Bio-based and Applied Economics 4(3): 201-234, 2015

ISSN 2280-6180 (print) © Firenze University Press 
ISSN 2280-6172 (online) www.fupress.com/bae

Full Research Article

DOI: 10.13128/BAE-17151

Accounting for growth in global agriculture

Keith Fuglie 

Economic Research Service, U.S. Department of Agriculture, Washington, DC

Date of submission: August 21st, 2015; accepted October 4th, 2015

Abstract. Rising prices of agricultural commodities have renewed concerns about 
constraints to agricultural productivity. To assess productivity trends, total factor pro-
ductivity (TFP) is generally preferred to partial productivity indexes as an indicator of 
technical and efficiency changes because it is more closely related to the unit costs of 
production. But measuring TFP is demanding of data, and developing comprehensive 
and comparable indexes of international agricultural TFP has been challenging. This 
study proposes a growth accounting approach, using FAO data on quantity changes 
in inputs and outputs and aggregating input changes using cost shares derived from 
other sources, as a consistent way of constructing agricultural TFP indexes for world 
agriculture. This produces aggregate growth rates for agricultural output, input and 
TFP at the country, regional and global levels. Results suggest that the rate of agri-
cultural TFP growth accelerated in recent decades, especially in developing coun-
tries. Most regions of the world now rely on productivity-based growth rather than 
resource-based growth to raise agricultural output.

Keywords. Growth accounting, technical change, total factor productivity 

JEL Codes. Q16, O13

1. Introduction

The reversal of the long-run decline in global prices for agricultural commodities that 
occurred in the first decade of the 21st Century raised concern that the rate of productiv-
ity growth in world agriculture may have slowed. If so, this would pose serious challenges 
to meeting projected future growth in the global demand for food and exacerbate envi-
ronmental degradation as more resources would be converted to produce food. In fact, 
there is evidence of a slowdown in the rate of growth in cereal grain yields (Alston et al., 
2009). However, single-factor productivity measures like crop yield mix the effects of tech-
nical change with intensified use of other inputs like fertilizer and irrigation. Since total 
factor productivity (TFP) accounts for the contributions of all inputs to production, it is a 
better indicator of technical or efficiency improvements and more closely associated with 
changes in production cost. 

* Corresponding author: kfuglie@ers.usda.gov.



202 K. Fuglie

TFP is the real output produced by a firm or industry over a period of time divided 
by the real input used by that firm or industry over the same period of time (where real 
input refers to the combined use of land, labor, capital and material resources employed 
in production). It is often difficult to provide meaningful definitions of real output or real 
input due to the heterogeneity of outputs produced and inputs used. However, it is pos-
sible to provide meaningful definitions of the output growth and input growth between any 
two periods of time using index number theory (Caves et al., 1982). Using information 
on output and input quantities and prices alone from two or more points in time, one 
can derive indexes of output and input growth, with the difference in their growth being 
defined as the growth in TFP. 

The measurement challenge for TFP growth is whether information is available to rep-
resent output and input quantities and prices at a sufficiently disaggregated level to account 
to quality differences in these measures. Most of the readily available data for measuring 
TFP growth is found in high-income countries, which is why most agricultural TFP studies 
have focused on these countries. From a series of studies compiled by Alston et al. (2010), 
it appears that the growth rate of agricultural TFP may have slowed in a number of devel-
oped countries (namely, Australia, the United Kingdom, South Africa and possibly the 
United States), lending further support to the productivity slowdown hypothesis. 

For most countries, and for the world as a whole, estimates of agricultural TFP can 
only be approximated. While there are limitations with the coverage of output data, the 
larger challenge is with the input data. While the FAO provides quantity information on 
several of the major inputs used in agricultural production, data on their costs and prices 
is mostly lacking. Arnade (1994) used Data Envelope Analysis (DEA) to get around the 
lack of input price data, solving linear programming problems to trace out a world agri-
cultural productivity frontier and determine the distance of each country to that frontier. 
In this framework, shifts in the frontier over time (defined over all or a group of coun-
tries) represents technical change, whereas the distance of a particular county to the fron-
tier represents technical inefficiency of that country. Together, changes in technology and 
efficiency sum to the rate of growth in TFP. Coelli and Rao (2005) give a summary of 
studies that have used this approach and provide updated estimates comparing agricul-
tural TFP among 93 countries over 1980-2000. However, the DEA method is sensitive to 
the “dimensionality” issue: as more or fewer countries or inputs are added to the analysis, 
results change. It is also sensitive to outliers. Further, as Coelli and Rao (2005) point out, 
the solutions to the linear programming problems provide shadow values for the inputs, 
which in many cases appear to be implausible, at least at the country level (implying, for 
example, that the marginal values of land, labor and/or other inputs are often zero). 

Another approach to dealing with the lack of input prices (or actually, lack of input 
cost shares), was proposed by Avila and Evenson (2010). They used input cost shares esti-
mated from agricultural censuses in Brazil and India to impute cost shares for other devel-
oping countries (specifically, they applied India cost shares to Asia and Africa and Brazilian 
cost shares to Latin America, making some adjustments based on relative input intensities per 
hectare). Applying these cost shares to input quantities from FAO, they estimates agricultural 
TFP growth rates for 78 developing countries for 1961-1980 and 1981-2000. Unlike many of 
the DEA models which found that agricultural TFP growth was apparently negative – even 
for countries experiencing the Green Revolution of the 1970s and 1980s – Avila and Evenson 
(2004) reported positive and accelerating TFP growth for these developing countries. 



