122 SAJEMS NS Vol I (1998) No I Free Trade and Deregulated Domestic Markets: Attitudes of Commercial Farmers in K waZulu- Natal R C Stockil and G F Ortmann Department of Agricultural Economics, University of Natal, Pietennaritzburg ABSTRACT This study analyses farmers' attitudes towards free trade and deregulated domestic product and input markets using a survey conducted in 1996 among 112 commercial farmer~ in KwaZulu-Natal. Most respondents were in favour of Iiberalised trade and deregulated domestic markets, but expected a decline in product prices, farm profits and land values. Logit analyses of farmers' personal and business characteristics that influence their attitudes towards free trade and deregulated domestic markets were conducted. Results indicate that improved information on risk: management practices and impon tariff levels may help commercial farmers to adapt to a changing economic environment and reduce resistance to free trade. INTRODUCTION The Uruguay Round of the General Agreement on Tariffs and Trade (GA Tn. which was formally concluded on 15 December 1993 among 117 participating countries, aims to Iiberalise intemational trade (Giardini, 1995). In 1990. South Africa had 21 agriCUltural marketing boards which had a profound influence over domestic product prices. imports and exports, and the manner in which agricultural products were marketed. With the exception of most vegetables and subtropical fruit, marketing of primary food and fibre agricultural products was managed in some or other form by producer-dominated marketing boards (Swart. 1996), R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SATEB NR Vol 1 (1998) Nr 1 123 South Africa is a signatory to the GAlT, resulting in increased market access here and abroad, reduced domestic support and export subsidies, and a revision of sanitary and phytosanitary measures (Department of Agriculture. 1994), Consequently, the elimination of producer controls over imports, tarification of agricultural imports, the removal of most single channel marketing schemes and a general reduction in the authority of marketing boards has been experienced, The remaining 15 marketing boards perform mainly non-trade distorting functions such as generic advertising and market information (Swart. 1996). Amendments to international trade policy. together with improved dispute settlement procedures and monitoring provided by the Committee on Agriculture in the World Trade Organisation (WTO). will create conditions favourable for more stable international agricultural trade flows. Agricultural production will therefore be determined more by comparative advantage rather than the size of national budgets. Trade Iiberalisation will also increase import competition forcing industries to improve productivity through the adoption of innovations (Maclaren. 1995). International prices of agricultural products are expected to increase in the medium term following reduced supply from developed countries that pay high export subsidies (Department of Agriculture, 1994). However, technological improvements may reduce prices in the long term (Goldin and Knudsen, 1990). A greater degree of price uncertainty faced by commercial farmers is also expected in the event of reduced government protection (Lyne and Ortmann, 1992). The objectives of this paper are (I) to ascertain farmers' responses to GAlT and deregulated domestic markets, and (2) to determine the factors influencing attitudes of commercial farmers in KwaZulu-Natal towards free trade and deregulated domestic markets by using logit models. Education and improved information flows can· help farmers to survive in a changing policy and trade environment, and it is for this reason that factors affecting farmers' perceptions towards free trade are important. Following Kastens and Goodwin (1994), this paper analyses farmers' attitudes regarding a non-specific policy environment in which no particular agricultural programme is evaluated, which differs from other studies. The paper concludes with a discussion of the management and policy implications of the results obtained. PASr RESEARCH ON A TI1TUDES TOWARDS LIBERALISED TRADE Farmers are assumed to be rational, profit maximising decision makers and will therefore support agricultural policies in which they perceive the greatest benefit (Orazem et ai, 1989; Barkley and Ainchbaugh, 1990). Research in the USA has R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 124 SAJEMS NS Vol I (1998) No 1 shown that fann operators with higher levels of education and experience display a resistance towards policy liberalisation because they are better able to comprehend the complex regulations and resulting benefits of agricultural progranunes and are less suspicious of these programmes (Kastens and Goodwin. 1994; Barkley and Ainchbaugh, 1990). Farmers who, assessed on a subjective basis, are less risk averse are less opposed to the riskier marketing conditions resulting from trade Iiberalisation. Farmers operating a greater proportion of rented land were more likely to support a free trade policy environment (Kastens and Goodwin, 1994). Specialised farms. probably because of their riskier nature, are less likely to support liberalisation than diverse operations and farm types enjoying relatively high levels of government support, which are again expected to oppose policy liberalisation (Kastens and Goodwin, 1994). Larger farms in the USA show a greater degree of support for policy liberalisation. TIris may be as a result of limitations placed on individual farmers' benefits, or may indicate that larger fanners perceive greater benefits from economies of sire in a free trade and free market environment. If farmers' perceptions are correct, policy Iiberalisation may lead to increased farm sire (Kastens and Goodwin, 1994). Edelman and Lasley (1988) indicate that amongst USA fanners, acres owned and com acres were significantly positively related to pursuing an open market and there was a positive relationship between decoupling (which provides direct support to farmers but does not distort production, conswnption and trade) and farm sire. Conversely, research by Orarem et al (1989) indicates that the largest farms tend to oppose decoupling because it would result in lower crop prices