewness in the distribution to the right, i.e. there are relative- ly more transactions of low per hectare price in the data. Consequently, the choice of ap- proach has a considerable effect on the final results of the search for the average price and probable market price of arable land. The price of supplementary arable land proved highest in the Agricultural Advisory Center of Varsinais-Suomi (24 650 FIM/ha), while in Satakunta, the price was also almost on the same level. The price of supplementary arable land was lowest in the Agricultural Ad- visory Centers of Etelä-Karjala, Kymenlaak- so, and Pirkanmaa, where it varied from 13 900 FIM to 14 700 FIM per hectare. Standard deviations of per hectare prices were relatively large. The standard deviation Table 27. Price of arable land (FIM/ha) in transactions of solely supplementary arable land in 1982 in southern Finland, calculated by different measures. Agricultural No. Arithmetical Standard Variance Weighted Median Advisory average deviation % from average price Center price the average price price Uusimaa 17 17 912 4 518 25.7 17 114 17 200 Nylands Svenska 7 17 571 6 628 37.7 16 219 15 500 Varsinais-Suomi 44 24 725 9 253 37.4 24 650 23 650 Satakunta 26 23 708 8 537 36.0 24 344 23 250 Pirkanmaa 14 14 671 4 981 33.9 14 720 13 850 Province of Häme 14 17 571 6 252 35.6 18 247 17 500 Itä-Häme 7 16 786 10 931 65.1 16 907 14 000 Kymenlaakso 16 12 831 5 072 39.0 14 217 14 000 Etelä-Karjala 11 14 182 5 011 35.3 13 877 15 000 Southern Finland 156 19 628 8 602 43.8 19 872 17 000 Table 27 shows that the area-weighted aver- age price (19 872 FIM/ha) is a little higher than the price determined as an arithmetical average value (19 628 FIM/ha). In fact, the average price of arable land represents the average price of arable land for sale, for which the price of each hectare of arable land is as- sessed separately, i.e. by weighting the unit area price of each transactionby the area for sale. As for the arithmetical, non-weighted average value, it represents the average price of transactions of arable land. Transactions of areas of different size do not have an effect. in the whole data was 8 602 FIM. The largest dispersion was noted in the Agricultural Ad- visory Center of Itä-Häme which had the smallest number of observations. Figure 12 shows the price development of supplementary arable land by Agricultural Advisory Centers and a sharp price rise dur- ing the period 1982—1986. During the period in question, the price of arable land increased by 60 % in southern Finland. On the other hand, in 1986 the nominal price level of ara- ble land increased at a yearly rate of 12.5 %, as determined by the formula calculating the 212 compound interest of the value for the first and the last year2 . In the regional examination, the Agricul- tural Advisory Centers of Satakunta and Var- sinais-Suomi are of interest because the prices of supplementary arable land rose during each year investigated and the price levels were the highest. The prices of transactions of sup- plementary arable land were the lowest in Etelä-Karjala, Itä-Häme, and Pirkanmaa. When examining prices and their develop- ment by Agricultural Advisory Centers, it must be kept in mind that there have been few annual transactions in some of the districts. As a result, a more reliable picture of the price development can be gained from centers with the greatest number of transactions. Table 28 describes the classification of transactions of supplementary arable land in eight different price classes with respect to the central year of the investigated period. I V _| " ' —1 x 100, in which y„ =value of the \y°l last year, and in which y 0 = value of the first year Most transactions were made in Satakun- ta, Varsinais-Suomi, and Uusimaa. In addi- tion, it is worth noting that in the first two centers almost half of the transactions be- longed to the upper price class. In Finska Hushällningsällskapet only one transaction was made. Similarly, in Itä-Häme and Etelä- Karjala only a few transactions were made in 1984. Despite these inadequacies, a conception of the price differences of arable land in different centers can be formed. Earlier in this chapter, the price of sup- plementary arable land and its development was examined as nominal price. This indicated the regional price differencesand the changes in price development. However, inflation can considerably alter the picture conveyed by prices and their development. Therefore, in figure 13, the price level of supplementary arable land and its development is based on 1986 prices. The cost-of-living index has been used as the deflator. The figure shows that the real price of sup- plementary arable land was highest in 1985. The following year the average price of ara- Figure 12. Weighted average price of transactions of supplementary arable land (FIM/ha) by Agricultural Advisory Centers in southern Finland, 1982—1986. 213 Table 28. Division of transactions of supplementary arable land (without buildings) in price classes and lowest and highest per hectare prices by Agricultural Advisory Centers in 1984. Price Uusi- Nyl. Vars. Finska Sata- Pirkan- Häme Itä- Kymen- Etelä- Southern FIM/ha maa svenska Suomi Hush, kunta maa Häme laakso Karjala Finland total < 15000 1 32 242235 24 15000—19999 3 5 1 2 1 5 1 6 5 29 20000—24999 5 1 3 I 2 4 6 1 2 1 26 25000—29999 10 1 5 4 5 1 1 27 30000—34999 21 8— 2232 20 35000—39999 3 4 3 3 2 1 16 40000—44999 2 4 3 1 10 >45000 2 16 17 1 4 40 Total 24 13 36 1 37 16 31 8 14 12 192 Min 8831 12889 5269 23132 10961 4324 5463 9613 9682 8602 4324 Max 39568 51958 92869 23132 80663 56967 65729 38364 43008 27242 92869 Weighted average price 25120 19996 37713 21400 40554 17368 28471 25691 18105 14373 28850 ble land in the research region was c. 1 000 FIM/ha lower. The average real price of arable land in the research region varied from 25 000 FIM to 33 000 FIM during the period 1982—1986. Though the real price of arable land mainly decreased during the last year, the price of arable land reached its peak value in 1986 in Varsinais-Suomi. During this same year, the price of arable land exceeded 43 000 FIM/ha in Satakunta and Varsinais-Suomi. In these areas, the per hectare prices of arable land were the highest in each year. The next highest prices were found in the Agricultural Advisory Figure 13. Real weighted average price (FIM/ha) of transactions of supplementary arable land, by Agricultural Advisory Centers, in southern Finland, 1982—1986, at 1986 prices. 214 215 Center of Häme, the annual price variation being between 23 000 FIM/ha and 35 000 FIM/ha. On the other hand, in Pirkanmaa the prices (18 600—22 500 FIM/ha) remained lower than those of Häme as a whole. The lowest prices for arable land were paid in Itä- Häme, Etelä-Karjala, and Kymenlaakso. The prices in these areas and Pirkanmaa were clearly lower than in the Finnish and Swedish speaking areas of Uusimaa. In addition to the transactions of solely ara- ble land examined above, the market price register also includes transactions of sup- plementary arable land with buildings. The number of transactions including buildings de- veloped in the following way: Year Number of transactions 1982 41 321983 251984 1985 36 491986 in three cases, transactions included only dwellings. Dwellings and farm buildings were included in 14 transactions, farm buildings in 38 transactions, and other buildings in 128 transactions. There was no specification of the value of buildings, but they were included in the price declared in the market price regis- ter. Table 29 indicates the weighted average prices of transactions of supplementary ara- ble land including buildings, at 1986 prices. The prices in transactions of supplementary arable land areas including buildings were not significantly higher than those in transactions of solely arable land areas, during 1982 and 1983. However, during the period 1984—1986 buildings, especially dwellings and farm build- ings, have been a source of considerable price rises. It is worth noting that thereal prices of transactions, including farm buildings, were higher only during the last two years than the prices of supplementary arable land without farm buildings. Nonetheless, the real prices of transactions, including other buildings, ex- ceeded in every year the prices of supplemen- tary arable land without buildings. A general picture of the price level of ara- ble land can be seen in table 30 which presents appraised weighted average prices by Agricul- tural Advisory Centers during the period 1984—1986. Except for Finska Hushällnings- sällskapet, arable land transactions in differ- ent areas have been so numerous that the aver- age prices in each area can be considered as reliable indicators of the price level. Table 30 indicates that during the period 1984—1986, arable land price has been clear- ly highest (41 600—42 400 FIM) in Satakunta and Varsinais-Suomi. In these centers the price was 30 % higher than the average level in southern Finland. Also in the Häme area, the price level was higher than the average level in the research region as a whole. The fourth highest price level was observed in the Finnish- and Swedish-speaking areas of Uusi- Table 29. Weighted average price of transactions including buildings (FIM/ha) during 1982—1986a . Year No buildings Including buildings Dwellings Dwellings Farm Other and farm buildings buildings buildings 1982 25 240 24 326 22 814 24 875 27 907 1983 27 692 37 193 21 149 31 815 1984 31 641 95 817 31 922 39 532 1985 33 004 114 307 55 415 33 648 36 246 1986 31 920 59 915 38 792 38 386 Notes : Prices are deflated to the 1986 level by the cost-of-living index. Table 30. Prices of arable land (FIM/ha) in transactions of supplementary arable land without buildings during the period 1984—1986, by Agricultural Advisory Centers3 . Agr. Adv. Center Number of trans- actions Weighted average price Relative price of arable land (Southern Finland = 100) Uusimaa 68 28 832 29 706 41 592 25 369 42 376 20 243 33 746 22 028 24 767 19 381 89 Nylands Svenska Varsinais-Suomi 36 92 127 129 Finska Flushällningss. Satakunta 4 79 116 131 Pirkanmaa Fläme 48 63 90 105 Itä-Fläme 31 68 Kymenlaakso Etelä-Karjala 64 77 50 60 Southern Finland 634 32 238 100 Notes: Prices are deflated to the 1986 level by the cost-of-living index. maa, where the weighted price of arable land remained 10°/o lower than the average level in southern Finland. The prices of transactions of supplementary arable land were lowest in Pirkanmaa, Itä- Häme, Kymenlaakso, and Etelä-Karjala. In these centers, prices were about a third lower than the average level. 5.2.3. The price of supplementary arable land and its development according to purchases of arable land by the National Board of Agriculture In purchases by the National Board of Agriculture the appraisal of arable land is based on the local general price level of free Table 31. Price of arable land (FIM/ha) in the purchases of the National Board of Agriculture in southern Finland, 1972—1986. Year All transactions Quality classes of arable land Average Change from I class II class 111 class price preceding year % 1972 3 127 3 830 2 866 1 985 1973 4 811 +53.9 5 136 4 631 3 788 1974 5 516 + 14.7 5 906 4 971 3 989 1975 5 694 + 3.2 6 657 5 036 3 739 1976 6 575 +15.5 7 724 6 027 4 413 1977 8 129 +23.6 9 323 7 564 6 153 Subsurface drained Open drainage 1978 8 973 + 10.4 11 298 8 252 1979 10 091 +12.5 12 952 8 907 1980 9 236 + 8.5 12 715 8 730 1981 11 912 +29.0 15 088 9 993 1982 13 655 +14.6 17 777 11 206 1983 13 557 - 0.7 17 893 11 185 1984 15 968 +17.8 19 067 13 686 1985 15 315 -4.1 23 224 13 991 1986 22 993 +50.1 25 642 17 826 216 market transactions. Appraising the arable land price by the local current price means that appraisals are based on the price level of trans- actions of supplementary arable land. Thus, the National Board of Agriculture is excercising extreme caution with respect to the pricing of purchases. It has the policy that pricing must not be the cause of a rise in the price of land. Those prices are not considered which result from price competition between differentbuyer groups. The average prices of thepurchases of the National Board of Agri- culture, weighted by area, are given in table 31. In the purchases of the National Board of Agriculture, from 1972 to 1986, the price of arable land has risen 7.4-fold. The average price rise of arable land in purchases was 15.3 % a year. The price difference between subsurface drained and open drained arable land developed in the following way during the research period: 1978 3 046 ha/FIM 1979 4 045 » 1980 3 985 » 1981 5 095 » 1982 6 571 » 1983 6 708 » 1984 5 381 » 1985 9 233 » 1986 7 816 » 37 % 45 % 46% 51 % 58 % 60% 39% 66% 44%. In 1978, subsurface drained arable land cost 37 % more and in 1983, 60 % more than open drained arable land. In 1984 the difference was less, but increased again during the last two years investigated to a level that almost corresponded to the drainage cost per hectare. If the prices paid annually by the National Board of Agriculture for arable land are trans- formed to 1986 prices, the real price develop- ment of arable land can be determined (ap- pendix 7, table 7.1). The table shows that the price of arable land was the highest in 1973, 1979, 1982, 1984,and 1986. During the latter Figure 14. The development of nominal and real prices of arable land (FIM/ha) and the real price trend, 1972—1986, according to the purchases by the National Board of Agriculture. 217 year, the price of arable land was excep- tionally high, i.e. about 40 % higher than the 1980—1985 level. The real price of arable land, as a three year mean for the last three investigated years (1984—1986), was 18 800 FIM/ha. All in all, the real price of arable land in the purchases of the National Board of Agriculture, despite rather large annual variations, has undergone only a slight in- crease (figure 14). The trend line in figure 14 is estimated from the following model: y= 14 504 + 284t, where t = 0... 