203Accounting for growth in global agriculture

Nin-Pratt and Yu (2010) used the Avila and Evenson (2010) cost shares to set upper 
and lower bounds on the input shadow values in a DEA application (which they called 
“constrained” TFP) to generate agricultural TFP estimates for 63 developing countries over 
1967-2006. They found significant differences between estimates of constrained and uncon-
strained TFP growth and their constrained TFP growth results were closer to those of Avila 
and Evenson (2010). Findings from these studies suggest the average rate of agricultural 
TFP growth for developing countries had risen since the 1980s but there remained large 
variation among countries. This first result, of higher average TFP growth in recent dec-
ades, seems at odds with the productivity slowdown hypothesis, but these estimates typi-
cally lag at least a decade from the present and rarely have complete global coverage. 

The present study uses growth accounting to construct TFP indexes for agriculture 
world-wide. The approach is similar to what Avila and Evenson (2010) proposed, but with 
a much broader and representative set of input cost shares for constructing agricultural 
input indexes. The analysis includes industrialized, developing and transition countries for 
a near-complete global coverage. It extends my earlier work (Fuglie 2008, 2010b, 2012) 
which has steadily developed and improved upon this approach. In addition to providing 
updated evidence on agricultural TFP growth (extending the estimates through 2012), the 
principal methodological innovation in this paper is to include a comprehensive meas-
ure of animal feed in the aggregate input index. Most studies of international agricultural 
productivity have ignored animal feed as an input in agricultural production, using the 
FAO measure of net agricultural output (which subtracts from gross output the portion of 
crops kept on farms for use as feed) to net out animal feed. However, this measure does 
not account for by-products from food manufacturing and other industries that are pro-
cessed into animal feed nor crops that are imported for use as feed. The present study 
extends the approach suggested by Nin et al. (2003) to construct a more complete estimate 
of animal feed using the FAO Commodity Balance Sheets. It is expected that the revised 
estimates of aggregate agricultural inputs will be higher (and thus agricultural TFP some-
what lower) as a result of this more complete accounting of inputs used in agriculture. 

More complete coverage of inputs provides better estimates of TFP and improves our 
ability to track growth in productivity. But as the input quantities are generally not adjust-
ed for quality, these estimates should be considered as “raw TFP” (Avila and Evenson, 
2010). In particular, labor inputs are measured in terms of the number of economically 
active adults employed primarily in agriculture. Part-time employment and the improve-
ment in schooling and skills are not factored in. Many of the national TFP studies from 
industrialized countries have gone to great lengths to account for the rising quality of 
labor and other inputs in measuring the change of total inputs used in agriculture (e.g., 
see Ball, 1985; Ball et al., 1997). Controlling for changes in input quality will generally 
raise the estimate of input growth and therefore lower the residual between output and 
input growth. We should expect that the raw TFP growth estimates reported here should 
be somewhat higher than the TFP growth estimates from these national studies. 

Another point regarding the interpretation of TFP growth is that since it is estimated 
as a residual between output and input growth, it will tend to reflect not just pure techni-
cal change but also economies of scale, improvements in technical and allocative efficien-
cy, and changes in the quality of natural resources like soil, water and climate. The latter 
is especially important in light of the impact of climate change on agriculture. If climate 



204 K. Fuglie

change is negatively affecting agricultural productivity, say due to the effects of increased 
heat stress on crop yields, then estimates of TFP growth would indicate the net effect of 
the gains from technical, efficiency and scale improvements against the losses in produc-
tivity from natural resource degradation. 

A final point to keep in mind is that simply measuring TFP growth does not tell us 
anything about the causes of this growth. However, in order to have an explanation for 
TFP growth, it is first necessary to have a good measure of it. The next section of the 
paper describes in detail the methods and sources of data used to construct agricultural 
TFP indexes for each country, region and the world as a whole. This is followed by a dis-
cussion of some of the main findings on trends in global agricultural productivity growth 
during 1961-2012. The final section concludes and offers some suggestions for future 
research. 

2. Methods and Data

2.1 Measuring Total Factor Productivity Growth 

Here, I sketch out the procedures used to construct internationally comparable meas-
ures of agricultural TFP growth relying primarily on FAO data on agricultural inputs and 
outputs, and supplementary information on production costs from other studies. 

Define total factor productivity (TFP) as the ratio of total output to total inputs in 
a production process. Let total output be given by Y and total inputs by X. Then TFP is 
simply:

TFP = Y/X (1)

Changes in TFP over time are found by comparing the rate of change in total output 
with the rate of change in total input. Expressed as logarithms, changes in equation (1) 
over time can be written as:

= −
d TFP
dt

d Y
dt

d X
dt

ln( ) ln( ) ln( )  (2)

which simply states that the rate of change in TFP is the difference in the rate of change in 
aggregate output and input. 