and returns per hectare. As farmers' wealth (net worth) increases, their preference for a Iiberalised policy environment rises (Kastens and Goodwin, 1994). Farmers under financial stress are less likely to support market-orientated domestic policies or the elimination of domestic policies because of increased price risk (Edelman and Lasley, 1988). The greater the number of days worked off the farm, the lower is the support for the continuation of present government programmes as producers with significant off-farm income are less dependent on government programmes and are thus more likely to support market liberalisation (Edelman and Lasley, 1988; Barkley and Ainchbaugh, 1990). R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SATED NR Vol 1 (1998) Nr 1 125 DATA SOURCE A"'D CHARACTERISTICS OF RESPOl\'DESTS This study was conducted in 1996. The target population of 4436 corrunercial farmers in KwaZulu-Natal was stratified following Lyne and Ortmann (1996) and randomly sampled. The first stratum, constituting bioclimatic groups 1 and 2 (Phillips, 1973), is termed the Coastal Belt. Bioclimatic groups 3, 4, 5, 6 and 8 represent the KwaZulu-Natal Midlands. whilst bioclimatic groups 7, 9, 10 and 11 comprise the Lowveld region. A questionnaire was mailed to 578 farmers yielding 149 responses (25,8 percent) of which 112 (19,4 percent) were usable. The non- usable questionnaires were due to missing values, the sale of farming operations, or the retirement of some farmers. Twenty-six usable questionnaires were received from each of the Coastal and Lowveld regions, whilst 60 usable questionnaires were returned by Midlands farmers. Respondents were on average 49,5 years of age, had 24,3 years of fanning experience and 14, I years of formal education. Sixty percent of respondents were individual owners of their farm business. 12 percent of the operations were close corporations. 10 percent companies, nine percent trusts and nine percent of respondents operated partnerships. Thirty-nine percent of respondents were employed to manage the fanning operations. Computers were owned and used in the farm business by 64 percent of respondents, which is much higher than the 48 percent reponed by Woodburn et al.(I994). The average area operated in the Coastal Belt was 511 hectares (median 313 hectares). 1121 hectares in the Lowveld region (median 253 hectares. due to a predominance of irrigated sugar-cane farms and a few large extensive beef enterprises). and 866 hectares in the Midlands (median 604 hectares). More than 70 percent of gross farm income was derived from sugar-cane production by 75 percent of farmers in the Coastal Belt. Sixty-two percent and 15 percent of farmers in the Lowveld region derived more than 70 percent of gross farm income from sugar-cane and beef production respectively. The Midlands region is characterised by mixed farming activities. Among the farmers who received 70 percent or more of their gross income from a single enterprise. 27 percent were beef farmers, 12 percent dairy farmers. 10 percent sugar-cane farmers. eight percent timber farmers and six percent pig farmers. About 86 percent of respondents were full-time farmers. Median household income from off-farm employment (including spouse's income) was R47 375 for the 35 percent of respondents who indicated off-farm employment (by themselves or their spouse). Estimated market values of assets, debt/asset levels and farm turnover (gross income) in a normal year for the three regions are presented in R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 126 SAJEMS NS Vol 1 (1998) No 1 Table 1. The Coastal Belt had the highest mean twnover and debt/asset ratio, followed by the Lowveld and Midlands regions respectively. The debt/asset ratio should preferably be less than 0,5 (Barry et ai., 1995). At a nominal interest rate of 15 percent, farmers will experience cash flow problems if borrowed capital exceeds one-third of the value of fannland. as the retwn to farmland in approximately five percent (Nieuwoudt and Vink, 1995). The ratios presented in Table 1 indicate that sample farmers were solvent. Ninety-two percent of respondents had ratios of less than 0,30. Table 1: Fann asset values, debt/asset ratios and turnover of sample fanns in three regions of KwaZulu-NataI, February 1996. Asset value (Rm) Debt/asset ratio Turnover (Rm) n Mean Median n Mean Median n Mean Median Coastal 25 5,612 3.691 18 0,149 0,141 22 2.086 1.471 (n=26) Lowveld 25 5,676 3,800 25 0.131 0,116 26 1.594 1.076 (n=26) Midlands 50 2,709 1.670 48 0,083 0,012 54 1.147 0,620 (n=60) Land is cash-rented by 21 percent of respondents, with the area rented ranging from 10 to 900 hectares (mean of 264 hectares). Two respondents were involved in share-lease agreements (50 and 100 hectares), while a single respondent rented out land (150 hectares). RESPONSES TO GAIT AND DEREGULATED DOMESTIC MARKETS Given coverage of GAIT (today the WTO) in the popular press and farming magazines, it was anticipated that most respondents would be aware of changes to international agricultural policy. GAIT focuses primarily on reducing tariff barriers to trade in industrial and agricultural products (Schuh, 1995). Signatories to GAIT are required to reduce internal support (Giardini, 1995), all non-tariff barriers are to be convened to tariffs, individual tariff lines reduced and import opportunities must be allowed for to facilitate opening-up of markets (Josling, 1993). The volume of subsidised exports is to be reduced and sanitary and phytosanitary (SPS) measures revised and tightened (Giardini, 1995). R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SATEB NR Voll (1998) Nr 1 127 Only one respondent indicated that he was not aware of GA IT. Seventy percent of respondents gained information on GAIT by reading newspapers, whilst 45 and 31 percent obtained information from Effective Fanning and Fanner's Weekly respectively. Twenty-eight percent indicated attendance at farmers' days whilst Financial Mail, Latulbou Weekblad and industry publications (eg Sugar Journal) were used by 23 percent of respondents to gain information on GAIT. Private consultants were employed by 10 percent of respondents. nine percent read Finance Week, six percent read Finansies en Tegniek, and only five percent made use of