14 (years 1972—1986). The yearly change was on an average 1.7 °7o, calculated as compound interest. In the pricing of arable land purchased by the National Board of Agriculture, the aver- age value (above), weighted by area, has been used along with data from the market price register. If the arable land purchased by the National Board of Agriculture was represen- tative with respect to the proportion of sub- surface drainage, approximately 50 % in southern Finland (Anon. 1980 b), the price level it paid for arable land rises closer to the general price level (when appraisals are made by the arithmetical average value of subsur- face or open drained arable land). Table 32. The average value of prices of subsurface and open drained arable land in the purchases of the National Board of Agriculture, 1978— 1986». Year Price level of arable land (FIM/ha) 1978 9 775 1979 10 930 1980 10 723 1981 12 541 1982 12 492 1983 14 539 1984 16 377 1985 18 608 1986 21 734 The percentage of subsurface drained arable land for free sale does not correspond to the subsurface drained percentage of the cultivated arable land as a whole (com- pare Ala-Kantti 1981, p. 38), The results in table 32 indicate that the price level examined in this way was, except in 1986, higher than theprice level shown in table 31. At the same time, annual price changes are smaller than when using the weighted average price, because in the figures of table 32, the ratio between subsurface and open drained arable land prices remains constant. 5.2.4. Price differences between transactions of supplementary arable land in the market price register and thepurchases of arable land by the National Board of Agriculture Transactions of arable land under cultiva- tion (without buildings) in the period 1982 1986, derived from the statistics of the Na- tional Board of Land Surveying, primarily de- scribe the prices paid for land by farmers. On the other hand, the market price register also includes transactions, by the National Board of Agriculture, concerning supplementary ara- ble land (including arable land under cultiva- tion). Thus, the price information of the reg- ister does not entirely represent the price level on the free market. In addition, transactions between farmers also include transactions financed with loans in accordance with the Farm Act. In transactions of supplementary arable land without buildings, the area-weighted average price in southern Finland has been de- termined using both nominal and real prices. By comparing arable land prices paid b the National Board of Agriculture, with info ma- tion given by the market price register, the general relationship between these two price statistics and price development can be found (table 33). Table 33 shows that the price level in the purchases of arable land by the National Board of Agriculture is clearly lower than the price level in the market price register. Ara- ble land purchases by the National Board of Agriculture have varied on a yearly basis from 48 % to 72 % of the prices in the market price register. The price ratio was the lowest (0.48) 218 in 1985, when the purchases of arable land by the National Board of Agriculture remained relatively insignificant, as well as being weighted towards open drained land. The cor- respondence between market prices was the highest during 1986 when about two thirds of the land purchases of the National Board of Agriculture concerned subsurface drained ara- ble land, while during the preceding years the arable land purchased by the National Board of Agriculture mainly concerned open drained arable land. The average price level of the pur- chases of the National Board of Agriculture represented 60 % of the level of the market price register during the period 1982—1986. Table 33. A comparison of the price of arable land (FIM/ha) according to the National Board of Land Surveying (NBLS) and the National Board of Agriculture (NBA), for the period 1982—1986 in the investigated region. Year Transactions of Purchases NBA/ supplementary by NBA NBLS arable land by NBLS 1982 19 872 13 655 0.69 1983 23 636 13 557 0.57 1984 28 850 15 968 0.55 1985 31 823 15 315 0.48 1986 31 926 22 993 0.72 Average 0.60 It would seem that the differences related to the quality and location of arable land have had an effect on the rather considerable dif- ferences between the market price register (NBLS) and the National Board of Agricul- ture concerning the annual market prices. A more detailed examination of these effects would have required complementary informa- tion and was not possible within the frame- work of this investigation. Comparisons in different regions between arable land prices paid by the National Board of Agriculture and price information from the market price register are impeded by the Na- tional Board of Agriculture’s small numberof transactions. To improve the situation, a price comparison is made by using three year moving averages; the total number of pur- chases amounting to 106 during the three years. First, the prices of 1984 and 1985 are converted to the 1986 price level. The aver- age value, weighted by the area of purchases of arable land by the National Board of Agri- culture, was then assessed at 19 555 FIM/ha, which is about 12 700 FIM lower than the ara- ble land price appraised in a corresponding way from the market price register. There are also rather large regional variations in the price level (table 34). Table 34. Regional variations in the price of arable land (FIM/ha) appraised according to data from the market price register (NBLS) and the National Board of Agriculture (NBA), 1984—1986, at 1986 prices. Agr. Adv. Transactions of Purchases by NBA/NBLS Center supplementary NBA arable land by NBLS Uusimaa 28 832 19 134 0.66 Nylands Svenska 29 706 16 431 0.55 Varsinais-Suomi 41 592 20 197 0.49 Finska Hushällningss. 25 369 22 908 0.90 Satakunta 42 376 16 781 0.40 Pirkanmaa 20 243 19 854 0.98 Häme 33 746 24 679 0.73 Itä-Häme 22 028 14 032 0.64 Kymenlaakso 24 767 20 870 0.84 Etelä-Karjala 19 381 15 972 0.82 Southern Finland 32 238 19 555 0.61 219 Thus, in transactions for supplementary arable land, the National Board of Agricul- ture paid approximately 61 % ofthe price paid in transactions in the market price register, with a regional variation of 40 % to 98 %. It is worth noting that the relative prices paid by the National Board of Agriculture were par- ticularly low in Varsinais-Suomi and Satakun- ta, where the price level of transactions of sup- plementary arable land in the market price register were the highest. Besides the small number of arable land transactions by the National Board of Agricul- ture, price differences were mostly influenced by the lack of competition in the purchases of the National Board of Agriculture. In ad- dition, it is commonly believed that the pur- chases of the National Board of Agriculture concern remotely located arable land and land otherwise of low market value. 220 6. COMPARISON BETWEEN CAPITALIZED VALUE AND MARKET PRICES OF SUPPLEMENTARY ARABLE LAND The comparison between capitalized value and market prices can only be an approxima- tion. It is clear that productive and economic conditions on individual farms may vary con- siderably. The comparisons that will be presented cannot thereforebe generalized for all farms. Similarly, factors related to produc- tion structures, or types of farms, and the size of arable lands, prevent the generalization of the results. Capitalized value is defined above only for two typs of farms, cattle farms specialized in dairying and grain farms. In the appraisal of the capitalized value of supplementary arable land, the final results depend on, in addition to the importance of return, its duration and the capitalization rate employed. Consequently, appraisals were made using three different interest rates and durations. The marginal tax rate was also taken into consideration for the determination of the net marginal operating margin that is to be capitalized on the basis of the results of the regression models. For cattle farms a mar- ginal tax rate of 40 % was selected, and 50 % for grain farms. Figure 15. Comparison between the capitalized value of supplementary arable land and market prices on cattle farms, at 1986 prices. 221 Market prices for supplementary arable land are based on the prices of arable land purchased by the National Board of Agricul- ture during the period 1972—1986, as well as market price statistics of the National Board of Land Surveying from 1982 to 1986. The level of the capitalized value and market prices are described in figure 15. In figure 15, the columns for each year de- scribe present values capitalized at interest rates of 3 °/o, 5 °7o, and 7 % for different du- rations. The results indicate a continuous in- crease of the capitalized value of the arable land on cattle farms during the research pe- riod. In general, when the returns from sup- plementary arable land were capitalized for a period of 15 years, the capitalized value of supplementary arable land on cattle farms in- creased at least to the price of arable land paid by the National Board of Agriculture. The capitalized value determined for ten years did not reach the price level paid by the National Board of Agriculture during all the years in- vestigated. Further, a duration of five years proved too short for the capitalized value to rise to the market price level in any of the years investigated. The capitalized value of supplementary ara- ble land, determined for the period 1982— 1986 (which describes the value of arable land in the most reliable way, rather than for sep- arate years), exceeded the price paid for ara- ble land by the National Board of Agriculture when appraised at each interest rate and for a duration of 15 years. On the other hand, when an interest rate of 5 % and a duration of 10 years were used for capitalization, the capitalized value did not reach the market price level. On the basis of these comparisons, it may be noted that therelationship between the capitalized value of supplementary arable land on cattle farms, and the price of land paid by the National Board of Agriculture, has remained almost unchanged during the years investigated. On the other hand, the price according to the market price register of the National Board of Land Surveying, almost 30 000 FIM/ha, proved to be clearly higher than the capitalized value of supplementary arable land. Figure 16. Comparison between capitalized value and market prices of supplementary arable land on grain farms, at 1986 prices. 222 The correspondence between the capitalized value and market prices of supplementary ara- ble land on grain farms is presented in figure 16. The return from supplementary arable land on grain farms was assessed for three consecutive years to avoid yield variation problems, but excluding the last period of ex- amination. Farms that had maintained the same production structure during the three or five year periods were accepted for consider- ation. Because the capitalized values of ara- ble land on grain farms remained generally lower than for cattle farms, they reached the level of prices paid by the National Board of Agriculture only during 1982—1986 when using a capitalization rate of 3 % and a dura- tion of 15 years. During all other investigated periods, the capitalized value of supplemen- tary arable land on grain farms remained lower than the market prices of the National Board of Agriculture and still lower than the market prices of the National Board of Land Surveying. 223 7. RESULTS AND CONCLUSIONS 7.1. Return from supplementary arable land Returns from supplementary arable land have been determined, for given periods, for cattle and grain farms using the Cobb-Douglas production functions. The farms considered were those which had practiced continuous bookkeeping since 1968. During this period, the current tax system in agriculture has been in force. The results of the study are based on both cross sectional and time series data. In both farm types examined, production functions analyses showed an obvious increase in return from supplementary arable land as a function of time. After dummy variables (fu/ha) were added to the regression model, the effect of yield level on return from sup- plementary arable land was assessed. Especial- ly, the marginal value product on grain farms proved, according to expectations, sensible to variations in yield level. On the other hand, on cattle farms, yield levels had a smaller ef- fect on the return from supplementary arable land. This is understandable, because during rainy summers, crop yields may remain low, whilepasture and grass yields increase with no notable deterioration in quality. The results of cattle farms for the years 1972, 1976, and 1980 are based on the use of cross sectional data for those years. However, because of the rather large annual variation in yields on grain farms, the determinationof their marginal value product demanded the use of three-year means. Therefore, the return from supplementary arable land on grain farms was estimated for the periods 1975— 1977 and 1979—1981. Thus, the return from supplementary arable land in both farm groups could be reliably determined only for the last reseach period, 1982—1986. Only farms that had continuously maintained the same production structure during theresearch period were accepted for analysis. The effect of the yield level variable describing the quality of arable land on the marginal value product from arable land during 1982—1986 is shown in table 35. Table 35. The effect of the yield level variable. Marginal value Yield (fu/ha) product (FIM/ha) <25004013 2500—3300Cattle farms 4147 4294 >3300 <25004089 Grain farms 2500—33004919 5512 >3300 The average feed unit yield (fu/ha) in both farm types was 2 500—3 300 fu/ha during the period concerned. The variation in arable land quality on cattle farms, measured