Agriculture is a multi-output, multi-input production process, so Y and X are vec-
tors. When the underlying technology is represented by a constant-returns-to-scale Cobb-
Douglas production function and where (i) producers maximize profits so that the output 
elasticity with respect to an input equals the cost share of that input and (ii) markets are 
in long-run competitive equilibrium so that total revenue equal total cost, then equation 
(2) can be written as:

∑ ∑= −
− − −

TFP
TFP

R
Y
Y

S
X
X

ln ln lnt

t
i

i t

i ti
j

j t

j tj1

,

, 1

,

, 1
 (3)

where Ri is the revenue share of the ith output and Sj is the cost-share of the jth input. 
Total output growth is estimated by summing over the growth rates for each output com-



205Accounting for growth in global agriculture

modity weighted by its revenue share. Similarly, total input growth is found by summing 
the growth rate of each factor of production, weighted by its cost share. TFP growth is just 
the difference between the growth of total output and total input. 

One difference among growth accounting methods is whether the revenue and cost 
share weights are fixed or vary over time. Paasche and Laspeyres indexes use fixed weights 
whereas the Tornqvist-Thiel and other chained indexes use variable weights. Allowing the 
weights to vary reduces potential “index number bias.” Index number bias arises when 
producers substitute among outputs and inputs depending on their relative profitability or 
cost. In other words, the growth rates in Yi and Xj are not independent of changes Ri and 
Sj. For example, if labor wages rise relative to the cost of capital, producers are likely to 
substitute more capital for labor, thereby reducing the growth rate in labor and increasing 
it for capital. For agriculture, index number bias in productivity measurement appears to 
be more significant for inputs than outputs. Cost shares of agricultural capital and materi-
al inputs tend to rise in the process of economic development while the cost share of labor 
tends to fall. Commodity revenue shares, on the other hand, appear to show less change 
over time. To reduce potential index number bias in TFP growth estimates, cost shares are 
varied by decade whenever such information is available. For outputs, however, base year 
prices (or equivalently, base year revenue shares) are fixed, since these depend on FAO’s 
measure of constant, gross agricultural output (described in more detail below). The base 
period for output prices is 2004-2006.

A key limitation in using equation (3) for measuring agricultural productivity change 
is a lack of representative cost share data for most countries. For the present study, direct 
estimates of cost shares were assembled for 22 countries representing about two-thirds of 
world agricultural output. For another set of countries where input prices are not available 
or market-determined, (Sub-Saharan Africa and transition economies of the former Soviet 
Union and Eastern Europe) econometric estimates of production elasticities were used in 
place of cost shares. For remaining countries, representing about 25% of world agricultur-
al output, cost shares are approximated by applying cost shares from a “like” country. The 
section below on “input cost shares” provides details on the data sources and assumptions.

The framework outlined above provides a simple means of decomposing the relative 
contribution of TFP and inputs to the growth in output. Using g(Z) to signify the annual 
rate of growth in a variable, the growth in output is simply the growth in TFP plus the 
growth rates of the inputs times their respective cost shares: 

∑ ( )( ) ( )= +
=

g Y g TFP S g X  
j

J

j j
1

 (4)

I call equation (4) a cost decomposition of output growth since each Sj g(Xj) term gives 
the growth in cost from using more of the jth input to increase output.1 It is also pos-

1 Strictly speaking, input prices are held constant when estimating total input growth, so any increase in cost comes 
from using more quantity of the input and not from changes in its price. If input and/or output prices actually 
change between any two periods over which TFP growth is estimated, this would affect the distribution of the eco-
nomic gains in TFP but not the measure of TFP growth itself. For example, if output prices fell between the two 
periods, some of the gains in TFP would be passed on to consumers in the form of lower food prices. If fertilizer 
prices increased between two periods, some of the gains in TFP would be distributed as higher payments for ferti-
lizers. In competitive equilibrium, any TFP benefits that are retained by the farm sector will be capitalized into the 
price of sector-specific capital inputs (land) so as to maintain the zero profit (total cost= total revenue) condition. 



206 K. Fuglie

sible to focus on a particular input, say land (which I will designate as X1), and decom-
pose growth into the component due to expansion in this resource and the yield of this 
resource:

( ) ( )= +g Y g X g Y
X

  1
1

 (5)

This decomposition corresponds to what is commonly referred to as extensification 
(land expansion) and intensification (land yield growth). We can further decompose yield 
growth into the share due to TFP and the share due to using other inputs more intensively 
per unit of land:

∑( ) ( ) ( )= + +
=

g Y g X g TFP S g
X
X

   
j

J

j
j

1
2 1

 (6)

I call equation (6) a resource decomposition of growth since it focuses on the quantity 
change of a physical resource (land) rather than its contribution to changes in cost of pro-
duction. 

Figure 1 gives a graphical depiction of the growth decomposition described in equa-
tion (6). The height of the bars indicate the growth rate of real output. Growth in real 
output is first decomposed into growth attributable to agricultural land expansion (exten-
sification) and growth attributable to raising yield per hectare (intensification). Finally, 
yield growth itself is decomposed into input intensification (i.e., more capital, labour and 
fertilizer per hectare of land), and TFP growth, where TFP reflects the efficiency with 
which all inputs are transformed into outputs. Improvements in TFP are driven by tech-
nological change, improved technical and allocative efficiency in resource use, and scale 
economies. The decomposition of output growth into these components is both intuitively 
appealing and has some direct policy relevance: land expansion and input intensification 

Figure 1. Growth in Output, Yield and TFP.