extension officers. Sixty percent of respondents intended seeking additional information on GAIT. Fifty-two percent of respondents expected a decrease in their crop prices in the event of the successful implementation of the GAIT requirements, whilst 25 percent expected their crop prices to increase. A decrease in livestock prices was expected by 61 percent of respondents and 12 percent thought that livestock prices would increase. Approximately equal proportions of respondents expected input prices to decrease (43 percent) and increase (39 percent). However, a greater proportion of crop and livestock producers (i.e. those who realised more Iban 55 percent of their gross income from either crop or livestock production respectively) thought a decrease in input prices likely. As is shown in Table 2, about one-half of respondents foresaw a decline in farm profits and 38 percent a fall in land values if the GAIT requirements are successfully implemented. Following the reduction of import tariffs on meats in 1995. beef respondents experieoced a 24 percent decrease in beef prices. Poultry farmers indicated a 29 percent reduction in broiler prices, mutton farmers a 16 percent decrease in mutton prices. and pork farmers a 17 percent decrease in pigmeat prices. Seventy-seven percent of respondents did not know the level of import tariffs on the products they produced, whilst 90 percent did not know the levels of import tariff on the inputs they used. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 128 SAJEMS NS Vol 1 (1998) No 1 Table 2: Percentage of respondents expecting changes to product and input prices, farm profits and land values foDowing the GATT provisions, 1996. Decrease Increase No change Uncertain (percent) (percent) (perrent) (percent) Crop prices (n=103) 52 25 14 9 Livestock prices (n=81) 61 12 10 17 Input prices (n= 108) 43 39 11 7 Crop farmers' (n=54) 45 35 11 9 Livestock farmers . 49 38 8 5 (n = 31) Farm profits (n = 109) 49 19 16 16 Land values (n = 110) 38 14 22 26 "Fanners who received more than 55 perrent of their gross income from either crops or livestOCk. A five-category scale, ranging from 'strongly disagree' with the statement given to 'strongly agree' (Table 3), was used to elicit farmers' attitudes towards free trade. For the various questions posed, between 59 and 92 percent of respondents were in favour of free trade. Ninety-two percent of respondents were in favour of the deregulation of domestic product and input markets, indicating that they perceived the recent deregulation of product marketing boards to be of benefit to their fann businesses. This conclusion may not apply generally as approximately 74 percent of the original 578 farmers in the sample did not return the questiormaire, while those who did may have definite views on liberalised trade. Thirty-nine percent of all respondents, including 31 percent of livestock producers and 40 percent of crop farmers, would respond to further reduction in import tariffs by adjusting their farming operations. In the event of complete deregulation of domestic product and input markets, 48 percent of respondents, including 49 percent of livestock producers and 46 percent of crop farmers, would alter their farming operations. A greater proportion of crop farmers are corx:erned with further reduction in import tariffs, while nearly one-half of livestock producers are concerned with the complete deregulation of domestic markets. The recent high maize price, following on the deregulation of the maize industry. may be of concern to livestock producers. A large proportion of sugar-cane producers R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SATES NR Vol I (1998) Nr I 129 in the sample may have been concerned about changes in the domestic sugar price resulting from reduced impon tariffs. Of the respondents who considered altering their cost structure in response to successful implementation of the GAIT provisions, 29 percent would reduce labour employment, increase use of machinery, and maintain fenilizer use at current levels. Of the 45 percent of respondents who would change the size and mix of enterprises currently operated, 46 percent would increase the size of current enterprises. Both enterprise diversification (adding a new enterprise) and changing the relative sizes of current enterprises were suggested by 23 percent of respondents. Reducing enterprise size (four percent of respondents) and changing farm enterprises totally (four percent) were also considered. Table 3: Response of sample farmers to statements regarding a free trade environment, 1996. Percentage of respondents in each category Statement Strongly Disagree • Un~ Agree Strongly disagree certain agree Free, unrestrained international 5 16 20 50 9 trade (without the interference of governments both here and abroad) is beneficial for SOUID African farmers (n= 109). A free. open market system of 2 18 13 51 16 trade (ie. one wiIDout government intervention) should be pursued by all food exporting and importing countries by reducing all trade barriers (n=IIO). Farmers in all countries exporting 4 11 7 52 26 agricultural commodities should not receive any government support (n= III). SOUID African farmers should 0 3 5 71 21 compete in a deregulated (free) domestic product market. if input markets are also deregulated I (n= I 10). R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 130 SAJEMS NS Vol I (1998) No 1 Of the 64 percent of respondents who would respond to reduced import tariffs and/or complete domestic market deregulation, 46 percent would seek additional marketing infonnation to aid management decisions and 44 percent would add value to their products. Forty percent of respondents considered purchasing additional technology to improve productivity, 35 percent considered exporting products, 32 percent employing machinery contractors and 24 percent would employ labour contractors. Employing the services of a private consultant or restrucruring debt commitments were considered by 22 percent of respondents, 19 percent indicated engaging in forward contracts, 18 percent borrowing less capital or reducing family drawings, and 10 percent would trade on the futures market (SAFEX). The relatively small proportion of respondents considering trading on the futures market