by yield level, appeared to be smaller than on grain farms. With an increase in yield level, the mar- ginal value product also increased in both farm groups. Consequently, the use of the yield level dummy variable to describe the quality of agricultural landproved a viable so- lution. The solution had, in fact, been success- fully applied previously, e.g. by Locken et al. (1978). In addition, the results given by the present investigation regarding the return from supplementary arable land receive sup- port from the results of Elstrand (1980). El- strand applied production functions to the analysis of the importance of land rent on 224 Norwegian dairy and grain farms. Cross-sec- tional data were used concerning Norwegian bookkeeping farms. For these two farm types, the return from supplementary arable land on grain farms also showed itself to be higher than on dairy farms. The estimation of the return from supple- mentary arable land has been based on the use of a production function explaining gross re- turn in agriculture. Therefore, the marginal value product from supplementary arable land, assessed by parameter estimates, de- scribes the growth in gross return in agricul- ture when the arable land area is increased by one hectare. The farmer is unable to use the additional gross return exclusively for the ac- quisition of supplementary arable land, but variable costs derived from cultivating the supplementary land must be subtracted. As a result, those variable costs that were not in- cluded as independent variables in the regres- sion models, were subtracted from the mar- ginal value product of supplementary arable land. The operating margin assessed in this way does not, by definition, absolutely cor- respond to the concept of operating margin, which represents a compensation for the fixed production factors of the farm. The operating margin so determined indicates the compen- sation per hectare that remains for arable land acquisition before taxation. In practice, arable land acquisition is financed either in totality or partially by tax- able income. Therefore, taxes were subtract- ed from the marginal operating margin. In the determination of taxes, reference was made to Siirola (1988, p. 55 —58). On this basis, the marginal tax rate, due to additional income in agriculture, was set at 40 ®/o for cattle farms and 50 % for grain farms. The marginal tax rates were selected because the investigated bookkeeping farms can be considered to be rationally managed and practicing intensive production, i.e. they have participated in bookkeeping activities for a long period of time. The dependency between the marginal value product from arable land and the size of farm is of interest. The benefit gained from sup- plementary arable land is supposed, following theory, to be the highest in the smaller farm size classes. This contention was investigated by dividing the research data into two size classes. By using production function analy- sis, the marginal value product from sup- plementary arable land on cattle farms of less than 25 ha, determined for theperiod 1982— 1986, varied from 4 775 FIM to almost 5 200 FIM. On larger farms, the marginal value product was no more than 1 360—1 700 FIM. The results, therefore, supported the expec- tations, and indicated that economic results on cattle farms in the larger size class are prob- ably more dependent on other production in- puts and their utilization than on supplemen- tary arable land. On the other hand, on grain farms, the ad- dition of arable land seemed to favour the larger farm size class. On farms of less than 30 hectares, the marginal value product from supplementary arable land per hectare varied from 3 400 to 4 500 FIM, while the variation in larger farms was between 4 300 and 5 400 FIM. The different development of marginal value product from supplementary arable land between the farm types results because rather small grain farms are, to a large extent, a secondary occupation for the farmer. Further, grain farms have more machinery, with their attendant costs, compared to cattle farms, and unit costs decrease along with an increase in the area of arable land. Moreover, the years 1982—1986 were characterized by good weather conditions and satisfactory yield levels. Thus, an increase in arable land area on rather large grain farms proved economi- cally justified within the framework of the production technology employed. The analysis of the differential return, cal- culated on the basis of the taxable net return in agriculture, also supported the different de- velopment of the return from supplementary arable land as a function of farm size with re- spect to both cattle and grain farms. The anal- ysis was based on the method presented by Elstrand (1980), which showed that the 225 differential return from supplementary arable land was largest (4 769 FIM/ha) when moving from the smallest size-class of cattle farms, less than 20 ha, to a larger class. As for grain farms, the greatest increase in differential re- turn (4 023 FIM/ha) was realized when moving from the 20—30 ha class to the 30—50 ha class. It is worth noting that the differen- tial return, remaining as interest on capital in- vested in agriculture on farms of 30 ha or larger, was still higher than the average tax- able net return for the farm groups in ques- tion. 7.2. Market prices of supplementary arable land Two different price statistics were used when price information on arable land was collected: the market price register of real es- tates of the National Board of Land Sur- veying, and the statistics on land acquisitions of the National Board of Agriculture. From the former statistics, only information con- cerning the years 1982—1986 was available, while the arable land acquisitions of the Na- tional Board of Agriculture have been re- corded since the 1960’5. The market price reg- ister also includes the land acquisitions by the National Board of Agriculture, as well as transactions on supplementary arable land financed according to the Farm Act (arable land areas of more than 2 ha). The charac- teristics of both price statistics have been ex- amined in more detail in section 5.1. The price development of arable land in southern Finland can be determined for the period beginning in the early 1970’s only by using the price statistics on transactions on supplementary arable land by the National Board of Agriculture (figure 14> This indi- cates a considerable rise in nom. ial prices, but it is rather moderate in real terms. The price development of supplementary arable land in land acquisitions of the Na- tional Board of Agriculture therefore diverges considerably from the price development in Sweden, Denmark, Holland, or France. In these countries, the real rise in land prices was rapid during the final years of the 1970’s (Andersson 1989, p. I—4).1 —4). Andersson (1989 p. 88—89) considered that the main reason for the rise of land prices in Sweden was the low real interest rate prevailing during that period and the simultaneous price rise of agricultural products on the world markets. The downturn in the trend of land price in the 1980’s was connected with the decrease in prices of agricultural products. This would mean that Swedish agricultural protectionism did not completely guarantee profitability in an agricultural sector that was facing price effects from international markets. Comparisons show that the price level of arable land in acquisitions by the National Board of Agriculture was c. 60 % of the prices recorded in the market price register. The majority of additional land is acquired, in practice, by transactions between farmers, and financed by the National Board of Agricul- ture. Thus, the statistics on loans involved in purchases of supplementary arable land would have been, in principle, a useful source mate- rial for price comparisons of arable land. During the period 1982—1983, in transac- tions in accordance with the loan statistics, the price of arable land was even higher than prices recorded by the market price register. The price development during these years sug- gests that loans granted for the purchase of additional land contributed to the increase in the price of arable land. A more probable in- terpretation might be that the loan-related price statistics include agricultural properties (not just land), the price effect of which is reflected in price per hectare. Therefore, be- cause of the duration of the investigated pe- riod and the unambiguous interpretation of the price statistics, the price statistics representing purchases of arable land by the National Board of Agriculture were chosen for the study and not the loan statistics. The market prices of supplementary arable land in Finland can be explained by factors endogenous to agriculture (Laurila 1988). These endogenous factors (market activity, 226 profitability in agriculture, quality of arable land, and yield level) explained 63 °7o of the variation of arable land prices. After an ex- ternal factor (employment) was integrated in the price model, the degree of determination of the model increased only by 4 percentage points. In this respect, the situation differs considerably from the price development of agricultural land in the United States, where Castle and Hoch (1982) have shown that capital gains explain half of the increase in land prices. Even though expectation values near large population centers in southern Fin- land have had some impact on the prices paid for land, the significance of such land acqui- sition on the agricultural land prices seems to be rather doubtful. 7.3. Correspondence between the capitalized and market values Capitalized value depends on both the in- terest rate used for capitalization and the du- ration of the return. In this study, the capitali- zation of return was realized at 3 %, 5 %, and 7 °7o interest rates while the corresponding durations were 5, 10, and 15 years. The de- preciable agricultural property items were ad- justed to correspond better with real market value(appendix 1). If bookkeeping values for property had been used, the value of depreci- able property items would have decreased from 1968 to 1986, even though investments were, on average, higher than tax deprecia- tions (compare Latukka 1989, p. 81 —82). Thus, the depreciable property, adjusted by the cost-of-living index, on cattle and grain farms, clearly increased. In fact, it doubled in 1986 compared to the taxable depreciable property items. Yet, according to Kukkonen (1990), a real interest of 5 % or 7 °7o may be considered to be quite high because, the level of real interest in the long term should follow the annual growth percentage of the gross na- tional product. The results of the study by the Farmers’ So- cial Insurance Institution were taken into con- sideration when selecting the duration of the return for the assessment of capitalized value. Within the framework of the Farm Closure Pension Act and those farmers continuing agriculture, following generation transfers, the mean age of farmers who had acquired supplementary arable land was found to be 44 years (Tolonen 1985, p. 17). In addition, with respect to loans via the Farm Act, the amortization period of loans for acquisition of supplementary arable land proved to be relatively short, mainly five to eighteen years in southern Finland. On the other hand, the social time preference for the capitalization of return is essentially longer than the individ- ual time preference. In this case, returns are supposed to continue infinitely. Comparisons between capitalized and mar- ket values showed that the capitalized value of supplementary arable land on cattle farms, during the whole period investigated, attained the market price paid by the National Board of Agriculture when the duration of return was 15 years. When a duration of return of 10 years was employed, the capitalized value of land no longer reached the price paid by the National Board of Agriculture in all pe- riods during the investigated years. The results for cattle farms are shown in figure 17. Each set of three columns in figure 17 represents a single time period or year and the column on the left of each triad describes the case in which the net marginal operating mar- gin of supplementary arable land is capital- ized at 3 % interest. The darkest part of the column describes a capitalized present value for a duration of 5 years to which have been added the cumulative difference of capitalized present values between 10 and 5 years, and be- tween 15 and 10 years. The height of the column indicates the capitalized present value for a duration of 15 years, at an interest rate of 3 °7o. Correspondingly, the middlemost column describes the cumulative present value at 5 % and the one on the right at 7 % interest. On the other hand, the present value on grain farms (figure 18) has remained clearly lower than the market prices paid by the Na- tional Board of Agriculture, except during the 227 228 Figure 17. Cumulative capitalized value (FIM/ha) of arable land on cattle farms and the price of arable land in purchases by the National Board of Agriculture, at 1986 prices. Figure 18. Cumulative capitalized value (FIM/ha) of arable land on grain farms and price of arable land in pur- chases by the National Board of Agriculture, at 1986 prices. period 1982—1986. Only during this latter period did the present value on grain farms reach the level of market prices paid by the National Board of Agriculture. Yield level has only a small effect on the capitalized value of cattle farms. If the aver- age feed unit yield during 1982—1986 was lower than 2 500 fu/ha, the capitalized value of arable land would decrease by little more than 3 %. If the quality of arable land, mea- sured by yield level, exceeded the productivi- ty of 3 300 fu/ha, the present value of sup- plementary arable land would be increased by 3.5 %, as may be observed in table 18. On the other