207Accounting for growth in global agriculture

are strongly influenced by changes in resource endowments and relative prices. For exam-
ple, increasing population density or higher crop prices can induce more intensive use of 
existing farmland and investments in land improvement (Boserup, 1965). But in the short 
run, the ability to raise yield through intensification is largely confined to existing technol-
ogy. Changes in TFP, on the other hand, are driven by changes in technology and alloca-
tive efficiency. Yield growth resulting from incremental improvements to technology can 
be sustained over the long-run through investments in research and development (R&D).

2.2 Data

FAO’s 1961-2012 annual time series of crop and livestock commodity outputs and 
land, labor, livestock, farm machinery, inorganic fertilizers and animal feed inputs are the 
primary source used to construct the national, regional and global quantity measures. In 
some cases these data are modified or supplemented with data from other sources (such 
as national statistical agencies) when they are considered to be more accurate or up-to-
date, as described below. 

Output

For agricultural output, FAO publishes estimates of annual production of 198 crop 
and livestock commodities by country since 1961. FAO also aggregates production into 
a measure of the gross agricultural output using a common set of global average com-
modity prices from 2004-2006 and expresses this in constant 2005 international dollars. 
FAO excludes production of animal forages but includes crop production that is used for 
animal feed and seed in estimating gross agricultural output. The FAO also provides a 
measure of output net of domestic production used for feed and seed. However, the net 
production measure does not exclude imported grain that may be used as feed or seed, or 
grain that is exported and used in another country for these purposes. 

Because current (or near current) prices are fixed to aggregate quantities and meas-
ure changes in real output over time, the FAO gross agricultural output is equivalent to a 
Paasche quantity index. The set of common commodity prices is derived using the Geary-
Khamis method. This method determines an international price pi for each commod-
ity which is defined as an international weighted average of prices of the i-th commod-
ity in different countries, after national prices have been converted into a common cur-
rency using a purchasing power parity (PPPj) conversion rate for each j-th country. The 
weights are the quantities produced by the country. The computational scheme involves 
solving a system of simultaneous linear equations that derives both the pi prices and PPPj 
conversion factors for each commodity and country. The FAO updates these prices every 
five years and recalculates its index of gross production value back to 1961 using its most 
recent set of international prices. See Rao (1993) for a thorough description and assess-
ment of these procedures. 

I use the FAO value of gross agricultural output in constant 2005 international dol-
lars as the basis for a consistent measure of output for each country and the world. How-
ever, due to the influence of weather and other factors, agricultural production is vola-
tile from year to year, and it can be difficult to disentangle short-run fluctuations from 



208 K. Fuglie

long-term trends. To relieve the data of some of these fluctuations, I smooth the output 
series for each country using the Hodrick-Prescott filter (setting λ = 6.25 as recommended 
for annual data by Ravn and Uhlig, 2002). Even with smoothing there is still considerable 
curvature in the output series, although much of the year-to-year fluctuation in output has 
been removed from the data. I assume that the smoothed series provides a better indica-
tor of productivity trends and that annual variation around this trend is primarily due to 
short-term disturbances like weather. 

Inputs

Inputs are divided into six categories: farm labor, agricultural land, two forms of capi-
tal inputs – farm machinery and livestock, and two types of intermediate inputs – inor-
ganic fertilizers and animal feed. The primary source of information is FAO, which pub-
lished annual estimates beginning in 1961 (and for farm labor beginning in 1980) for each 
country, except for former Soviet Socialist Republics (SSRs) for which data begin in 1992. 
I extend the time series for each of the SSRs back to 1980 from Shend (1993) and further 
to 1965 using Lerman et al. (2003). 

Farm labor is the total number of adults (males and females) who are economically 
active in agriculture. FAO currently publishes farm labor estimates and projections for 
each country of the world from 1980 to 2020, although previously FAO also published 
estimates for 1961-1979. FAO estimates are used for each country except China, Nigeria 
and transition economies (former Soviet Union and Eastern Europe). Estimates are back-
cast from 1980 to 1961 using the agricultural labor force growth rates from the 2006 ver-
sion of FAO labor force statistics, which included estimates from 1961 onward. For China, 
agricultural labor estimates are from the Statistical Yearbooks of the National Bureau of 
Statistics of China. For Nigeria, labor force estimates are from Fuglie and Rada (2013), 
who determined that FAO farm labor force estimates for this country were grossly under-
counted. To derive more plausible estimates, Fuglie and Rada (2013) used FAO data (2006 
version) for 1961-1966 and then extrapolated them to the present assuming a 2% annual 
growth rate. For transition economies, national agricultural statistical sources are used, as 
reported in EUROSTAT for the Baltic countries and Eastern Europe, CISSTAT for Russia, 
Belorussia and Moldova, the International Labor Organization’s LABORSTA for Ukraine, 
and the Asian Development Bank for Asiatic former Soviet republics. Pre-1992 labor esti-
mates for these countries are from Shend (1980) and Lerman et al. (2003). 