could possibly be due to the limited range of contracts currently available (eg there are no sugar contracts at present). Selling land was considered by 12 percent of respondents, 10 percent intended seeking off-farm employment, nine percent would introduce capital invested off the farm into the farm business. and six percent would sell assets or borrow more capital. Sample farmers were also asked to rate their level of management skill relative to other farmers in their district on a five-point Likert-type scale (where 1 ;: low and 5 = high). The average rating for their managerial ability in farm production and overall farm management was 3,72, in farm finance 3.61, and 3, II for product marketing. In the past, 21 marketing boards marketed about 90 percent of the total value of agricultural production (Lyne and Ortmann, 1992), thereby reducing the need for farmers to market their products. For farmers to survive in a deregulated policy and trade environment with increased price risk, product marketing skills will need to be improved and/or marketing experts employed. FARMERS' ATTITUDES TOWARDS FREE TRADE AND DEREGULATED DOMESTIC MARKETS The personal. financial and farm size characteristics of farm operators may be important factors influencing their attitudes towards free trade (Kastens and Goodwin, 1994). Consequently, the objective of this section is to determine the personal and business characteristics of respondents that influence their perceptions regarding free trade. A logit model is developed for this purpose. The dependent variable (yi) is dichotomous, scoring one if the respondent agreed with free trade (or deregulated domestic markets), and zero otherwise. Equation (I) can be used to estimate the log odds of the probability of an event occurring. in this case the probability that a respondent wiJI favour a free· R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SATEB NR Vol 1 (1998) Nr 1 131 trade enviromnent. Respondents who 'strongly disagreed' or 'disagreed' with the statement regarding free trade (first question in Table 3) were allocated a zero,while respondents who 'agreed' and 'strongly agreed' were coded a one. 'Uncertain' responses were excluded from the analysis. n In [pd(l-p; )] = a + L f3kXki (1) k=l where pi is the probability of the ith respondent favouring a free-trade environment and Xk the kth explanatory variable. The dependent variable In[p;/(l-pi)] is the natural log of the odds ratio in favour of the ith respondent being in favour of free trade (Gujarati, 1995). Independent variables considered in the model Based on previous studies discussed above. personal characteristics considered include years of farming experience. level of education. and willingness to take risks. Proponion of land rented, farm type. turnover (farm size). operator's wealth (net worth), off-farm employment. distance to nearest town. the debVasset ratio, and a measure of financial stress (incorporating both solvency and liquidity measures) were the business characteristics considered. Exploratory analysis was also pursued to determine other factors affecting farmers' attitudes towards free trade. Variables which were hypothesised to have an impact on free-trade attitudes are defined in Table 4. Hypothesised relationships between free-trade attitudes and EXP. EOU, RENT, SIZE, NWORTH and RISK are consistent with those of past studies. Increased price risk is expected with free trade, and thus farmers who have to contend with the additional fmancial risk associated with borrowed capital (01 A, FSTRESS. FFOPT and REPAY) are likely to oppose free trade. Beef (BEEF) and sugar-cane (SUGAR) farmers are likely to be faced with reduced product prices under free-trade conditions (lower tariffs), and most beef fanners are unlikely to benefit substantially from reduced input costs under free trade owing to the extensive nature of production. These farmers are also expected to oppose free trade. Milk is highly perishable and thus not easily exportable, and opportunities exist to add value to it. However, increased competition from dairy products originating mainly in the EU and New Zealand may lead to lower R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 132 SAJEMS NS Vol 1 (1998) No I product priceS. Dairy fanners (DAIRY) are, therefore, hypothesised to oppose free trade too. COMPUT is expected to have a positive effect on farmers' free-trade attitudes as it is a measure of farmer progressiveness. Rapid information flows are vital in a free market, and farmers using computers can develop a competitive edge over other fanners due to more timeous information. DIST is expected to have a negative relationship because proximity to markets is an imponant advantage to farmers if they are to market their own products. Table 4: Variable EXP EDU RENT SIZE NWORTH RISK D/A FSTRESS FFOPT REPAY BEEF SUGAR Definition of variables expected to influence farmers' attitudes towards free trade. Definition Expected relation- ship Years of farming experience. - Years of formal education. Percentage of the total area operated that is rented. Annual gross income from farming operations (Rand). Net wonh of fann business (Rand). + Fanners' willingness to take risks relative to other + fanners in the district (measured on a Liken-type scale ranging from 1 = much less willing to 5 = much more willing). Debt to asset ratio. Measure of financial stress (liquidity and solvency) - scoring 0 if low up to 4 if high (Orazem et ai, 1989). Fixed financial obligations as a prop onion of - turnover. Value of medium and long term debt repayments (RlOO 000). Dummy variable = I if more than 55 percent of - gross income is derived from beef production, 0 otherwise. Dummy variable = 1 if more than 55 percent of - gross income is derived from sugar-cane production, 0 otherwise. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SATEB NR Vol 1 (1998) Nr 1 133 Table 4 (continued) DAIRY Dununy variable = I if more than 55 percent of - gross income is derived from dairy production, 0 otherwise. COM PUT Dummy variable = 1 if a computer is used in the farm business, 0 otherwise. + DIST Distance of farm from nearest town (km). - ENVIRO Rating of changes in environmental reg.ulations as - a source of risk (I = low and 5 = high) . RISKFAC Factor consisting of variability in crop prices. - changes in the Rand exchange rate, cham.!es in costs of inputs and further reduction of tarIffs on i imported farm products as sources of risk, as per i section 5(b) (1 = low and 5 = high for all sources of risk). LPRICE Rating of variability in livestock prices as a source of risk (1 = low and 5 = high). - LANDP i Dununy variable I if farmer expects land prices to increase if GATT provisions are successfully implemented, 0 otherwise. + LANDRED Rating of further land redistribution and/or I + restitution by government as a source of risk (I i low and 5 = high). MFINAN Self-rating of management skill in farm finance (I = low and 5 == high). + PORF Dununy variable 1 if respondent is involved - full-time in the farming operation, 0 otherwise. PORF is an inverse measure of off-farm income. TLEVEL Dummy variable = I if respondent indicated some knowledge of tariff levels on products sold or inputs used, 0 otherwise . - • All ratings. where I = low and 5 high. are based on a Likert-type scale Farmers who attach greater imporraoce to various risk sources are expected to have a negative attitude towards free trade (which gives rise to greater price risk). Hence, the variables ENVIRO, RISKFAC and LPRICE are expected to have a negative relationship with the dependent variable. Note that RISKFAC was developed for the following reason: Variability in crop prices, changes in the Rand exchange rate, changes in the cost of inputs and funher reduction of import tariffs on agriCUltural products are sources of risk expected to influence respondents' R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 134 SAJEMS NS Vol 1 (1998) No 1 attitudes towards free trade. Multicollinearity was anticipated among these variables owing to significant (one percent) zero-order correlation coefficients. A factor (RlSKFAC). explaining 51 percent of the variation in these variables. was therefore created. Farmers expecting profits. and thus land prices (LANDP), to fall (rise) under free trade conditions are expected to have a negative (positive) attitude towards free trade. However, fanners who perceive further land redistribution as a source of risk (LAND RED) may see an opportunity in free trade to secure their land base if they can become relatively more competitive. If this is the case, a positive relationship between LANDRED and attitudes towards free trade is expected. MFINAN is hypothesised to have a positive effect on free-trade attitudes. Farmers who rate their financial skills highly most probably manage financial risk more effectively and are thus in a better position to manage increased price risk under free trade. Previous studies have reported a positive relationship between off-farm income earned by farmers and their attitude towards free trade; these farmers are less reliant on income from government programmes (eg Orazem et aI, 1989; Barkley and Flinchbaugh, 1990). Heoce, a negative relationship is expected between PORF and attitudes towards free trade. Knowledge of tariff levels is also expected to influence free-trade attitudes. Reduction of tariffs on imported products is expected to reduce local prices and heoce fanners' profits. However, lower impon tariffs on farm inputs are expected to have a positive effect on farmers' finances owing to lower input costs. Hence, if farmers have more knowledge of (or are more concerned with) tariffs on imported !ann products than of input tariffs, a negative relationship between TLEVEL and attitudes towards free trade is expected. The general logit model can now be defined as: Infpil(l-pl)] = {Ja + PI EXPi + ,lhEDUi + foRENTi + foSIZR + ,BsNWORTHi + ,BsRISKi + {JIDIAi + /lIFSTRESSi + ,&FFOPTi + PIOREPA Yi + PIlBEEFi + pl2SUGAR.J + PI3DAIRYi + jJl.COMPUTi + PI~DlSTi + PI6ENVIROi + jJl7RISKFACl + jJlsLPRICB + PI9LANDPi + /hlTLEVEL (2) R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SA TEB NR Vol 1 (1998) Nr 1 135 Results of the analysis Independent variables included in other studies proved to be poor determinants of fanners' attitudes towards free trade in this study. This may be due to a relatively small sample and lack of variation in the data. Only variables with coefficients significant at the ten percent (or higher) level of probability were retained in the logit model. Results of the analysis are presented in Table 5. The model Chi-square statistic is highly significant. The goodness of fit statistic shows no significant lack of fit in the overall model. The estimated model correctly classifies 92,59 percent (50 of 54) of those in favour of free trade and 75,00 percent (12 of 16) of those opposed to free trade, yielding an overall correct classification rate of 88,57 percent (62 of 70). Large standardised parameter estimates indicate that a unit change in the independent variable will have a large effect on the log of the odds ratio of a respondent agreeing with free trade relative to other independent variables. RISKFAC has the largest effect on trade attitudes followed by LANDP and LANDRED. Computer adoption (COMPUn, a proxy for a farmer's progressiveness, has a positive coefficient as hypothesised, implying that respondents who own and use a computer are more likely to favour a liberalised trade environment. Progressive farmers are more likely to adopt relevant technologies which increase their competitive advantage. However, this variable may also caprure the effects of other factors such as farm size, farmer's age and level of education as larger, younger and better educated farmers are more likely to use a computer (Woodburn et ai, 1994). A negative relationship for DIST indicates that fanners more distant from a town will oppose free trade. In the past, 21 marketing boards reduced the need for farmers to market their products. In the event of trade liberalisation, farmers will be required to market their products (either themselves or by agents), and locality of potential markets will therefore be important. Respondents who consider changes in environmental regulations (ENVIRO) as a risk to their businesses are likely to oppose a free-trade environment, as evidenced by a negative coefficient. The South African commercial farming industry has relied on chemicals and commercial fenilizers to produce high yields (Lyne and Ortmann, 