hand, the impact of yield level was ob- viously greater on grain farms than on cattle farms. If yield levels remain below the median class, the present value would decrease by al- most 17 %. If the yield level per hectare ex- ceeded 3 300 fu, capitalized value would in- crease by approximately 12 %. Consequently, the effects of variation in yield level were manifested quite differently in the farm types investigated. Nevertheless, the variation did not change the general pic- ture concerning the capitalized value in the farm groups. On cattle farms, throughout the periods investigated, capitalized valuereached the price level for arable land paid by the Na- tional Board of Agriculture when the return was capitalized using all three capitalization rates based on a 15 year duration. On the other hand, the present value of supplemen- tary arable land on grain farms reached the price level of purchases by the National Board of Agriculture only during the period 1982—1986, and when the capitalization rate was 3 % and 5 % for a duration of 15 years. During the earlier periods investigated, the present valueof supplementary arable land on grain farms remained lower, even for the best land, than the prices paid by the National Board of Agriculture. It should be noted that many other factors, besides return on the cultivation of arable land, have an influence on the level of mar- ket prices paid for arable land. These other factors may be changes in Finnish agricultural policy and the financial means of its im- plementation. Because southern Finland has been chosen as the research region, the regional support scheme for agriculture is only weakly effec- tive in the areas investigated. On the other hand, different voluntary agreements, as well as legislation, concerning changes in produc- tion aimed at decreasing livestock production, have somewhat restrained agricultural prac- tices. The dual price system for milk, im- plemented in thebeginning of 1985, has been effective in controlling milk production, al- though its impact can be observed only dur- ing the two last years researched. Thus, the conditions of neoclassical theory, requiring perfect competition are not fully realized in dairy farming. On grain farms these conditions are better realized, because production has not been so affected by agri- cultural policy means. Nonetheless, even on grain farms, the cultivation of special crops has long been based on contracts. As for financial factors, real interest will be considered in more detail. Real interest rate was 4.5 % negative during the period 1975— 1977, while for the period 1979—1981 real in- terest was only slightly negative, turning clear- ly positive (3.2%) during the period 1982—1986. The turning point was 1982 and a continuous growth of real interest followed during the subsequent years. By 1986, a real interest of 5.4 % had to be paid on loans granted by cooperative and savings banks (ap- pendix 8, figure 8.1.). Moreover, the price development of cash crops appeared less favourable than for ani- mal products during the 1970’s (appendix 8, table 8.1.). With crop yields also remaining low and quality poor, the 1970’s saw a decrease in profitability, and consequently, in capital- ized value. Therefore, in both farm groups ex- amined, but especially grain farms, there may have been some speculative demand for sup- plementary arable land, despite the low level of capitalized value. It was possible to pay a higher price for arable land than its capital- ized value when the project was financed with 229 external capital and the real interest for the loan was negative. Further, the expectations on return from agriculture, and the expected increase in the value of land near population centers, could have encouraged higher prices for supplementary arable land than that as- sessed on the basis of their productive value. Since 1984, the state has tried to stabilize the increase in land prices by changing the financial practices concerning land acquisi- tions by not issuing loans for land acquisition, if the market price exceeded the estimated cur- rent price. It is probable that landowners reacted to this “financing ceiling”. It can therefore be supposed that in at least some of the arable land transactions not involving state loans, higher prices were knowingly paid than in transactions which qualified for financing, even though it meant losing the exemption from stamp duty. This supposition is sup- ported by the fact that prices in transactions recorded in the market price register of the National Board of Land Surveying exceeded the prices of farm-loan financed transactions during the years investigated. 7.4. Conclusions The capitalized valueof supplementary ara- ble land depends not only on the amount of return, but also on its duration and the interest rate used for capitalization. If the duration of return is assumed to be 100 years, capitalized value on cattle and grain farms during the period 1982—1986 would be as follows: Capitalization rate % Cattle farms Grain farms 66 832 FIM/ha 52 865 FIM/ha 41 979 FIM/ha 33 206 FIM/ha 30 179 FIM/ha 23 872 FIM/ha. 3 5 7 As may be observed from this, the capital- ized valueof supplementary arable land would reach the market value (c. 30 000 FIM/ha) recorded by the National Board of Land Sur- veying, except at the 7 % capitalization rate on grain farms. Such a long period of time is not realistic for an individual farmer, even with the hypothesis of constant return and in- terest. Also, generation transfers in farming occur approximately every 30 years, and this led to the choice of shorter durations of re- turn for the assessment of the present values of supplementary arable land, as was ex- plained in more detail in section 4.6.2. Problems that impede the determination of the capitalized value of supplementary arable land, and the use of the market prices paid for them, are related to the representiveness and properties of the statistical data. The em- pirical data in this study are based on infor- mation collected from different sources. Each statistical source has its special characteristics, as discussed earlier. Common to these sources are problems created when generalizing the results assessed with them. For example, the capitalized values of supplementary arable land assessed for the bookkeeping farms of southern Finland might be somewhat higher than the values of all other farms in the same region. This hypothesis is derived from Suo- mela (1958, p. 81—82), who estimated that gross return, production cost, and agricultural surplus on bookkeeping farms during the 1950’s were approximately 20 % higher than on other farms of more than two hectares of arable land in the country as a whole. Conse- quently, the capitalized values of supplemen- tary arable land on cattle and grain farms as determined in this investigation may be con- sidered to be about one fifth too high, and therefore primarily indicative. The representativeness of the data on mar- ket prices partly involves the same problems encountered with the use of bookkeeping data, namely, the majority of annual trans- actions of agricultural land concerns the trans- fer of whole farms from parents to children as part of the normal generation transfer process. According to theFarm Register, pur- chases of arable land destined for supplemen- tary arable land in southern Finland amounted to a little over 7 300 ha/year during the pe- riod 1982—1986. The market price register of the National Board of Land Surveying, the source of prices in this study, consisted of en- 230 tries of only 1 233 ha per year, on average, for the same period. Conversely, in the statis- tics on purchases of arable land by the Na- tional Board of Agriculture, the state only ac- quired c. 500 ha/annum during the whole period of investigation, 1972—1986. These statistics are nonetheless considered to best de- scribe the price of arable land in the research region. The National Board of Land Surveying’s information on supplementary arable land prices represents the highest and that of the National Board of Agriculture the lowest level of prices paid for arable land. Previous studies concerning land value and price (chapter 3) focused on the determination of either capitalized value or market value, at the exclusion of each other. The investigation in hand has sought to determine both the capitalized value of supplementary arable land and its market value, thereby enabling mutual comparisons. The results have shown that the average capitalized value on cattle farms reached, on the average, the price level of pur- chases of arable land by the National Board of Agriculture, given the reservation that the results of bookkeeping farms were still about a fifth higher than in other farms in southern Finland. On the other hand, the average capitalized value on grain farms remained, af- ter the adjustment in level, clearly lower than the market prices paid by the National Board of Agriculture. The prices in accordance with the market price register of the National Board of Land Surveying therefore exceed even the highest capitalized value on cattle farms by 10000 FIM during 1982—1986, while the corresponding difference with re- spect to grain farms reached 14000 FIM. Capitalized value and market prices are sub- ject to variations, both regionally and between different types of farms. Consequently, the results presented in this study are not valid in other regions or for other types of farms. It must also be emphasized that the results of the study indicate the average capitalized value and the development and level of market values in southern Finland only in an indica- tive way. Deviations concerning individual farms may be considerable. On the other hand, the state’s agricultural policy and social actions may also have an im- pact on the determination of land value. The economic conditions prevailing in society at any given time are reflected or capitalized in land value. Decreasing prices of agricultural products or diminishing agricultural support both decrease land prices, while probably a freer land market would have the opposite ef- fect. 231 8. SUMMARY The aim of the study was to determine the capitalized and market values of supplemen- tary arable land in the bookkeeping regions of southern Finland during theperiod 1972— 1986. The starting point was an examination of appraisal methods and the assessment of their applicability to the task in hand. The value of arable land was considered to be determined exclusively by the return from its cultivation. Also, the prices paid for arable land were con- sidered to describe the valueof arable land ac- quired for cultivation purposes only, without taking into consideration questions of expec- tation value. First, central concepts of value and price used in appraisals of arable land were exam- ined, as well as their applicability for differ- ent purposes. According to the terminology accepted by the Association of Finnish Real Estate Valuers, present value means “current value determined as the capital value of return assessed on the basis of present use or poten- tial use”. Correspondingly, market value means current value or current price deter- mined on the basis of market prices. Sup- plementary arable land means arable land al- ready acquired or to be acquired, which is used or to be used exclusively for agricultural production. These definitions were applied in this study. In chapter 2, the central factors influencing the value and price of supplementary arable land were examined. First, the effect of the quality of arable land, as well as that of de- mand and supply, were examined. It was ob- served that the effect of measures related to cultivation techniques and economic measures affecting the economic result in agriculture received emphasis and, thus, had an impact on land value. Conversely, the significance of land quality has weakened over time. The strong demand for supplementary ara- ble land, in comparison to its weak supply, increased the price level of arable land, espe- cially when related to technological develop- ments in agriculture and the increased use of machines, which have enabled the manage- ment of large-sized farms. The gap between the prevailing size of farm and the size of farm employing modern technology proved to be considerable. Previous investigations indicated that fac- tors affecting agricultural production and its profitability are also reflected in land values. Consequently, the distinct effect on the price of supplementary arable land of each of the contributing factors is difficult to define, even though they could be divided into general, quality, and individual factors. The third chapter examined former studies dealing with present value and market value, as well as the procedures applied. First, the fundamentals of the appraisal of capitalized value and its dependence on different factors were examined, following which, an approach based on land rent, which has had application in both economics and agricultural economics, was presented. The most common method for appraising the value of supplementary arable land has been to assess the return of supplementary arable land as a residual when the costs of other factors and production inputs have been subtracted from the gross return. In this ap- proach, a central problem is the differences in the use of inputs and the difficulties related to their pricing. This method results in low or 232 even a negative return for land for small farms. A third method of appraising the return of supplementary arable land employs an ap- proach based on the theory of marginal productivity, in which production inputs re- ceive compensation equal to their marginal value product. The second part of chapter 3 (section 3.2.), examined investigations which have dealt with market value. Problems concerning the avail- ability and validity of data on market prices were also discussed. Several price investiga- tions