Agricultural land is the area in permanent crops (perennials), annual crops, and per-
manent pasture. Cropland (permanent and annual crops) is further divided into rainfed 
area and area equipped for irrigation. The areas of rainfed cropland, irrigated area and 
permanent pasture are then aggregated into a quality-adjusted measure that gives greater 
weight to irrigated cropland and less weight to permanent pasture in assessing agricultural 
land changes over time (see the next section on Land Quality). However, for agricultural 
cropland in Sub-Saharan Africa total area harvested for all crops is used rather than the 
FAO cropland series (Fuglie and Rada, 2013). For China I use sown crop area (Nation-
al Bureau of Statistics of China) for cropland, given unreasonable discontinuities in the 
cropland series of both the FAO and Chinese government sources (Fan and Zhang, 2002). 
For New Zealand, FAO cropland series prior to 2002 fails to reflect changes in a consist-
ent definition over time. For cropland, I use the area in grain, seed, fodder and horticul-



209Accounting for growth in global agriculture

tural crops from Statistics New Zealand (2003) for 1961-2001, and FAO data from 2002 
onward. For similar reasons, for cropland in Indonesia prior to 1990 national estimates 
from the Badan Pusat Statistik as described in Fuglie (2010b) are used. 

Farm machinery is the total metric horse-power (CV) of major farm equipment in 
use.2 It is the aggregation of the number of 4-wheel riding tractors, 2-wheel pedestrian 
tractors, power harvester-threshers, and milking machines, expressed in “40-CV tractor-
equivalents.” The average CV per machine is assumed to be 40 CV per 4-wheel tractor, 
12 CV per 2-wheel tractor, 20 CV per power harvester-thresher, and 1 CV per milking 
machine. However, due to insufficient information no adjustment is made for differenc-
es across countries or over time in farm machinery sizes within these categories, except 
for China, which reports farm machinery inventories in power units (National Statistical 
Bureau of China). 

The FAO reports continuous time series data for 4-wheel tractors, harvest-threshers 
and milking machines, but not 2-wheel walking tractors. For many developing countries, 
particularly in Asia, 2-wheel tractors have been a major component of farm mechaniza-
tion. For 2-wheel tractors, FAO reports numbers in use for 1970s but then discontinued 
this series until recommencing it in 2002. For interim years, I collected national farm 
machinery statistics on 2-wheel tractors in use from the agricultural censuses of China, 
Japan, South Korea, Taiwan, Thailand, Philippines, Indonesia, Indian, Bangladesh, Paki-
stan, and Sri Lanka, and interpolated between census years. These countries constitute 
most of the global use of 2-wheel tractors in use on farms. 

Presently, FAO farm machinery statistics only extend to 2009 (and for many countries 
they may not extend past 2005).3 To extend estimates of farm machinery to 2012, national 
statistics on the number of tractors and combine-harvested from more recent years were 
collected for a number of countries: Bangladesh (Hassan, 2013), China (National Statisti-
cal Bureau of China, 2014), Europe (Eurostat), India (Singh et al., 2015), Japan (Ministry 
of Agriculture, Forestry and Fisheries), Russia (Russian Federation Federal State Statis-
tics Service, 2015), and the United States (National Agricultural Statistical Service, 2014). 
For remaining missing data, farm machinery stocks were extrapolated using the average 
growth rate from the three most recent years of available data. 

Livestock Capital is the aggregate value of animals used for breeding, milking, egg lay-
ing, wool production, and to provide animal traction. To approximate livestock capital, 
total inventories of animals on farms, measured in “cattle equivalents” are used.4 Inven-

2 This measure of capital stock is based on physical inventories. An alternative is to estimate capital stock as 
the sum of accumulated past investments with depreciation (perpetual inventories). Larson et al. (2000) used 
the perpetual inventory method to estimate agricultural capital stocks for 62 countries over 1967-1992. This is a 
promising effort but coverage remains incomplete. 
3 In addition to the number of farm machines in use, FAO reports the value of gross capital stock of farm 
machinery and equipment annually from 1975 to 2007. This value is computed by multiplying the number of 
4-wheel tractors, harvester-threshers, and milking machines in use by a fixed unit price and adding $35 of hand 
tools per agricultural worker. Thus, the two measures, of machines in use and the value of gross capital stock of 
farm machinery, are highly correlated. However, 2-wheel tractors are not included in the FAO estimate of gross 
capital stock. 
4 The FAO agricultural capital stock series makes a distinction between livestock “fixed assets” and livestock 
“inventories” simply by treating 85% of value of farm animals as fixed assets (breeding stock) and the rest as 
pure inventories. Since we are primarily interested in the growth rate of livestock capital, it makes no difference 
which measure is used since they are directly proportional to one another. 



210 K. Fuglie

tories include dairy cows, other cattle, water buffalo, camels, horses, other equine species 
(asses, mules, and hinnies), small ruminants (sheep and goats), pigs, and poultry species 
(chickens, ducks, and turkeys), with each species weighted by its relative size. The weights 
for aggregation are based on Hayami and Ruttan (1985, p. 450): 1.38 for camels, 1.25 for 
water buffalo, dairy cows and horses, 1.00 for other cattle and other equine species, 0.25 
for pigs, 0.13 for small ruminants, and 12.50 per 1,000 head of poultry. 

Fertilizer is the amount of major inorganic nutrients applied to agricultural land 
annually, measured as metric tons of N, P2O5, and K2O nutrients. The source of the data 
is the International Fertilizer Association, except for small countries, which is from FAO.