1992), whilst environmental restrictions under GAIT could include limits or bans on the use of fenilizers, pesticides and herbicides (LaFrance, 1992). Provisions under GAIT include revision and tightening of sanitary and phytosanitary measures (Giardini, 1995). In addition to these concerns, issues R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 136 SAJEMS NS Vol 1 (1998) No I relating to the landscape, conservation, water quality. quality of foodstuffs and animaI welfare are to be addressed (Maclaren, 1995). Adherence to certain environmental measures would require additional invesunent for farmers. thereby increasing their costs of production. Environmental groups are also becoming an effective counter force to traditional farm lobbies (Maclaren, 1995). A negative coefficient for RISKFAC indicates that respondents who perceive variability in crop prices, changes in the Rand exchange rate, changes in input costs and further reduction in import tariffs as sources of risk. are opposed to liberalisation of trade. This is to be expected as crop prices would be more variable under free trade. Changes in the Rand exchange rate would affect the domestic cost of inputs and prices of imported products, whilst further reduction in import tariffs on agricultural products would also affect local product prices. As hypothesised. respondents who perceive these factors as important sources of risk to their operations would oppose liberaIisation of trade. This independent variable has the highest standardised parameter estimate and thus has the largest effect on free-trade attitudes relative to other independent variables. As expected, respondents who think that land prices (LANDP) will rise in the event of the successful implementation of GAIT are supportive of a free-trade environment. GAIT is synonymous with free trade and this response is therefore rational seeing that increased land prices are of benefit to land owners. LA NDRED , the importance of further land redistribution as a source of risk, is positively related to attitudes towards free trade. Respondents' supportive of free trade probably consider themselves competitive and are likely to survive in a free-trade environment. Respondents' perceptions may indicate that more productive units will not be redistributed ahead of other less competitive operations. As expected, variability in livestock prices (LPRICE) has a negative association with the dependent variable. Operators who regard variable livestock prices as an important source of risk are likely to oppose free trade as livestock prices are likely to be more variable under free-trade conditions. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SATEB NR Voll (1998) Nr 1 137 Table 5: Logit model of attitudes towards free trade by sample farmers in KwaZulu-Natal, 1996 (0=70) • Variable Parameter Standard • Asymptotic Standardised estimate error I-statistic parameter estimate COMPUT 3,2968 1,7066 1,93' I 1,5868 DIST -0,1996 0,0732 -273'" I ' -2,7686 ENVIRO ! -1,8245 0,6970 -2,62" 1_2,8680 RISKFAC -6,4153 2,2851 -2,81'" • -6,4153 LANDP ~3'8559 2,63" 3,5825 LANDRED , 0,7684 2,63" 3,4635 LPRICE -1,0460 0,4452 -2,35" -2,1820 MFINAN 2,8454 1,0248 2,78'" 2,4708 PORF 3,8046 1,9115 1.99" 1.3373 ! REPAY -2.600 0.9552 -2.72'" -2,4226 TLEVEL -3,7165 1,7032 1-2,18" -1,5958 CONSTANT i -4,2732 2.8689 -1,49 7.2305 Model Chi- 46,163'" on 11 degrees of freedom square Goodness of fit 27,393 (N.S.) on 58 degrees of freedom Correct prediction (percent) Total 88,57 Agree 92.59 Disagree 75.00 I , " ... Note: " indicate Significance at the 10 percent, 5 percent and I percent levels of probability respectively. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 138 SAJEMS NS Vol I (1998) No 1 Respondents with a higher self-rating of management skill in farm finance (MFINAN) are supportive of a free-trade environment. Above-average financial managers are better equipped to manage iocreased price risk associated with market forces by controlling financial risk. Following economic Iiberalisation, farmers in New Zealand reduced fixed financial repayment obligations by reducing debt levels (Robinson, 1995). The positive PORF coefficient suggests that full-time farmers are in favour of a Iiberalised trade environment. This is contrary to expectations and the results reported by Edelman and Lasley (1988) and Barkley and Flinchbaugh (1990). In this study, respondents farming on a full-time basis may have a greater incentive to adapt to a rnarket-orientated economy in order to satisfy household consumption needs. These farmers may have more time available to investigate different marketing alternatives and to identify and respond to market signals. A negative relationship exists between attitudes towards free trade and debt repayment levels (REP A V). Following economic Iiberalisation in New Zealand in 1984, the real price offarmland decreased from 1095 dollars per hectare in 1984 to 569 dollars per hectare in 1989 (Johnston and Frengley, 1991). Consequently, farm households whose debt exceeded 50 percent of total assets increased from 10 pereent of households in 1985 to nearly 24 percent in 1986. This proponion of fanners with high debt levels subsequently declined due to debt restructuring and farm sales (Johnston and Frengley, 1994). Lower levels of debt are now evident as fanners seek to have a fmaocial buffer to protect themselves from a future economic downturn (Robinson, 1995). Given the potential decline in land values, operators with higher debt levels are possibly aware of the probable effect of free trade on their debt/asset ratio. The anticipated iocrease in price risk associated with trade liberalisation, in addition to the level of financial risk associated with borrowed capital, may be a matter of coocem. Respondents who indicated that they knew the level of impon tariffs on products they produced or inputs they used (fLEVEL) were opposed to free trade. Seventy-seven and 90 percent of respondents, respectively, did not know the tariff level on products they produced or inputs they used. More respondents have an expectation regarding the effect of tariff removal on reduced product prices (negative effect on farm business) than on reduced input prices (benefit to farm business), which might explain the negative relationship. This result may imply that fanners have not sought information, or need more information, on the level of tariffs on imported goods, particularly on the potential benefits (decreased costs) to the farm business of reduced (abolished) tariffs on imported inputs. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SATEB NR Vol 1 (1998) Nr 1 139 Logit model of attitudes towards deregulated domestic markets Logit analysis was also used to establish the personal and business characteristics (considered in a previous analysis) influencing farmers' attitudes towards deregulated domestic markets. The logit model adapted from equation (1) is used to predict that a respondent will strongly agree with the statement given. Only three percent of respondents disagreed with the statement regarding deregulated domestic product and input markets (last question in Table 3). Consequently, respondents who 'agreed' with the statement given were coded as zero, whilst those who 'strongly agreed' were coded as one. After investigating variables considered in the previous analysis, only variables with coefficients significant at the ten percent (or higher) level were retained in the model. These include DAIRY, EXP, MFINAN and RISK all of which are defined in Table 4. Results of the analysis are presented in Table 6. The model's Chi-square statistic is highly significant. The goodness of fit statistic shows no significant lack of fit in the overall model. The estimated model correctly classifies 96,10 percent (74 of 77) of those who 'agree' and 42,86 percent (9 of 21) of those who 'strongly agree' with market deregulation, yielding an overall correct classification rate of 84,69 percent (83 of 98). Zero-order correlation coefficients were all less than 0,19 and there were no relationships significant at the five percent level, thus multicollinearity is not suspected. Of the independent variables included, MFINAN has the highest standardised coefficient indicating the relatively large effect it has on attitudes towards market deregulation. Respondents who derive more than 55 percent of their gross income from dairy production are more likely to strongly agree with deregulation of domestic product and input markets. These (dairy) respondents possibly perceive significant benefits from the deregulation of the maize (feed) industry and improved marketing opportunities, as product differentiation and adding value to milk (eg cheese and yoghurt) may be feasible alternatives Years of farming experience (EXP) is negatively related to deregulation attitude, indicating that less experienced farmers are more likely to agree with market deregulation than more experienced producers. The latter could be less suspicious of government programmes' complex regulations (Kastens and Goodwin, 1994). Respondents who have a higher self-rating of management ability in farm finance (MFINAN, which also has the highest standardised parameter estimate), are more likely to strongly agree with market deregulation. Skilled financial R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 140 SAJEMS NS Vol I (1998) No I managers may be able to exercise prudent purchasing of inputs and should also be better able to manage increased price risk associated with deregulated markets. Willingness to take risks (RISK) was positively related to deregulation attitudes. Increased price risk associated with the deregulation of product markets is less likely to be perceived as a problem by fanners who are more willing to take risks. Table 6: Logit model of attitudes towards deregulated domestic markets by sample farmers in KwaZulu-Natal. 1996 (n=98). Variable Para- Standard Asymptotic Standardised meter error I-statistic parameter estimate estimate DAIRY 1,74943 1,1011 1,59' 0,3957 EXP -0,0641 0,0261 -2,46" -0,8653 MFINAN 1,3482 0,4602 2,93'" 1,1707 RISK 0,7865 1 0,3427 2,30" 0.7756 CONSTANT -7,5856 2,3161 -3,28'" 1-1,9576 Model Chi-square 24,600'" on 4 degrees of freedom Goodness of fit 90,266 (N.S.) on 93 degrees of freedom Con-ect prediction (percent) I Total 84,69 Agree 96,10 Strongly agree 42,86 Note: ., ", ••• indicate significance at the 10 percent. 5 percent and I percent levels of probability respectively. CONCLUSIONS Most respondents support a liberalised trade environment but foresee a reduction in product prices, farm profits and land values if GAIT provisions are successfully implemented. A possible reason for this apparent anomaly is the additional source R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SATEB NR Voll (1998) Nr 1 141 of risk associated with potential changes in government agricultural policy. Following agricultural Iiberalisation in New Zealand, farmers were faced with changes in prices and the economic envirorunent but considered these more modest and predictable than government's reactions to political demands (Robinson, 1995). Respondents are also aware of the imponance of becoming more competitive on domestic and international markets. A large proportion of respondents, however, did not know the level of import tariffs on the products they produced (77 percent) or inputs they used (90 percent). The perceived knowledge of tariff levels (TLEVEL) negatively affects attitudes towards free trade because more farmers knew about the negative effect of reduced import tariffs on their product prices than about the cost-saving effect (positive) of lower tariffs on imported inputs. More information on import tariffs and their effects could improve farmers' perceptions of free trade, particularly information on how lower tariffs affect prices of imported inputs. Information channels could include publications frequently used by farmers. more industry- specific information provided by industrial publications or the electronic media. Sixty percent of respondents intended seeking additional information about GAlT. Information regarding expected price changes and the implications of the GAlT provisions may also be useful to farmers in planning the future direction of their businesses. Seeking more marketing information to aid management decisions and adding value to their products. were cited as responses to reduced agricultural support. Production responses included increasing the size of current enterprises to take advantage of economies of