on supplementary arable land have been made in Finland, but long time-series of prices are available only in national statistics concerning land acquisitions and from the price statistics of notaries. Consequently, former Finnish price investigations usually focused on the examination of market prices of a small area or locality during specific years. The empirical part of this investigation con- cerned southern Finland, where the natural conditions for agriculture (soil and climate) are the most favorable. For the determination of the capitalized value of supplementary ara- ble land, the records of bookkeeping farms were used as a starting point. These farms have continuously participated in the profit- ability survey in agriculture since 1968. The data finally employed consisted of only those cattle and grain farms which were based prin- cipally on the use of arable land. Before estimation, the taxable values of the depreciable property items (machines, build- ings, and land improvements) on the book- keeping farms were adjusted by the cost-of- living index to the price level of 1986. The choice of variables was mainly based on the use of correlation analyses and experiments. First, a linear, Cobb-Douglas function, and a transcendental function were tested. The Cobb-Douglas function was chosen for the as- sessment of the marginal value product of sup- plementary arable land. Estimations were made by using both cross section and time se- ries analyses from the years 1972—1986. The marginal value product on cattle farms was determined for 1972, 1976, 1980, and the pe- riod 1982—1986. Due to the annual variations in the yields on grain farms, estimations were made on the basis of three- to five-year data: 1975—1977, 1979—1981, and 1982—1986. Only farms that had maintained the same pro- duction structure throughout theperiod were included in the data. The average marginal value product of supplementary arable land was estimated most accurately for the period 1982—1986 for both farm groups. Yield level dummies were used in regression models as in- dicators of the quality of arable land (table 16). Next, such variable costs that were not in- cluded in the explaining variables were sub- tracted from the marginal value product of arable land on both cattle and grain farms. The marginal operating margin so assessed in- dicated the pre-tax compensation, in Finnish marks per hectare, available for arable land acquisition. For the determination of the taxable part of the additional income consequent upon the acquisition of additional land, a marginal tax rate of 40 °Jo was chosen for cattle farms and 50 % for grain farms. By multiplying the marginal operating margin by a coefficient (1-marginal tax rate), the net marginal oper- ating margin per hectare of arable land could be derived for both farming types (table 18). For the capitalization of the capitalized valueof supplementary arable land 3 %, 5 %, and 7 % interest rates were used with cor- responding durations of return of 5, 10, and 15 years. The capitalized values assessed in this way increased continuously with time on cat- tle farms. Conversely, the capitalized value re- mained very low on grain farms, except dur- ing the last part of the research period (table 23). When the dependency between the marginal value product of arable land and the area of arable land was examined, farms of both types were classified into two size classes. The results indicated that the marginal value product of arable land on cattle farms of less 233 than 25 hectares was approximately three times greater than for the larger farm group. On grain farms, the marginal value product of arable land continued to increase with in- creasing farm size. In addition, the evaluation of differential return, based on the average taxable net return, similarly showed that the return gained from supplementary arable land was highest in the smallest farm-size class in the case of cattle farms. On the other hand, on grain farms, the marginal value product of arable land and the differential return re- mained high as the size of thefarm increased. The market prices paid for supplementary arable land were collected from the statistics on land acquisitions of the National Board of Agriculture from 1972—1986 and from the market price register of the National Board of Land Surveying. The latter statistics only covered the period 1982—1986, during which the number of transactions amounted to ap- proximately 1 233 ha a year, while the acqui- sitions of the National Board of Agriculture amounted to only 505 ha per year. As a re- sult, the annual purchases of arable land represented a little over one thousandth of the total 1.1 million hectares of arable land in the research region. The average price, weighted by the area of supplementary arable land in the transactions of the National Board of Agriculture, in- creased considerably until 1985, but decreased during the last year investigated, (prices de- flated by the cost-of-living index to 1986 values): Year FIM/ha 1982 25 240 27 692 31 641 33 004 31 920. 1983 1984 1985 1986 The average price per hectare of supplemen- tary arable land in the study region area was 29 899 FIM during the investigated years. In 1986, the highest prices for arable land were paid in the Agricultural Advisory Centres of Satakunta and Varsinais-Suomi, in which the price of arable land exceeded 43 000 FIM per hectare. During other years, the prices of ara- ble land in these areas were also the highest. The next highest price for arable land during the period 1982—1986 was observed in the area of the Agricultural Advisory Center of Häme, the range being 23 000—35 300 FIM/ ha. The lowest price for arable land was paid in the areas of the Agricultural Advisory Centers of Etelä-Karjala, Pirkanmaa, Itä- Häme, and Kymenlaakso. The real price of arable land varied from 15 700 to 28 300 FIM/ha during the years in question. The increase in the real prices of the pur- chases of the National Board of Agriculture were very moderate during 1972—1986. The average value of the annual changes was as- sessed to be 1.7 °7o. The quantity of subsur- face drained and open drainedarable land had a strong effect on the price of arable land ac- quired by the National Board of Agriculture. During the period 1978—1986, subsurface drained arable land cost 37 °70—60 % more than open drained arable land. In the pur- chases of the National Board of Agriculture, the price level was about 60 % of that re- corded by the National Board of Land Sur- veying. Because the results of bookkeeping farms are about a fifth higher than for other farms of the region, a corresponding reduction in level was made in the capitalized values of sup- plementary arable land. When the reduction in level was taken into consideration, the com- parison between capitalized value and market prices showed that the capitalized value of supplementary arable land on cattle farms reached the level of prices paid by the National Board of Agriculture during all investigated years when a duration of return of 15 years and a capitalization rate of 3 % (compare fig- ure 15) were employed. The capitalized value assessed on a duration of ten years failed to attain the price level paid by the National Board of Agriculture. In addition, the level of market prices recorded by the National Board of Land Surveying exceeded, by about 234 10000 FIM, the highest capitalized value of cattle farms during the period 1982—1986. For grain farms, the capitalized value of supplementary arable land did not reach the level of market prices paid by the National Board of Agriculture during the investigated period as a whole (compare figure 16). The capitalized valueof supplementary arable land on grain farms was clearly at its highest dur- ing 1982—1986, when, despite the reduction, it almost reached the level of market prices paid by the National Board of Agriculture with a duration of return of 15 years and a capitalization rate of 3 %. The highest capital- ized value of supplementary arable land on grain farms was approximately 16000 FIM/ ha and the market prices recorded by the Na- tional Board of Land Surveying almost 30 000 FIM. Market prices were thus almost twice as high. The capitalized value of supplementary arable land on grain farms proved to be more sensitive to variations in yield level than on cattle farms. In the assessment of the results of this in- vestigation, it should be noted that book- keeping farms represent farms that are larger than average, in addition to which they are managed in a rational and intensive way. For this reason, caution is required when gener- alizing the results. The same argument applies to market prices. Consequently, the results of the investigation only indicate the average capitalized value and market prices of sup- plementary arable land in southern Finland and there n ay be considerable deviations at the individual farm level. 235 REFERENCES Aalstad, S., Bjorä, E., Holm, A., Hustad, T., Kaurin, T., Stokke, K. & Gronn, A. 1979. Bruksverdi i land- bruket. 134 p. Oslo. Aalto, W. 1951. Tilusten jyvitys maanjakotoimituksis- sa. Maanmittaus 26: 71—139. Abel, W. 1958. Agrarpolitik. 465 p. 2. Aufl. Göttingen. Aereboe, F. 1919. Die Beurteilung von Landgiitern und Grundstiicken. 532 p. 52 Tafeln. 2. Aufl. Berlin. 1923. Allgemeine landwirtschaftliche Betriebslehre. 691 p. Berlin. Ahonen, L. 1970. Diskonttausarvo metsän hinnoitusin- formaationa. Acta Forest. Fenn. 105: I—Bl. Ala-Kantti, E. 1981. Lisäpellon hinta ja siihen vaikut- tavia tekijöitä eräissä Varsinais-Suomen kunnissa vuo- sina 1975—1978. Hels. yliop. maat.ekon. lait. Pro gradu-työ. 83 p.+ 11 liit. Andersson, Y. 1989. Markpriser i jordbruket. En teore- tisk studie belyst med svenska data. Inst, för ekon. och stat. Rapport 15: 1—97. Anon. 1967. Suomen Asetusk. 1967; 543. 1972—1986.Maatalouden taloudellisen tutkimuslai- toksen laskemat tuottaja- ja tarvikehintaindeksit. Maat. tilasi, kuuk.kats. Maatilahan, julk. vv. 1973— 1987. 1974—1989.Maat. tal. tutk.lait. julk. 31, 32, 34, 36, 41, 42, 44, 46 ja 59. Tilivuodet 1972—1987. 1977. Maatilalaki. Annettu 18.2. 1977/188. 1978.Laki oikeudesta hankkia maa- ja metsätalous- maata 26.5.1978/391. 1980 a). Peltojen inventointitutkimus. Maatilahan, julk. 3: 1—26, 2 Uit. 1980b). Sara 2000. Salaojitusohjelma 1980—2000. Salaojakeskus 48 p. Helsinki. 1980 c). Maatilarekisteri. SVT XLIILI. Maatilahal- litus. Helsinki. 1982.Ajankohtaista maatalousekonomiaa. Eri tuo- tantosuuntaa harjoittavienkirjanpitotilojen tuloksia. Tilivuosi 1980. Maat. tai. tutk.lait. tied.ant. 88: 1—39. 1986 a) Kiinteistöarviointisanasto. Suomen kiinteistö- arviointiyhdistys r.y. 54 p. Helsinki. 1986 b). Tutkimuksia Suomen maatalouden kannat- tavuudesta. Tilivuodet 1982—1984. Maat. tai. tutk. lait. julk. 51: 1 135. 1987. Suomen Tilasi. Vuosik. 557 p. Helsinki. 1989. Maatilalain lainoitusohje. Maatilahallituksen ra- hoitustoimisten 1.1.1989 laatima ohje. Barlowe, R, 1958. Land resource economics. 585 p. New Jersey. Barnard, L. & Nix, J. 1973. Farm planning and control. 549 p. Cambridge. Björk, B-C. 1984. Kansantaloudellinen laskentakorko- kannan empiirinen määrittäminen. Liiketal. Aikak. 4: 446—472. Blohm, G. 1966. Die Neuorientierung der Landwirt- schaft. 143 p. Stuttgart. Brandes, W. & Woermann, E. 1969. Landwirtschaftliche Betriebslehre. Allgemeiner Teil. 234 p. Hamburgund Berlin. Brown, M. 1968. On the theory and measurement of technological change. XII+ 214 p. Cambridge. Busch, N. 1969. Taxationslehre fiir Landwirtschaft und Gartenbau. 407 p. Hamburg und Berlin. Carey, H. C. (ref. Abel, W. 1958). Cassel, G. 1938. Teoretisk sosialekonomi. 692 p. Stock- holm. Castle, E. N. & Hoch, J. 1982. Farm real estate price components 1920—1978. Amer. J. Agr. Econ. 2: B—lB. Clark, C. 1973. The value of agricultural land. 117 p. New York. Clark, J. B. 1923. The distribution of wealth. A theory of wages, interest and profits. XXVIII + 445 p. New York. Crowley, W. 1974 a. Farmland use value versus market prices in the three Oregon land markets. U.S. Dept. Agric. 550; I—l9. 1974 b. Actual versus apparent rates of return on farmland investment. Agric. Finance Rev. 35, 10: 52—57. Doll, J. P. 1974. On exact multicollinearity and estima- tion of the Cobb-Douglas production function. Amer. J. Agr. Econ. 56: 556—563. Doll, J., Widdows, R. & Velde, P. 1983. The value of agricultural land in the United States. A report on re- search. Agr. Econ. Res. 2: 39—44. Elstrand, E. 1962. Okonomisk klassifisering av jord. Et metodestudie. Norg. Landbr.okon. Inst. Stensilert sasrmeld. 8: 1—77. 1980. Verdsetting av landbrukseiendommer og jor- darealer. Norg. Landbr.okon. Inst. 156 p. Oslo. & Sonju, K. 1978. Bruksverdier og kostnadsverdier for landbrukseiendommer. Prinsipper for verdsetting 236 og eksempler pä verdier for hele eiendommer. Norg. Landbr.okon. Inst. 129 p. Oslo. Feuerstein, H. 1970. Bestimmungriinde der Preise und des Transfers land-und forstwirtschaftlich genuzten Bodens, Eine ökonometrische Analyse des schleswig- holsteinischen Bodenmarktes von 1954 bis 1968. Diss. 184 p. Kiel. Found, W. 1971. A theoretical approach to rural land- use patterns. 189 p. London. Gabr, M. 1972. Modelle zur Charakterisierung der Gren- zertragsböden bei landwirtschaftlicher Nutzung. Diss. 165 p. Giessen. Goltz, T. von der 1905. Handbuch der landwirtschaft- lichen Betriebslehre. 706 p. 3. Aufl. Berlin. Conner, E. C. K. 1891. Kansantalouden ja verotuksen periaatteet. Suom. B. Tuunanen v. 1937. 451 p. Por- voo. Griesbach, H. 1966. Der Einfluss des technischen Fort- schritts auf die Preise landwirtschaftlich genutzten Bodens in Industrieländern. Volksw. Schr. 104: I 126. Gryst, W. & Timmons, J. 1961. The economic role of land resource institutions in agricultural adjustment. In lowa State University ofScience and Technology. Center for Agricultural and Economic Adjustment. Dynamics of land use: needed adjustment, p. 252— 277. Ames, lowa. Gulbrandsen, O. & Lindbeck, A. 1969. Jordbruksnä- ringens ekonomi. 