Animal Feed is the total amount of crop (except fodder), animal and fish products 
used for feed, measured in tonnes of dry-matter (DM) equivalents. Data on commodities 
used for animal feed are from the FAO Commodity Balance Sheets. In addition to total 
DM, total metabolizable energy (ME)5 and total crude protein (CP) of animal feeds were 
estimated.6 Parameters for the DM, CP and ME Mcal/kg (for ruminants) for each type of 
feed are from the National Research Council (1982). See Appendix Table A1 for details. 

Table 1 shows how the composition of global animal feed evolved over the three dec-
ades between 1976-1980 and 2006-2010 (reported in 5-year average annual quantities). 
There was a significant shift toward greater use of oilcrops and oilcrop meals (or cakes) 
in feed, contributing to an overall rise in the protein content of animal feeds. Over these 
three decades, the amount of CP in the global feed mix rose by 84%, while total DM 
and ME increased by 58 and 56%, respectively (DM and ME are highly correlated with 
each other). By 2006-2010, cereal grains (including processing by-products such as brans 
and distiller grains) contributed about 64% of total ME and 41% of total CP. The share 
of animal and fish products (whey, milk, meat and fish products) in global animal feeds 
declined over time. 

While these six inputs account for the major part of total agricultural input usage, 
there are a few types of inputs for which complete country-level data are lacking, name-
ly, use of chemical pesticides, seed, veterinary pharmaceuticals, energy, and services 
from farm structures. However, more detailed input data are available from several of 
the national studies from which input cost shares are derived (see section below on Input 
Cost Shares). To account for these inputs, I assume that their growth rate is correlated 
with one of the six input variables just described and include their cost with the related 
input. For instance, services from capital in farm structures as well as irrigation fees are 
included with the agricultural land cost share; the cost of chemical pesticide and seed is 
included with the fertilizer cost share; costs of veterinary medicines are included in the 
animal feed cost share, and energy costs are included in the farm machinery cost share. So 
long as the growth rates of the observed input and its unobserved counterparts are simi-
lar, then the model captures the growth of the unobserved inputs in the aggregate input 
index. 

5 Metabolizable energy is total energy of feed consumed after accounting for energy in feces, urine and gasses. 
6 For a few small countries the FAO commodity balance sheets do not report feed utilization data. For these 
countries, regional average amounts of DM, CP and ME per livestock unit are multiplied by the number of live-
stock units in that country to estimate total feed. All of these countries are estimated to use substantially less 
than one million tonnes of feed per year and the total feed use of these countries combined amounts to less than 
0.005% of global feed use. 



211Accounting for growth in global agriculture

Table 1. The Changing Composition of Global Animal Feed (average annual quantities).

Million metric tons  
(dry matter) % change

Share of total

1976-80 2006-10 1976-80 2006-10

Quantity
Cereals (grain & bran) 644 891 38.4 0.72 0.63
Oilseeds (crops, meal, oil) 93 278 199.0 0.10 0.20
Roots and tubers 49 58 19.0 0.05 0.04
Other crops 30 71 139.6 0.03 0.05
Milk, whey & butter 65 91 40.4 0.07 0.06
Meat & fish (meat, meal, oil) 15 19 23.5 0.02 0.01
TOTAL 896 1,409 57.3 1.00 1.00

Metabolizable Energy
Cereals (grain & bran) 1,878 2,613 39.2 0.72 0.64
Oilseeds (crops, meal, oil) 259 794 206.6 0.10 0.20
Roots and tubers 137 162 18.1 0.05 0.04
Other crops 67 147 120.7 0.03 0.04
Milk, whey & butter 211 278 32.0 0.08 0.07
Meat & fish (meat, meal, oil) 50 63 26.5 0.02 0.02
TOTAL 2,601 4,058 56.0 1.00 1.00

Crude Protein
Cereals (grain & bran) 73.9 100.1 35.4 0.55 0.41
Oilseeds (crops, meal, oil) 35.5 108.6 205.9 0.27 0.44
Roots and tubers 2.7 2.6 -3.8 0.02 0.01
Other crops 3.4 11.4 238.5 0.03 0.05
Milk, whey & butter 9.9 12.8 29.9 0.07 0.05
Meat & fish (meat, meal, oil) 8.0 9.7 21.4 0.06 0.04
TOTAL 133.4 245.3 83.9 1.00 1.00

Source: Feed quantities from FAO Commodity Balance Sheets; feed composition from National 
Research Council (1982).

2.3 Land Quality

The FAO agricultural database provides time-series estimates of agricultural land 
by country and categorizes this as either permanent pasture or cropland (which is fur-
ther divided in arable and permanent crop land). It also provides an estimate of area 
equipped for irrigation. The productive capacity of land among these categories and 
across countries can be very different, however. For example, some countries count vast 
expanses of semi-arid lands as permanent pastures even though these areas produce very 
limited agricultural output. Using such data for international comparisons of agricultural 
productivity can lead to serious distortions, such as significantly biasing downward the 
econometric estimates of the production elasticity of agricultural land (Peterson, 1987; 
Craig et al., 1997). 