size. adding a new enterprise (diversification) and changing the size and mix of current enterprises. Opportunities exist for private consUltants. who are considered by 22 percent of respondents as a potential source of information (but currently used by only 10 percent of farmers), to advise farmers in these areas, particularly in product marketing as farmers consider their skills in this area to be relatively poor. The low envisaged use of futures markets (by 10 percent of respondents) could be improved by increasing the range of futures contracts (eg more crop contracts) and/or educating farmers on the functions and principles of futures markets. This could help farmers in managing price risk as more variable prices are expected with trade Iiberalisation. Respondents who perceive sources of risk captured by ENVIRO. RISKFAC and LPRICE to be important are opposed to free trade. Farmers who are relatively risk averse (RISK) are also less likely to favour market deregUlation. High standardised parameter estimates for RISKFAC and LANDRED indicate the R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 142 SAJEMS NS Vol I (1998) No I relative importaIX:e of these risk sources on perceptions of free trade. Increased price risk is expected with free trade and market deregulation. To improve farmers' ability to survive in the changing policy environment and to avoid resistance towards trade Iiberalisation from risk-averse producers. improved information flows and education regarding methods of altering financial. marketing or production techniques to manage risk could be pursued (Barry el ai, 1995). Self-rating of managerial ability in farm finance may improve, thereby reducing debt repayment levels which would also improve perceptions of free trade. Small-scale emerging farmers are likely to have less information and knowledge of risk management strategies than large-scale commercial farmers. Consequently, government extension services may have an important role to play in disseminating information and educating small-scale producers on risk management strategies in a changing trade environment. REFERENCES 1. BARKLEY, A.P. and FLINCHBAUGH, B.L. (1990). Faml Operator Opinion and Agricultural Policy; Kansas Survey ResulJs. North Central Journal of Agricultural Economics. 12(2),223-239. 2. BARRY, P.J., ELLINGER, P.N., BAKER. C.B. AND HOPKIN, J.A. (1995). Financial Management in AgriculJure. Fifth Edition, Interstate Publishers. Inc. Danville. Illinois. USA. 3. DEPARTMENT OF AGRICULTURE (1994). The General Agreement on Tariffs and Trade: BackgrouruJ aruJ ImplicaJions for South Africa. DirectoraJe: Marketing, Pretoria. February 1994. 4. EDELMAN. M.A. and LASLEY, P. (1988). An Analysis of the Agricultural aruJ Trade Policy Preferences of Iowa Operators. North Central Journal of Agricultural Economics, 10(2). 243-254. 5. GIARDINI, G. (1995). Farm Policy Reform in the European Community: Main Features and Trade Effects. Unpublished paper, Institute of Agricultural Economics and Policy. University of Bologna, Italy. 6. GOLDIN, I. and KNUDSEN. O. (1990). Agricultural Trade liberalization: ImplicaJioTlS for Developing Countries. OrganisaJion for Economic Development. The World Bank. Washington DC, 475-485. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . SATEB NR Vol I (1998) Nr I 143 7. GUJARATI, D.N. (1995). Basic Econometrics (Third Ed.). McGraw-Hili Publishing Co., Singapore. 8. JOHNSTON, W.E. and FRENGLEY, G.A.G. (1991). The Deregulmion of New Zealand Agricu/Jure: Market Intervention (1964-84) and the Free Market Readjustment (1984-90). Western Journal of Agricultural Economics, 16(1), 132-143. 9. JOHNSTON, W.E. and FRENGLEY, G.A.G. (1994). Economic Adjustments and Changes in Financial Viability of the Fanning Sector: The New Zealand Experience. American Journal of Agricultural Economics, 76(5), 1034-1040. 10. JOSLING, T. (1993). Multilmeralism: A Constraint on Unilateralism and Regionalism in Agricultural Trade. American Journal of Agricultural Economics, 75,803-809. 11. KASTENS, T.L. and GOODWIN, B.K. (1994). An Analysis of Fanners' Policy Attitudes and Preferences for Free Trade. Journal of Agricultural and Applied Economics, 26(2), 497-505. 12. LAFRANCE, J.T., (1992). Envirorunental Policy and GATT Negotiations: Discussion. American Journal of Agricultural Economics, 74(3), 782-784. 13. LYNE, M.C. AND ORTMANN, G.F. (1992). SuggestionsforAgricultural Economics Research in South Africa. Agrekon, 31(4), 205-209. 14. LYNE, M.e. AND ORTMANN, G.F. (1996), Estimating the Potentialfor Creating Additional Livelihoods on Commercial Fannland in KwaZulu- Natal. In M. Lipton. F. Ellis and M. Lipton (Eds.), Land, Labour and Livelihoods in Rural South Africa, Vol. 2: KwaZulu-Natal and Northern Province, Ch. 3. Indicator Press, Durban. 15. MACLAREN, D. (1995). The Uruguay Round Agreement 011 Agriculture: A New World Order for Agricultural Trade? Review of Marketing and Agricultural Economics, 63(1). 51-63. 16. NIEUWOUDT, W.L. AND VINK, N. (1995). Financing of land purchase by small-scale fanners. Development Southern Africa, 12(4), 507-517. 17. ORAZEM, P.F., OTTO, D.M. AND EDELMAN, M.A. (1989). An Analysis of Fanners' Agricultural Policy Preferences. American Journal of Agricultural Economics, 71(2),837-846. 18. PHILLIPS, J. (1973). The Agricultural and Related Development of the Tugela Basin and its Influent Su"ounds. A report by Town and Regional Planning, Pietermaritzburg. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) . 144 SAJEMS NS Vol 1 (1998) No 1 19. ROBINSON. S.A. (1995). Trade Policies and their Effects 011 Fanning: A New Zealand Fanner's Experience. Proceedings: Tenth International Farm Management Congress, University of Reading, United Kingdom. 23-35. s20. SCHUH. G.E. (1995). Trade Policies and their Effects on Famling. Proceedings: Tenth International Farm Management Congress. University of Reading. United Kingdom. 9-35. 21. SWART. PJ.A. (1996). South African Agriculture in the World Economy. Agrekon. 35(4), 200-203. 22. WOODBURN, M.R .• ORTMANN. G.F. AND LEVIN, J.B. (1994). Computer Use and Factors Influencing Computer Adoption Among Commercial Fanners in Natal Province, South Africa. Computers and Electronics in Agriculture. 11. 183-194. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 00 9) .