284 p. Stockholm. Gustafsson, C., Lind. h.. Lundström, S. & Persson, E. 1978. Värdering av landbruksfastigheter.Metoder och principiella frägor. Tekn. högsk. Medd, 5: 6. 221 p. Stockholm. Hannila, P. 1980. Lisämaaksi ostetun pellon hinta Sei- näjoen maatalouspiirin alueella. Hels. yliop. maat. ekon. lait. Pro gradu-työ. 74 p. + 12 liit. Hayami, Y. & Ruttan, V. W. 1971. Agricultural de- velopment. An international perspective. Xiv + 367 p. Baltimore and London. Heady, E. O. 1949. Basic economic and welfare aspects of farm technological advance. J. Farm Econ. 31: 296—297. Heady, E. O. & Dillon, J. L. 1972. Agricultural produc- tion functions. 667 p. Ames, lowa. Heertje, A. 1977. Economics & technical change. Vll + 334 p. London. Heikkilä, A-M. 1987. Lypsykarjayritysten optimaalinen koko. Maat. tai. tutk.lait. tied.ant. 132: I—7o +liitt. Heikkinen, J. 1978. Lisämaan hinta eräissä Kainuun kunnissa vuosina 1973—75. Hels. yliop. maat.ekon. lait. Pro gradu-työ. 82 p. +2 liit. Heinonen, S. 1980. Peltoalan kasvun vaikutuksista maa- tilan talouteen. Maatilahan, julk. 1: I—l3B +25 liit. Heiskanen, O. 1977. Maa- ja metsätalousmaan hinnan- muodostuksesta Mikkelin läänissä 1.4.1975—31.8. 1976 tehdyissä kaupoissa. Maanmitt.hall. moniste. 67 p. 1983. Maa- ja metsätalousmaan jyvitysperusteisen ar- von määrittäminen. Maanmitt. hali. julk. 50; 1—153. 1987. Maa- ja metsätalousmaan jyvitysperusteisen ar- von määrittämisestä (jatkotutkimus). Maanmitt.hall. 116 p. Helsinki. Hemilä, K. 1983. Measuring technological change in agriculture. An application based on the ces produc- tion function. Maatal.tiet. Aikak. 54: 165 —223. Herdt, R. W. & Cochrane, W. W. 1966. Farm land prices and farmtechnological advance. J. Farm. Econ. 48: 243—263. Hjelm, L. 1952. Jordbruksfastigheternasprisutveckling. Medd. frän Jordbr. Utredningsinst. 7: 1—32.I—32. 1963. Det svenska lantbrukets efektiviseringsvägar. Undersökning utförd pä uppdrag av 1960 ärs jord- bruksutredning av professor Lennart Hjelm i samräd med en expertgrupp inom utredningen. Stat. off. utredningar 66: 1—241. Holmsten, M. & Myhrberg, O. 1986. Pellon hinta Suo- messa vv. 1981 —85. Maanmitt.hall.julk. 55.43 p+ 13 liit. Hovi, L. & Jokinen, O. 1974. Peltoalueen ja talouskes- kuksen välisten kulku- jakuljetuskustannusten mää- rittelemisestä. Valt.tekn.tutk.kesk.Maank. labor. 9: 1—35. Huhtamäki, A. 1985. Sukupolvenvaihdos maatilalla. Maatal.yritt. eläkelait, julk. 4: 1—93 +liitteet. Hyvönen, V. 1982. Kiinteistöjärjestelmä ja kiinteistön- muodostamisoikeus. 767 p. Espoo. Ihamuotila, R. 1968. Viljelijöiden työtulojen taso kirjan- pitotiloilla 1956—1965. Maat. tai. tutk.lait. julk. 10: 1—172. 1971.Pääoman merkityksestä tuotannontekijänä Suo- men maataloudessa vuosina 1948—1970. Osuuspank- kijärjestön taloudellinen katsaus 2: I—B. 1983. Suomen maatalouselinkeinon pääomakanta ja velkaisuusaste vuosina 1961 —l9BO. Hels. yliop. maat.ekon.lait.julk. 10: I—l3B. & Stanton, B. F. 1970. A balance sheet of agricul- ture for Finland 1948—1967. Maat. tai. tutk.lait. julk. 20: 1—122. Ikäheimo, E. 1979. Taajamien lievealueiden maatalous. Maatal.hall. Aikak. 9,1: 27—32. Jaooi, E. 1945. Methodik und Technik der Ertragswert- schätzung landwirtschaftlicher Betriebe und Grund- stiicke. 138 p. Brugg. Johsson, R. W. M. 1971. Trends in rural land prices in New Zealand 19541969. Agric. econ. res. unit techn. pap. 4: 1—44. Johnston, W. E. & Bischoff, T. 1971. Zur Betriebsgrös- se auf Ackerbaustandorten unter Beriicksichtigung neuer technischer Fortschritte. Agrarwirtschaft 20: 117—125. Jorgensen, A. & Jorgensen, N. 1971. Nogle faktorers indvirkning pä ejendomspriserne i landbruget. Det landokon. Driftsbur. 70 p. Kobenhavn. Juvonen, E. 1983. Työn ja pääoman vaikutus maatalous- 237 tuloon. Maat. tai. tutk.lait. tied.ant. 101: I—6l. Kaarlehto, P. 1954. Maatalousmaan arvioimisesta Yh- dysvalloissa. Maatal.tiet. Aikak. 2: 61—78. Kanerva, V. 1974. Asunto- ja liiketonttien hintojen ja niihin vaikuttavien tekijöiden määrittäminen moni- muuttujamenetelmän avulla. VTT:n rak.tal. lab. tied.ant. 35: I—lo9.1 —109. 1978. Tontin hinta ja rakennuskustannukset. Valt. tekn. tutk.kesk. Rakennus-ja yhdyskuntatekniikka. Julk. 21: I—B4+ll1 —84+11 liites. 1980. Kiinteistön arviointi. 144 p. Helsinki. Kantola, J. 1967. Tutkimus asunto-ja liiketonttien hin- nanmuodostuksesta eräissä Uudenmaan läänin asu- tuskeskuksissa. Tekn. korkeak. maanmitt.os. Lis.työ. 185 p. 1977. Rakennusarvioinnin käsikirja. Maanmitt.hall, julk. 43: I—los. 1978. Ekonometrinen tutkimus käytännön arvioinnis- sa. Insinöörijärj. koulutuskesk. julk. 87—78, II: I—ls. 1979. Tutkimus maan hinnasta sekä rakennuksen omaisuusosa-arvon ja nykyarvon suhteesta Hämeen- linnan ympäristössä. Maanmitt.hali. julk. 46: 1—172. Kehrberg, E. & Reisch, E. 1969. Wirtschaftslehre der landwirtschaftlichen Produktion. 263 p. Miinchen. Kettunen, L. 1966. Om produktionfunktionens form. NJF:s saert. 48,1: I—l 9. & Rouhiainen, J. 1972. Aggregate livestock and total production functions in Finnnish agriculture in 1956/ 57—1969/70. Maat. tai. tutk. lait. julk. 28, 2; 1—54. & Torvela, M. 1969. Tuoton ja tuotantopanosten in- tensiteettitaso jakeskimääräinen riippuvuus maata- loudessa. Hels. yliop. kansantal.tiet. lait. tutk. 7: I—Bo. Korpela, A. 1978. Taloutemme rakenne ja toiminta. 210 p. Helsinki. Koutsoyiannis, A. 1985. Modern microeconomics. 581 p. London. Kukkonen, P. 1990. Suullinen tiedonanto. Kulmia, H. P. 1943. Mallitilusten jyvitys. Suom. maan- mitt.yhd. Aikak. 6—B: 160193. Kurki, M. 1982, Suomen peltojen viljavuudesta 3. 181 p. Helsinki. Larsson, G. 1954. Värdering av fastighetens jordbruks- del. Svensk lantmäteritidskr. 1: 56—76. Latukka, A. 1989. Poistonalaisten omaisuusosien ar- vostaminen ja sen vaikutus maatalouden liiketulok- siin vuosina 1968—1986. Hels. yliop. maatal.ekon. lait. Pro-gradu-työ. 120 p. Laur, E. 1928. Grundlagen und Methoden der Bever- tung, Buchhaltung und Kalkulation in der Landwirt- schaft. 857 p. 3. Aufl. Berlin. 1930. Einfiihrung in die Wirtschaftslehre des Land- baues. 346 p. 2. Aufl. Berlin. & Hovvald, O. 1957. Bewertung, Buchhaltung und Kalkulation in der Landwirtschaft. 243 p. 4. Aufl. Hamburg und Berlin. Laurila, I. 1986. Pellolla on monta hintatasoa. Maa- tal.hall. Aikak. 3: 15—23. 1988. Pellon markkinahinta. Johdatus teoreettiseen keskusteluun ja empiiriseen analyysiin. Maatilahan, julk. 15: 1—134. Leponiemi, A. & Lammi, K. 1968. Maatalouskiinteistö- jenkauppahinnat Suomessa 1961,1962 ja 1966. Kyösti Haatajan rah. tutk.tston julk. D. 1968: 1: 1—44. Mimeogr. & Lindberg, J. 1966. Tilastollinen selvitys maatalous- kiinteistöjen hinnoista v. 1961. Osuuskassajärjestön tai. kats. 1966, 4: 141—145. Locken, G., Bills, N. & Boiswert, N. 1978. Estimating agricultural use values in New York state. Land Eco- nomics 54, 1: 50—63. 1979. Estimating agricultural use values in New York state: Reply. Land Economics 55, 3: 408—410. Marshall, A. 1920. Principles of economics. An in- troductory volume. 871 p. Bth Ed. London. Matikainen, K. 1980. Maa- ja metsätalousmaa-alueiden arvioiminen. Ohje. Valt. tekn. tutk.kesk. Rakennus- ta!. laborat. Moniste. 8 p. + 3 liit. Meimbero, P. 1956. Landwirtschaftliches Rechnungs- wesen. 540 p. Stuttgart. Murray, W. G. 1969. Farm appraisal and valuation. 534 p. Ames, lowa. Myhrberg, O. 1969, Tutkimus omakoti-ja ranta-asutuk- sen kehityksestä sekä kieltojen ja rajoitusten vaikutuk- sesta siihen eräissä Uudenmaan läänin kunnissa. Tekn. korkeak. maanmitt. os. Lis.työ. 160 p. 1981. Raakamaan arvioinnin perusteet. Valt. tekn. tutk.kesk. tied. 31: 1—62, Myhrbero, O. & Väänänen, J. 1988. Pellon ja haja- asutusrakennuspaikkojen hintaindeksit. Maan.mitt. hali. julk. 63:1—51 p. +liitt. Mäkelä, E. 1977. Lisämaaksi ostetun pellon hinta eräis- sä Kymenlaakson kunnissa vuosina 1973—1975. Hels. yliop. maanvilj.tai. lait. Laudaturtyö. 75 p. Mäki, A. 1945. Viljelysmaan pakkolunastuksen vaiku- tuksesta luovuttavan viljelmän maatalouteen ja sen kannattavaisuuteen. Suom. Maatal.tiet. Seur. julk. 58: 1—59. 1963. Maatalouden korvauskysymykset tietoimituk- sissa. Maanmittausinsinööri 4. Erip. 12 p. Turku. 1964. Maatalouden tuotantovälineet. Maanvilj. tie- tok. 3: 89—159. Porvoo. Mäkinen, P. 1988. Suomen maatalouden rakennemuu- tos. Tutkimus rakennemuutokseen vaikuttaneista te- kijöistä ja Markovin ketjujen käyttömahdollisuuksista rakennekehityksen ennustamisessa. Hels. yliop. maat, ekon. lait. Lis. työ. 178 p. Neukomm, W. 1947. Untersuchung iiber die Anwendbar- keit der Rohertragsmethode. 87 p. Bern. Niehans, J. 1966. Eine vernachlässigste Beziehung zwi- schen Bodenpreis, Wirtschaftswachstum und Kapital- zins. Schw. Z. fiir Volkw. und Statistik. 102: 195— 200. 238 4 Niitamo, O. 1969.Tuotantofunktio, sen jäännöstermi ja teknillinen kehitys. Tilasi, päätoim. monist. tutk. 9: 1—49. Nositschka, H. 1973. Ein Beitrag zur Erklärung der Preisbildung landwirtschaftlich genuzten Bodens. Diss. 211 p. Frankfurt. Ott, A. E. 1959. Technischer Fortschritt. Handwörter- buch der Sozialwissenschaften. p. 302—316. Stuttgart. Pekkarinen, J. & Sutela, P. 1981. Kansantaloustiede. Johdatus perusteisiin. 261 p. Porvoo. Penrose, E. 1959. The theory of the growth of the firm. 272 p. Oxford. Pessi, Y. 1964. Suomen ilmasto ja sääolot. Maanvilj. tie- tok. I: 19—34. Porvoo. Petrini, F. 1964. Competition between agriculture and forestry under Swedish conditions. Lantbr.högsk. Ann. 30: 1—333. Pihkala, K. U. 1965. Possibilities of comparing land use for farming and forestry. Acta Agric. Scand., Suppl. 14: 1—56. 1975. Economics of forest land clearing for agricul- ture. J. Sci. Agric. Soc. Finland. 47, 4: 283—368. & Lasola, T. 1973. Kainuulaisen uudistilan edullisin suuruus ja tuotantosuunta.Maatal.tiet. Aikak. 45, 4: 325—460. Plaxico, J. S. & Kletke, D. D. 1979. The value of un- realized farm land capital gains. Amer. J. Agr. Econ. 5: 327—330. Preuschen, G. 1968. Zweckmassig rationalisieren. 131 p. Hamburg. Rasmussen, S. 1981. Beregning af en landbrugsejendoms brugsvasrdi, Jordbr.okon. Inst. Rapp. 1: 1—34. Reichel, H. 1973. Überpriifung der Ergebnisse der Reichsbodenschätzungauf ihren ökonomischen Aus- sagewert unter heutigen Produktionsbedingungen. Diss. 101 p. Flohenheim. Renborg, U. & Karlsson, H. 1969. Företagets tillväxt. Akt. frän lantbr.högsk. 142: 1 —43. Renne, R. 1947. Land economics. Principles, problems and policies in utilizing land resources. 736 p. New York. Reynolds, J. E. 1966. An econometric investigation of farmland values in the United States. Diss. 185 p. lowa. & Timmons, F. 1969. Factors affecting farmland values in the United States. Agr. and home econ. exp. stat., lowa State Univ. Sci and Techn., Res. Bull. 566: 324—351. Ricardo, D. (ref. Conner, E.C.K. 1891). Rothkeoel, W. 1950. Geschichtliche Entwicklung der Bodenbonitieringund Wesen und Bedeutung der deut- schen Bodenschätzung. 147 p. Stuttgart. 1952. Landwirtschaftliche Schätzungslehre. 188 p. 2. Aufl. Stuttgart. Rouhiainen, J. 1972. Aggregate crop production func- tions in Finnish agriculture in 1956/57—1968/69. Maat, tai. tutk.lait. julk. 28:1 —71. Ryynänen, V. 1962. Maatalousmaan luovutus yleisiin tarpeisiin ja sen vaikutus viljelmän omaisuuden ar- voon sekä viljelmältä saatavaan tuloon eräissä Poh- jois-Karjalankunnissa. Hels. yliop. maan.vilj.tai. lait. Lis.työ. 188 p. 1967. Viljelmään kuuluvan maatalousmaan osittai- sen pakkolunastuksen aiheuttamien menetystenarvioi- minen. Suom. Maatal.tiet. Seur. Julk. 110: 1—165. 1970. Tutkimuksia maatalouden tuotantofunktiois- ta Sisä-Suomen kirjanpitoviljelmillä vuosina 1960— 1966. Acta Agr. Fenn. 120: 1—65. 1972. Teknisen kehityksen edellyttämän tilakoon ja vallitsevan tilakoon välinen kuilu. Virkaanastujais- esitelmä. Hels. yliop. maatal.ekon. lait. julk. 9: I—ll. 1978 a. Maatalouskiinteistöjenarvo ja arvioimisongel- mat. Maatalouspolitiikan saroilta. Samuli Suomelan 60-vuotisjuhlajulk. p. 77 —85. 1978 b. Arvioimistieteen perusteita ja menetelmiä. Hels. yliop. raaatal.ekon. lait. julk. 1: I—sB,1 —58, 11 Uit. 1981. Tavoitteena yritystoiminta ja elämänmuoto. Maatalouden kaksi roolia tekevät elinkeinosta vakaan. Lyh. Studia Agraria-sarjan luennosta. Erip. Maaseu- dun Tulevaisuus n:o 123. 1989. Laskentakorkokanta kiinteistöarvioinnin sovel- lutuksissa. Suom. kiint.arv.yhd. juhlajulk. 10 vuot- ta, p. 40—47. & Polkki, L. 1982. Maanviljelystalous. 265 p. Hel- sinki. Schuh, G. E. & Scharlach, W. C. 1966. Quantitative analysis of some farm and non-farm determinants of agricultural land values impact of economic de- velopment. Res. bull. 821: 1—24. Scofield, W. 1957. Prevailing land market forces. J. Farm Econ. Vol. 39, 5: 1500—1523. 1964. Land prices and farm earnings. Farm real es- tate market developments. U.S.D.A. Econ. Res. Serv. 66: 39—45. 1965 a. Land prices and farm earnings. Appraisal J. 33: 100—113. 1965 b. Land return and farm income. Farm real es- tate market developments. U.S.D.A. Econ. Res. Serv. 67: 44—52. Siebert, H. 1969. Einfiihrung in die Volkswirtschafts- lehre. Teil I: Preistheorie. 180 p. Stuttgart-Berlin- Köln-Mainz. Siirola, J. 1988. Ansiotulojen verotus maatalousvero- tuksessa. Hels.yliop.maatal.ekon. lait. Pro-gradu-työ. 68 p. Siltanen, L. & Ala-Mantila, O. 1989. Maatalouden kokonaislaskelmat 1980—1988. Maat. tai. tutk.lait. tied.ant. 151: 1—56. Skomroch, W. 1962. Bestimmungsgriinde und Berech- nung des Zinsanspruchs. Ber. iiber Landw. 40, 4: 813—844. Strobehn, R. 1966. Resource productivity and income distribution with implications for farm tenure adjust- ment. 111. Agr. Exp. Sta. Bull. 720: 1—35. 239 Suomela, S. 1958. Tuottavuuden kehityksestä Suomen maataloudessa. Maat. tai.tutk.lait. julk. 1: I—l2B Suter, R. 1976. The appraisal of farm real estate. 613 p. Illinois. Thaer, A. 1880. Grundsätze der rationellen Landwirt- schaft. 1100 p. Berlin. Thunen, G. H. von 1875. Isolierte Staat in Beziehung auf Landwirtschaft und Nationalekonomie. 