212 K. Fuglie

In this study, because I estimate only productivity growth rather than productivity 
levels, differences in land quality across countries is less of an issue. The estimates depend 
only on changes in agricultural land and other inputs over time. However, a bias might 
arise if changes occur unevenly among land classes. For example, adding a hectare of irri-
gated land would likely make a considerably larger contribution to output growth than 
adding a hectare of rain-fed cropland or pasture. To account for the contributions to 
growth from different land types, I derive weights for irrigated cropland, rain-fed crop-
land, and permanent pastures based on their relative productivity and allow these weights 
to vary regionally. In order not to confound the land quality weights with productivity 
change itself, the weights are estimated using country-level data from the beginning of the 
period of study (i.e., using average annual data from 1961-1965). I first construct regional 
indicator variables (Regioni, i = 1,2,…5, representing developed and former Soviet bloc 
countries, Asia-Pacific, Latin America and the Caribbean, West Asia and North Africa, 
and Sub-Saharan Africa). I then regress the log of agricultural land yield in a country (its 
total output Y divided by the sum of cropland and pasture area) against the proportions of 
agricultural land in rain-fed cropland (Rainfed), irrigated cropland (Irrig), and permanent 
pasture (Pasture). Multiplying the land-use proportions by the regional indicator variables 
allows the coefficients to vary among regions:

∑α ( )
+

=
=

ln Y
Cropland Pasture

Rainfed Region *
i

i i
1

5

∑ ∑β γ( ) ( )+ +Pasture Region Irrig Region*  *  
i

i i
i

i i  (7)

The coefficient vectors α, β and γ provide the quality weights for aggregating the three 
land types into an aggregate land input index. Countries with a higher proportion of irri-
gated land are likely to have higher average land productivity, as will countries with more 
cropland relative to pasture. The estimates of the parameters in equation (7) reflect these 
differences and provide a ready means of weighting the relative qualities of these land 
classes. 

The regression estimates show that, on average, one hectare of irrigated land was 
between 1.1 to 3.0 times as productive as rainfed cropland, which in turn was 10-20 times 
as productive as permanent pasture. The results give plausible weights for aggregating 
agricultural land across broad quality classes. In fact, this approach to account for land 
quality differences among countries is similar to one developed by Peterson (1987), who 
derived land quality weights by regressing average cropland values in U.S. states against 
the share of irrigated and unirrigated cropland and long-run average rainfall. He then 
applied these regression coefficients to data from other countries to derive an internation-
al land quality index. The advantage of my model is that it is based on international rather 
than U.S. land yield data and provides results for a larger set of countries.

The effects of this land quality adjustment on global land use change are shown in Table 
2. When summed up using unadjusted data, between 1961 and 2012 total global agricul-
tural land expanded from 4,429 million ha (mHa) to 4,930 mHa, or by about 11%. When 
adjusted for quality, “effective” agricultural land expanded by 28%, or nearly three times the 
rate of growth in raw area. The reason is that irrigated area expanded much faster than oth-



213Accounting for growth in global agriculture

er types of land and when weighted for its greater productivity, it implies a much greater 
expansion in “effective” agricultural land. For the purpose of TFP calculation, accounting for 
the changes in the quality of agricultural land over time increases the growth rate in total 
agricultural inputs and commensurately reduces the estimated growth in TFP.

Table 2. Global Agricultural Land Use Changes Between 1961 and 2012 (millions of hectares).

    Developed 
Countries

Transition 
countries

Developing 
countries World

(millions of hectares)

Rainfed Cropland
1961 357 272 567 1,196
2012 309 219 734 1,262
% change -14% -19% 30% 6%

Irrigated Cropland
1961 33 11 100 145
2012 50 23 242 315
% change 51% 110% 141% 118%

Permanent Pasture
1961 885 358 1,845 3,089
2012 774 380 2,199 3,353
% change -13% 6% 19% 9%

Total Agricultural Land
1961 1,276 641 2,512 4,429
2012 1,133 622 3,175 4,930
% change -11% -3% 26% 11%

(millions of hectares of rainfed cropland-equivalents)

Quality-adjusted 
Agricultural Land

1961 505 329 884 1,718
2012 482 305 1,419 2,206
% change -4% -7% 60% 28%

Source: Agricultural land area from FAO, with adjustments made for China, Indonesia and New Zea-
land based on national data sources. Cropland includes FAO’s measure of arable land and land under 
permanent crops except for sub-Saharan Africa, where cropland equals total area harvested. Cropland 
for China is total sown area. Land quality adjustments reflect the average productivity of different land 
types relative to rainfed cropland and are derived from regression analysis (see text).

This adjustment for changes in different classes of land allows us to further refine the 
resource decomposition of output growth in equation (6) to isolate the contribution of 
irrigation apart from expansion in agricultural area to output growth. Letting X1 be the 
quality-adjusted quantity of land (and for simplicity, dropping the Region subscripts on 
the land quality parameters), then a change in X1 is given by 

ΔX1 = αΔ(Cropland) + βΔ(Pasture) + (γ-α)Δ(Irrig). (8)

The first two right-hand-side terms indicate the expansion in land area (with growth 
in pasture area adjusted for quality to put in on comparable terms with cropland expan-