876 p. 3. Aufl. Berlin. Tolonen, S. 1985. Lisämaiksi siirtyneiden peltojen käyt- tö. Maatal.yritt.eläkelait, julk. sa; 1—43. Torvela, M. 1966. Tuotantopanostenkäytöstä ja käy- tön edullisuudesta maataloudessa Etelä-Suomen kir- janpitoviljelmillä. Maat. tai. tutk.lait. julk. 8: I—l4l. 1980. Yrityskoko kotieläintaloudessa. Maatalous 73: 104—106. & Mäki, S. 1974. Perheviljelmän koko rationaali- sessa maataloustuotannossa Maat. tai. tutk.lait. julk. 30: 1—79. Turkki, A. 1982. Tuotantopanosten käytön vaikutus maidontuotannon kannattavuuteen. Hels. yliop. maat.ekon. lait. julk. 8: I—6B.1 —68. Tweeten, L. 1969. Theories explaining the persistance of low resource returns in a growing farm economy. Amer. J. Agr. Econ. 51: 798—81. Ulvinen, O. 1980. Siemenviljan ja öljykasvien siemen- tavaran tuotannon alueellinen jakaumavuoden 1979 viljelystarkastuksissa. Kylvösiemen 2: 27—29. Upton, M. 1976. Agricultural production economics and resource use. 357 p. Oxford University Press. Valkama, H. 1979.Pellon hinnan kehityksestä maamme eri alueilla 1960—1977 sekä lisämaan käyvästä hin- tatasosta v. 1978. Pellervon tai. tutk.lait. raportteja ja artikkeleita 1: I—3B, 6 liit. Vartiainen, H. 1963. Hinnanmuodostus. Yhteiskuntatie- teiden käsikirja 1:135—151. Helsinki. Virolainen, J. 1950. Maatalousmaan arvioimisesta ja ar- vosta Suomessa vuosina 1934—1938. Suom. Maatal. tiet. Seur. Julk. 72: 1—283. Virtanen, P. V. 1966. Maan arvoon vaikuttavat seikat kaupungissa. Suomen kaupunkiliitto. 98 p. Karhula. 1978. Kaupunkimaan arvonmuodostuksen teoriaa. Maanmittaus 3—4:14—48. 1979. Kaupunkimaan arvioinnin perusteet. 189 p. Espoo. Vuorinen, J. 1952. Peltojen luokitus ja arvosuhteet vil- javuustutkimuksen valossa. Maanmitt.ins. Liiton Aikak. 10—11:419—427. Weber, A. 1973. Arbeit- und Bodensparende Tech- nologien. Agrarwirtschaft 2:57—68. Weckman, K. J. 1970. Production allocation in Finnish agriculture. Suom. Maatal.tiet. Seur. Julk. 117: 1— 116 Wentrup, C. 1978. Bestimmungsgriinde fiir Bodenpreise. Bonner Hefte fiir Agrarpolitik und Agrarsoziologie. Heft 5. 101 p. Stuttgart. Westermarck, N. 1967. Ohjeita katetuotto-menetelmän mukaisen taloussuunnitelman laatimiseksi. Maatal. seur. Keskusl. Julk. 534: 1 —55. Wiiala, A. 1958. Tiluslaji- ja jyvitysoppi. Maanjaon ar- vioimisoppi I: 1 —lOB. 1960. Tiekorvaukset. Maanjaonarvioimisoppi 111. 202 p. Vammala. 1966. Pakkolunastus- ja tilikorvaukset. Maanjaon ar- vioimisoppi 11. 249 p. Vammala. 1976. Kiinteistöarvioinnin käsikirja. 275 p, Otaniemi. Willer, H. 1967. Technischer Fortschritt und Landwirt- schaft. 218 p. Hamburg und Berlin. Ylätalo, M. 1978. Maan tuottoarvo ja siihen vaikutta- vat tekijät eräillä Etelä-Suomen kirjanpitotiloilla vuo- sina 1968—1972. Hels. yliop. maanvilj.tal. lait. julk. 5: 1 107. 1987. Maatalouden investoinnit ja tuottavuus. Peller- von tai. tutk.lait. julk. 8:1—94. Änkö, O. 1968. Maan hintataso Helsingin seudulla v. 1967. Valt. tekn. tutk. lait. julk, 143: 1—75. 240 APPENDIX I Basis for the assessment of capital values The estimation of production functions on the basis of farm data is made particularly problematic by the revaluation of agricultural property values. Following the fiscal reform of 1968, agricultural accounting adopted the same property values as used in taxing the values of farm buildings, equipment, and land improvements. Accord- ingly, depreciations in accordance with the taxation of the property parts concerned were implemented in book- keeping activities. Because the profitability survey, as well as other book- keeping activities, does not take into consideration the effect of inflation, agricultural property values have, lit- tle by little, fallen behind their real market value. This is particularly the case regarding the values of machines, equipment, farm buildings, and subsurface drainage, on which the estimations of production function are based. It was therefore, considered important in this study to adjust the tax depreciablepropertyvalues in order to pre- vent the decrease in real value of properties caused by in- flation. In this case, the adjustment of property items, in ac- cordance with bookkeeping practice, is made separately for farm buildings, machines, and land improvements'. The value determined for the depreciable property items in agriculture, in connection with the fiscal reform of 1968, is used as a starting point. The adjustment is made by multiplying the non-depreciated investment of each farm and property item at the beginning of 1968, by a coefficient of correction that corresponds to the percen- tile rise of the cost-of-living index or inflation in 1968, After that, the revised depreciation is subtracted from the revaluated assets, and, thus, the non-depreciated invest- ment at the end of the year is obtained. Investments and sales (mainly machines) are adjusted in the following way: investments and sales in accordance with agriculture taxation are assumed to have occurred in the middle of the fiscal year and, thus, their difference is multiplied by a coefficient of correction that cor- responds to six month’s inflation. After this, the depre- ciation on six months is subtracted from the revaluated 1 In this respect, the possibility to use the investment accumulation approach, in accordance with national ac- counting, was examined. It required gross investment ac- cumulation series for the properties in question cor- responding to their duration. The bookkeeping data used in this study does not contain information on investment accumulation for far enough back in time, especially con- cerning buildings and subsurface drainage. investments, giving the value of investments at the end of the year. The sum of depreciationson investments and depreciations on non-depreciated investment are the revaluated real depreciations of 1968.Correspondingly, the sum of investments at the end of the year and non- depreciated investments are non-depreciated investments in the beginning of 1969 that are to be adjusted. In this way, adjustment is continued until 1986. Agricultural investments are considered in accordance with taxation practices, i.e. they include costs of invest- ments, the value of own timber, and the value of exter- nal labour used for investments. Sales consist of sales in- come, compensation for damage, insurance and others, as well as grants. The revaluated procedure is shown as follows: At +1 =[A t (l + n,) —A,(I +Pi)p0] +[I,(1 +P|/2) 1,(1+p,/2)(p0/2)]- At+l =At(l+p,)(l-Po) + I t (l + p,/2)(l-p„/2), in which A, =value at the beginning of the year A,+ 1 =value at the beginning of the following year I, =net investment (total investment-sales) p, = percentage signifying increase in cost level each year p 0 = percentage of depreciations. For the determination ofcosts due to equipment, farm buildings and, land improvements, depreciations are as- sessed from the revaluated capital values as equal per- centages2 . The depreciationapplied was 19 % for machines, 8 % for buildings, and 6 % for land improvements. It is to be noted that the size of the adjusted depreciations de- pends on the rate of increase of the index used for ad- justing the capital values for inflation, as well as on the percentage of depreciation. The cost-of-living index was chosen for the adjustment because the special indexes (the index of equipment costs and the index of building costs) describe the price development ofnew property items and therefore include a certain price rise due to technological development, independent of inflation. The adjusted capital values and the depreciations esti- mated from them are used in this investigation for the assessment of costs derived from the cost of equipment, farm buildings, and land improvements. 2 The problems of capital value adjustment and depreciation have been discussed in more detail by Ylätalo (1987) and Latukka (1989); papers which form the basis of the method applied in the present investiga- tion. 241 APPENDIX 2 Estimates of parameters of the Cobb-Douglas function on cattle farms for different years, at 1986 prices. Table 2.1. Estimates of parameters in 1972. Production input Parameter Estimate t-value Constant Arable land area Cost of purchased feed Cost of purchased fertilizers Livestock cost Cost of equipment Agricultural works Number of farms =39 R 2 = 0.927 5.18966 0.249253 0.066866 0.080809 0.098730 0.055866 0.198038 0.110874 0.187752 6.9B*** 3.36** 3.37** 3.74*** 3.03** 1.23 3.73*** 2.25* 2.50** log A B, B,’ B,” B: B 3 B 4 B, B 6 Table 2.2. Estimates of parameters in 1976. Production input Parameter Estimate t-value Constant Arable land area Cost of purchased feed Cost of purchased fertilizer Livestock cost Cost of equipment Agricultural works Number of farms =44 R 2 =0.938 5.093482 0.240730 -0.013142 0.040965 0.175740 0.157887 0.161474 0.047505 0.157545 4.B9*** 3.07»* -0.78 2.12* 6.o3*** 2,91** 3.33** 0.84 1.31 log A B, B,’ B,” B 2 B, B 4 B, B 6 Table 2.3. Estimates of parameters in 1980. Production input Parameter Estimate t-value Constant Arable land area Cost of purchased feed Cost of purchased fertilizer Livestock cost Cost of equipment Agricultural works Number of farms=51 R 2 =0.954 4.704700 0.292734 0.019995 0.040190 0.239892 0.122267 0.084304 0.153962 0.078493 6.B7*** s.ol*** 1.43 2.61** B.29*** 3.15** 2.43*** 2.68»** 1.04 log A B, B,’ B,” B, B, B 4 B, B 6 242 Table 2.4. Estimates of parameters for the period 1982—1986. Production input Parameter Estimate t-value Constant Arable land area Cost of purchased feed Cost of purchased fertilizer Livestock cost Cost of equipment Agricultural works Number of farms = 155 R 2 = 0.896 5.259936 0.323629 0.011470 0.023014 0.199133 0.151382 0.102256 0.070035 0.114403 11.66*** 6.89»*» 1.25 2.36** 9.27*** 4.32*** 3.11*» 1.51 2.01* log A B. B,’ B,” B, B 3 B 4 B, B„ 243 APPENDIX 3 Estimates of parameters of the Cobb-Douglas production on grain farms for different periods, at 1986 prices. Table 3.1. Estimates of parameters for the period 1975—1977. Production input Parameter Estimate t-value Constant Arable land area Cost of purchased seeds Cost of equipment Agricultural works Number of farms =99 R 2 = 0.922 5.354686 0.303332 0.231422 0.328492 0.004665 0.288103 0.230291 7.79**» 3.28** 5.80**» 7.93*** 0.87 3.4l*** 4.92*** log A B, B.’ B,” B 2 b 3 b 4 Table 3.2. Estimates of parameters for the period 1979—1981, Production input Parameter Estimate t-value Constant log A B, B,’ B,” B 2 B, b 4 6.693207 0.589511 0.115958 0.194612 0.010625 0.184585 0.113499 B.79*** 6.52*** s.Bo*** 6.6B*** 1.43 Arable land area Cost of purchased seeds Cost of equipment 2.01* 1.87Agricultural works Number of farms = 102 R 2 =0.895 Table 3.3. Estimates of parameters for the period 1982—1986. Production input Parameter Estimate t-value Constant log A B, B,’ B,” B 2 b 3 b 4 7.855445 0.585661 0.119776 0.204096 0.011301 0.120806 0.073688 24.14*** 11.88*** 6.1 1»** 10.16**» 2.52* Arable land area Cost of purchased seeds Cost of equipment 3.05** 2.66**Agricultural works Number of farms = 195 R 2 = 0.936 244 APPENDIX 4 Table 4.1. Correlation matrix of cattle farm variables for the period 1982 —1986, at 1986 prices. Q X, X2X 3x4X 5Xj X, X8X, X| 0 X( j X| 2 X,3 X| 4 xl5 Q 1.0 0.383 0.846 0.768 0.115 0.035 0.747 0.717 0.374 0.522 0.651 0.208 0.040 0.563 0.196 0.644 X, 1.0 0.809 0.529 0.148 0.101 0.719 0.566 0.355 0.475 0.514 0.150 0.056 0.502 0.139 0.594 X21.0 0.582 0.158 0.089 0.713 0.634 0.317 0.459 0.609 0.213 0.127 0.489 0.198 0.637 X 31.0 0.175 0.042 0.505 0.586 0.228 0.327 0.457 0.194 -0.002 0.363 0.199 0.375 X41.0 0.042 0.084 0.129 0.029 0.081 0.076 -0.048 0.015 0.085 -0.040 0.046 X 51.0 0.015 0.175 -0.137 -0.089 0.110 -0.085 0.109 -0.086 -0.066 -0.078 X 61.0 0.609 0.459 0.603 0.497 0.321 0.109 0.613 0.312 0.609 X7 1.0 0.277 0.447 0.501 0.189 0.093 0.480 0.205 0.533 X 81.0 0.933 0.190 0.130 -0.021 0.874 0.133 0.350 X 91.0 0.309 0.168 0.029 0.974 0.168 0.496 X lO 1.0 0.118 0.124 0.336 0.105 0.498 X„ 1.0 0.120 0.177 0.954 0.129 X l2 1.0 0.031 0.223 0.096 X l3 1.0 0.177 0.492 X l4 1.0 0.144 X 1 5 10 Q = gross return X6=cost of equipment X l2 = investment in buildings X, = arable land area X7=livestock cost X l3 = agricultural works X2=purchased fertilizers X8=imputed wage of farmer and family X l4 = investment in mach./build. X 3 = purchased feed X9=labour cost X l5 = other costs X4=purchased seeds X, 0 = cost of buildings X 5=plant protection, grain drying X n = investment in equipment 245 Table 4.2. Correlation matrix of grain farm variables, for the period 1982 —1986, at 1986 prices. Q X, X2X 3X 4X, X* X7 Xg X9X| 0 X„ Xl2 Xj3 Xl4 Xjs Q 1.0 0.930 0.868 0.322 0.232 0.481 0.867 0.261 0.528 0.808 0.554 0.471 -0.006 0.793 0.471 0.622 X, 1.0 0.889 0.251 0.195 0.560 0.881 0.188 0.535 0.826 0.549 0.459 0.034 0.804 0.471 0.606 X21.0 0.271 0.195 0.520 0.794 0.223 0.497 0.751 0.480 0.399 -0.001 0.736 0.407 0.548 X 31.0 -0.059 0.147 0.182 0.695 0.433 0.463 0.080 0.090 - 0.217 0.476 0.058 0.012 X41.0 0.175 0.188 -0.113 -0.036 0.068 0.081 -0.028 0.133 0.055 0.001 0.193 X 51.0 0.584 0.144 0.424 0.538 0.320 0.320 0.051 0.533 0.321 0.252 X 61.0 0.117 0.508 0.746 0.492 0.543 0.127 0.725 0.557 0.569 X7 1.0 0.446 0.418 0.056 0.099 - 0.109 0.432 0.074 0.013 X g 1.0 0.782 0.112 0.370 -0.019 0.818 0.374 -0.092 X 91.0 0.477 0.487 -0.033 0.989 0.486 0.425 X, 0 1.0 0.190 0.073 0.458 0.219 0.616 X„ 1.0 0.067 0.479 0.937 0.248 X i: 1.0 - 0.042 0.221 0.059 X l3 1.0 0.480 0.379 X l4 1.0 0.250 Xl5 10 Q = grossreturn X6=cost of equipment X l2 = investment in buildings X, = arable land area X7=livestock cost X l3 = agricultural works X2=purchased fertilizers X8=imputed wage of farmer and family X l4 = investment in mach./build. X 3 = purchased feed X, = labour cost X l5 = other costs X4=purchased seeds X lO = cost of buildings X 5=plant protection, grain drying X M = investment in equipment 246 247 APPENDIX 5 Table 5.1. Average variable costs (FIM/farm) on cattle and grain farms during the period 1982—1986a . Variable Cattle farms Variable Grain farms Cost of supplies: Cost of supplies: (exl. purchased feed (exl. purchased and fertilizers) 27 141.70 seeds) 67 456.71 Repair and maintenance: Repair and maintenance: farm buildings 8 998.98 farm buildings 9 360.47 Other costs (exl. Other costs (exl. depreciations of depreciation of land improvements) 12 525.75 land improvements) 14 991.87 Total variable costs 48 666.43 Total variable costs 91 809.05 Gross return 318 958.30 Gross return 283 905.04 Variable costs % Variable costs % from gross return 15.26 from gross return 32.33 Notes: • Assessments are made at 1986 prices. The examination covers only those variable costs that have been excluded from the regression model. Consequently, purchased fertilizers and feed, for example, are not considered as vari- able costs on cattle farms. 248 APPENDIX 6 Figure 6.1. Variable costs per farm (FIM/ha) excluded from the models explaining gross return on cattle farms as a function of farm size during the period 1982—1986, at 1986 prices. 249 Figure 6.2. Variable costs per farm (FIM/ha) excluded from the models explaining gross return on grain farms, as a function of the farm size during the period 1982—1986, at 1986 prices. 250 APPENDIX 7 Table 7.1. Nominal and real price developmentof agricul- tural land (FIM/ha) in land acquisitions by the National Board of Agriculture, 1972—1986. Year Nominal moving Real price at 3-year average price 1986 prices 3 1972 3645 12495 1973 4485 17195 1974 5340 16813 1975 5298 14743 1976 6799 14863 1977 7892 16306 1978 9064 16741 1979 9433 17568 1980 10413 14400 1981 11601 16585 1982 13048 17389 1983 14393 15931 1984 14947 17521 1985 18092 15861 1986 22993 Notes: 3 The cost-of-living index is used as a deflator. The real price of arable land has been assessed using average prices weightedby the area acquisitions by the National Board of Agriculture, see figure 14. 251 APPENDIX 8 Table 8.1. Price development of plant and livestock products (1985= 100)a . Year Plant products Livestock products 1972 22.9 26.2 1973 27.6 29.2 1974 29.5 34.4 1975 34.9 43.5 1976 40.6 49.1 1977 41.3 53.2 1978 43.5 56.3 1979 47.5 59.5 1980 58.0 65.6 1981 69.0 73.0 1982 84.1 82.3 1983 85.9 88.5 1984 91.4 94.0 1985 100.0 100.0 100.91986 104.7 Source: • Indexes assessed by the Agricultural Economics Re- search Institute. Figure 8.1. Average interest on loans (%) of cooperative and savings banks and the development of the annual change of the cost-of-living index, 1972—1986. (Assessed using the quarterly lending rates according to the statistics of the Bank of Finland.) SELOSTUS Lisäpellon tuotto- ja kauppa-arvon määrittämisen perusteet ja soveltuvuus pellon arvon osoittamiseen Etelä-Suomessa vuosina 1972—1986 Matti Ylätalo Pellervon taloudellinen tutkimuslaitos, PTT Tutkimuksen tavoitteena oli lisäpellon tuotto- jakaup- pa-arvon määrittäminen Etelä-Suomen kirjanpitoalueella vuosina 1972—1986. Tämä edellytti tarkoitukseen sovel- tuvien laskentamenetelmien valintaan jakäyttöön liitty- vien kysymysten tarkastelua. Tällöin pellon arvon ajatel- tiin määräytyvän pelkästään sen käytöstä viljelytarkoi- tuksiin saatavasta hyödystä. Samoin pellosta maksetut hinnat kuvaisivat pelkästään viljelytarkoitukseen hanki- tun pellon arvoa ilman ns. odotusarvotekijöitä. Aluksi tarkasteltiin keskeisiä maatalousmaan arvioin- neissa käytettyjä arvo- ja hintakäsitteitä sekä niiden so- veltuvuutta eri käyttötarkoituksiin. Suomen kiinteistöar- viointiyhdistyksen hyväksymän terminologian mukaan tuottoarvolla ymmärretään ’’nykykäytön tai käyttömah- dollisuuksien perusteella arvioidun tuoton pääoma-arvona määritettyä käypää arvoa”. Vastaavasti kauppa-arvolla tarkoitetaan kauppahintojen perusteella määritettyä käy- pää arvoa eli käypää hintaa. Lisäpellolla tarkoitetaan puo- lestaan tilan muiden peltojen yhteydessä viljeltävää jo han- kittua tai hankittavaksi aiottua peltoaluetta, jota aiotaan käyttää maataloustuotantoon. Näitä määritelmiä sovel- lettiin myös tässä tutkimuksessa. Tutkimuksen toisessa luvussa tarkasteltiin lisäpellon ar- voon ja hintaan vaikuttavia keskeisiä tekijöitä. Ensiksi tutkittiin pellon laadun sekä kysynnän ja tarjonnan vai- kutusta. Tällöin todettiin viljelyteknillisten ja taloudel- listen toimenpiteiden vaikutuksen korostuneen maatalou- den taloudelliseen tulokseen ja siten myös maan arvoon vaikuttavina tekijöinä, kun taas maan laadun merkitys oli alentunut. Lisäpellon voimakas kysyntä sen vähäiseen tarjontaan nähden osoittautui pellon hintasoa kohottavaksi tekijäksi, varsinkin kun teknologinenkehitys ja erityisesti konei- den lisääntynyt käyttö olivat tehneet mahdolliseksi hoi- taa aikaisempaa suurempaa viljelmää. Kuilu vallitsevan janykyteknologian avulla hoidettavan viljelmäkoon vä- lillä osoittautui suureksi. Aikaisempien tutkimusten mukaan maataloustuotan- toon ja sen kannattavuuteen vaikuttavat tekijät heijas- tuvat myös maan arvoon. Siten yksittäisten tekijöiden osuuden erottaminen lisäpellon hintaan vaikuttavista kai- kista tekijöistä osoittautui ongelmalliseksi, vaikka ne kyet- täisiinkin ryhmittelemään yleisiin, laadullisiin ja yksilöl- lisiin tekijöihin. Luvussa kolme käytiin lävitse aikaisempia tuotto- ja kauppa-arvoperusteisia tutkimuksia sekä niissä sovellet- tuja menettelytapoja. Aluksi tarkasteltiin tuottoarvon las- kennan perusteita ja sen riippuvuutta eri tekijöistä. Sen jälkeen esitettiin kansantaloustieteessä käytetty, maankor- koon pohjautuva lähestymistapa, jotamyös on käytetty maatalousekonomisissa tutkimuksissa. Eniten käytetty li- säpellon arvon määrittämismenetelmä maataloudessa on kuitenkin ollut lisäpellon tuoton määrittäminen jäännök- senä, kun muiden tuotantopanosten käytöstä aiheutuvat kustannukset vähennettiin kokonaistuotosta. Keskeiseksi ongelmaksi siinä nousi tuotantovälineiden ja panosten käytön erilaisuuden lisäksi niiden hinnoitteluun liittyvät vaikeudet. Tällä menetelmällä saatiin pinta-alaltaan pie- nehköillä viljelmillä alhainen tai jopanegatiivinen tuot- to pellolle. Kolmantena lisäpellon tuoton laskentamene- telmänä tarkasteltiin rajatuottavuusteoriaan pohjautuvaa lähestymistapaa, jossa tuotantovälineet saavat niiden ra- jatuottoa vastaavan korvauksen tuotannon tuloksesta. Teorian soveltaminen edellytti tuotantofunktioiden esti- mointia. Kolmannen luvun jälkimmäisessä osassa (3.2.) esitel- tiin kauppa-arvoperusteisia tutkimuksia sekä todettiin kauppahinta-aineiston saatavuuteen ja edustavuuteen liit- tyvät ongelmat. Lisäpellon hintatutkimuksia oli kyllä tehty maassamme useita, mutta pitkän aikavälin hintatilasto- ja oli saatavissa ainoastaan valtion maanhankintatilastosta ja kaupanvahvistajien hintatilastoista. Siksi Suomessa teh- dyt aikaisemmat hintatutkimukset yleensä keskittyivät yh- den suppean alueen tai paikkakunnankauppahintojen tar- kasteluun muutaman vuoden osalta. Tämän tutkimuksen empiirisessä osassa tutkimus- alueeksi valittiin Etelä-Suomi, jolla alueella maatalouden luontaiset edellytykset (maaperä ja ilmasto) ovat parhaat. Lisäpellon tuottoarvon määrittämistä varten tutkimusai- neistoksi valittiin kirjanpitoviljelmät, jotka olivat yhtä- jaksoisesti olleet mukana maatalouden kannattavuustut- kimuksessa vuodesta 1968 lähtien. Lopulliseen tutkimus- 252 aineistoon kelpuutettiin kuitenkin vain nautakarja- ja vil- jatilat, joilla tuotannon katsottiin ensisijaisesti perustu- van viljelysmaan hyväksikäyttöön. Ennen tuotantofunktioiden estimointeja kirjanpitovil- jelmien poistonalaisten omaisuusosien verotuksenmukai- set arvot muutettiin elinkustannusindeksillä vastaamaan paremmin niiden markkina-arvoja. Muuttujien valinnassa käytettiin hyväksi korrelaatioanalyysejä ja kokeiluja. Aluksi testattiin eri funktiotyyppejä; lineaarista, Cobb- Douglas ja erästä transkendentaalifunktiota. Näistä va- littiin Cobb-Douglastuotantofunktio lisäpellon rajatuo- ton laskentaan. Laskelmat suoritettiin sekä poikkileik- kaus- että aikasarjatarkasteluakäyttäen vuosilta 1972— 86. Nautakarjatilojen rajatuotot määritettiin vuosilta 1972, 1976, 1980 ja 1982—86. Viljatilojen vuotuisen sa- tovaihtelun vaikutuksen tasoittamiseksi laskelmat laadit- tiin kolmen perättäisen vuoden tietojen pohjalta vuosil- ta 1975—77, 1979—81 ja ajanjaksolta 1982—86 koko ajan samassa tuotantosuunnassapysyneiden viljelmien osal- ta. Luotettavimmin lisäpellon keskimääräinen rajatuot- to pystyttiin määrittämään kummassakin tuotantosuun- taryhmässä viimeiseltä tutkimusperiodilta. Pellon hyvyy- den mittareina regressiomalleissa käytettiin satotaso- dummeja (taulukko 16). Pellon rajatuotosta vähennettiin tämän jälkeen sellai- set muuttuvat kustannukset, jotka eivät sisältyneet nautakarja- tai viljatilojen tuotantofunktion selittäviin muuttujiin. Näin saatu rajakatetuotto osoitti sen mark- kamääräisen korvauksen hehtaaria kohden, mikä nauta- karja- ja viljatiloilla oli käytettävissä pellon hankintaan ennen veroja. Verojen osuuden määrittämiseksi maata- louden tulonlisäyksen aiheuttamaksi marginaaliveroas- teeksi valittiin nautakarjatiloilla 40 ja viljatiloilla 50 Vo. Kertomalla lisäpellon rajakatetuotto kertoimella (1-raja- veroaste) saatiin molempien tuotantosuuntaryhmien pel- lon nettorajakatetuotto hehtaaria kohden selville (tauluk- ko 18). Lisäpellon tuottoarvoa laskettaessa käytettiin 3, 5 ja 7 %:n korkokantaa, kun tuoton kestoajaksi valittiin 5, 10 ja 15 vuotta. Näin määritetyt tuottoarvotnousivat ta- saisesti nautakarjatiloilla ajan myötä. Viljatiloilla tuot- toarvo jäi sitä vastoin hyvin alhaiseksi tutkimuskauden viimeistä periodia lukuunottamatta (taulukko 23). Tutkittaessa pellon rajatuoton ja viljelmäkoon välistä riippuvuutta tilat jaettiin kummassakin tuotantosuunta- ryhmässä kahteen suuruusluokkaan. Tulokset osoittivat pellon rajatuoton alle 25 hehtaarin nautakarjatiloilla ol- leen keskimäärin noin kolminkertainen suuremman vil- jelmäkokoryhmän tuloksiin verrattuna. Sitävastoin vil- jatiloilla pellon rajatuotto kasvoi edelleen siirryttäessä alle 30 hain tilaryhmästä suurempaan. Myös keskimääräisen verotettavan puhtaan tuoton käyttöön perustuva erotus- tuottotarkastelu osoitti pellon rajatuoton tapaan pelto- alan lisäämisestä saatavan hyödyn olevan suurin nauta- karjatilojen pienimmissä tilasuuruusluokissa, kun taas vil- jatiloilla pellon rajatuotto ja erotustuotto pysyivat kor- keina viljelmäkoon kasvaessa. Lisäpellosta maksetut kauppahinnat kerättiin maatila- hallituksen maanhankintatilastosta vuosilta 1972—86 ja maanmittaushallituksen tilastoimasta kiinteistöjen kaup- pahintarekisteristä. Jälkimmäisen tilaston tiedot oli saa- tavissa vain vuosien 1982—86 osalta. Tuona viitenä vuo- tena kauppahintarekisterin mukaiset kaupat Etelä-Suo- messa käsittivät keskimäärin 1233 ha vuodessa, kun taas maatilahallituksen pellon ostoissa maata vaihdettiin vain 505 ha vuodessa. Vuotuiset pellon ostot olivat siten tut- kimusalueen 1.1 milj. hain suuruisesta peltoalasta hieman yli promille. Lisäpellon pinta-alalla painotettukeskihinta maanmit- taushallituksen kaupoissa nousi voimakkaasti vuoteen 1985 saakka, mutta aleni viimeisenä tutkimusvuonna ku- ten seuraavastaelinkustannusindeksillä vuoden 1986 hin- tatasoon deflatoidusta aikasarjasta havaitaan: mk/havuosi 25 2401982 27 6921983 31 6411984 33 0041985 31 920.1986 Keskimääräinen lisäpellon hehtaarihinta tutkimus- alueella tarkasteltuina vuosina oli 29 899 mk. Vuonna 1986 pellosta maksettiin korkeimmat hinnat Satakunnan ja Varsinais-Suomen maatalouskeskusten alueilla, joilla pellon hinta ylitti 43 000 mk hehtaarilta. Myös muina vuo- sina pellon hinta näillä alueilla oli selvästi korkein. Seu- raavaksi kalleinta pelto oli vuosina 1982—86 Hämeen lää- nin maatalouskeskuksen alueella vaihtelurajojen ollessa 23 000—35 300 mk/ha. Halvinta pelto oli sitävastoin Etelä-Karjalan, Pirkanmaan, Itä-Hämeen jaKymenlaak- son maatalouskeskusten alueilla. Näillä alueilla pellon reaalihinta kyseisinä vuosina vaihteli 15 700—28 300 mk/ha. Maatilahallituksen pellon ostoissareaalihinnan nousu vuosina 1972—86 oli hyvin maltillinen. Trendin vuosi- muutosten keskiarvo oli 1.7 %. Salaojitetun ja avo-ojite- tun pellon määrällä oli merkittävä vaikutus maatilahal- lituksen ostaman pellon hintaan. Vuosina 1978—86 sa- laojitetusta pellosta maksettiin 37—60 % avo-ojitettua peltoa enemmän. Maatilahallituksen pellon ostoissa hin- tataso jäi selvästi maanmittaushallituksen hintoja alhai- semmaksi, sillä se oli keskimäärin vain noin 60 % kaup- pahintarekisterin hinnoista. Koska kirjanpitoviljelmien tulokset ovat noin viiden- neksen alueen muita tilojakorkeammat, lisäpellon tuot- toarvoihin tehtiin samansuuruinen tasoalennus. Taso- alennus huomioonottaen vertailu tuottoarvonjakauppa- hintojen välillä osoitti lisäpellon tuottoarvon nautakar- jatiloilla yltäneen maatilahallituksen maksamiin pellon hintoihin kaikkina tarkasteltuina vuosina, kun tuotonkes- toaikana käytettiin 15 vuotta ja laskentakorkokantana 3 %:a (vrt. kuva 15). Kymmenen vuoden kestoajalla las- kettu tuottoarvo ei enää yltänyt maatilahallituksen hin- 253 tatasoon kaikkina tarkasteltuina ajankohtina. Myös maanmittaushallituksen kauppahintojen taso ylitti vuo- sina 1982—86 noin 10 000 mk:lla nautakarjatilojenkor- keimman tuottoarvon. Viljatilojen lisäpellon tuottoarvo jäi maatilahallituk- sen kauppahintojen alapuolelle kaikkina tarkasteltuina ajankohtina (vrt. kuva 16). Viljatilojen tuottoarvo oli sel- västi korkein vuosina 1982—86, jolloin se ylsi tasoalen- nus huomioonottaen lähes maatilahallituksen kauppahin- tojen tasoon 15 vuoden kestoaikaa ja 3 %:n laskentakor- kokantaa käyttäen. Kun viljatilojenkorkein tuottoarvo tuolloin oli noin 16 000mk/ha, ja maanmittaushallituk- sen kauppahinnat lähes 30 000 mk, kauppahinnat olivat siten lähes kaksi kertaa korkeammat. Viljatilojen lisäpel- lon tuottoarvo osoittautui myös selvästi nautakarjatilo- ja herkemmäksi satotason vaihtelulle. Tuloksia arvioitaessa on syytä muistaa kirjanpitotilo- jen edustavan keskimääräistä suurempia viljelmiä, min- kä lisäksi ne ovat rationaalisesti ja voimaperäisesti vil- jeltyjä tiloja. Tästä syystä tulosten yleistämisessä on ol- tava varovainen. Samoin on asianlaita myös kauppahin- tojen suhteen. Siksi edellä esitetyt tutkimustulokset osoit- tavat lisäpellon keskimääräisen tuottoarvon ja kauppa- hintojen tason jakehityksen Etelä-Suomessa vain suuntaa- antavasti. Poikkeamat yksittäisillä viljelmillä voivat olla suuriakin. 254