214 K. Fuglie

sion). The third term isolates the contribution of irrigation expansion: (γ-α)*100% gives 
the percent augmentation to yield, holding other factors fixed, from equipping a hectare 
of cropland with supplemental irrigation. Dividing equation (8) by X1 converts the expres-
sion into percentage changes so that it shows the respective contributions of changes in 
rainfed cropland, pasture area and irrigation to output growth. Combined with equation 
(6), the resource decomposition expression shows the contributions to agricultural growth 
from expansion of agricultural land, extension of irrigation, intensification of other inputs 
per hectare, and improvements in TFP: 

∑θ α θ β θ γ α( )( ) ( ) ( ) ( ) ( )= + + − + +
=

g Y g X g X g X S g
X
X

g TFP  c c p p w w
J

J

j
j

1 1 1
2 1

 (9)

where θc, θp and θw are the shares of quality-adjusted agricultural land in crops (X1c), pas-
ture (X1p), and irrigated area (X1w), respectively (note: X1 = X1c + X1p + X1w). The first two 
terms [θcαg(X1c) + θpβg(X1p)] give the share of output growth attributable to land expan-
sion (holding yield fixed), while the third term [θw(γ-α)g(X1w)] indicates the share of out-
put growth due to the extension of irrigation (holding other inputs fixed). The fourth term 
of equation (9) gives the contribution to growth of input intensification and the last term 
the contribution of growth in total factor productivity. 

Input Cost Shares

The FAO (and supplementary) quantity data allow us to calculate the growth rates for 
six categories of production inputs (land, labor, machinery capital, livestock capital, and 
material inputs represented by fertilizer and feed), but to combine these into an aggregate 
input measure requires information on their cost shares or production elasticities. For this I 
draw upon 19 studies that have estimated nationally or regionally representative cost shares 
or production elasticities for agricultural inputs (see Appendix Table A2 for a list of these 
studies and the cost shares derived from them). These costs shares are assumed to be rep-
resentative of not only those nations but also for other countries in the same region. For 
instance, the cost shares from India were applied to other countries in South Asia, the cost 
shares for Indonesia were applied to other countries in Southeast Asia and the Pacific, the 
cost shares for Mexico were assigned to other countries in Central America and the Carib-
bean, and the cost shares for Brazil were applied to other countries in South America as well 
as the North Africa-West Asia region. These assignments were based on judgments about 
the resemblance among the agricultural sectors of these countries. Countries assigned to the 
cost shares from Brazil tended to be middle-income countries having relatively large live-
stock sectors, for example. For agricultural capital, some of these studies only reported an 
aggregate cost share for all capital services. To partition capital services into machinery and 
livestock capital services, I used the average proportions of capital stock in machinery, live-
stock and tree capital for low, middle and high income countries reported in Butzer et al. 
(2012), and assigned the cost share of capital services from trees to land. 

While the lack of direct observations on input cost shares for most countries intro-
duces uncertainty in the TFP estimation, the countries for which cost shares are observed 
represent about 65% of the global agricultural economy. This proportion rises to three-
fourths when Sub-Saharan Africa and the former Soviet Union are included – regions 
where econometrically-estimated production elasticities are used in place of cost shares. 



215Accounting for growth in global agriculture

Thus, countries to which input cost shares were imputed represent only one-quarter of 
world agricultural output. Another argument in support of this approach is that there is 
a significant degree of congruence among the cost shares reported for these country stud-
ies. For the developing countries for which cost shares data are available (India, Indone-
sia, China, Brazil and Mexico), farm-supplied inputs (land, labor, and livestock capital) 
account for between 60 and 90% of total costs, while inputs supplied by industry (machin-
ery, or fixed capital, and purchased materials such as fertilizers and processed animal 
feed), accounted for a far smaller share of resources. The cost share of inputs supplied by 
industry rises with the income of a country, and accounts for a third or more of total costs 
in the more highly industrialized countries. The use of modern inputs in transition coun-
tries, on the other hand, fell sharply after reforms were initiated in the early 1990s. These 
patterns of input use is reflected in cost shares estimated or imputed for these countries. 

Of perhaps greater concern is that some of the cost shares are becoming out of date. 
While the model attempts to adjust cost shares for each decade, it is still dependent on 
the information available from other published studies. If cost shares are unavailable for 
recent years, the model uses the last available data. In the case of China, the last nation-
ally comprehensive input cost shares for which we have estimates is for 1992. In China’s 
rapidly changing agricultural sector, we should expect that the use of costs modern indus-
trialize inputs (and their share of total cost) to continue to rise, and TFP may be overesti-
mated if the input index is not capturing the full extent of this transformation. Continued 
effort to extend and update national estimates of agricultural costs of production is neces-
sary to undertake global productivity assessments like the one described here. 

Country and Regional Productivity 

Using the methodology and data described above agricultural TFP indexes are esti-
mated for nearly every country of the world on an annual basis beginning in 1961 (and 
since 1965 for the independent states of the former Soviet Union). However, some coun-
tries have dissolved or are too small to have complete data. For the purpose of estimating 
long-run productivity trends, some national data are aggregated to create consistent politi-
cal units over time. For example, data from the nations that formerly constituted Yugosla-
via are added together to make comparisons with productivity before Yugoslavia’s dissolu-
tion. Similarly, data were aggregated for Czechoslovakia, Ethiopia and the former Soviet 
Union (TFP series for individual SSR’s begin in 1965). Because some small island nations 
have incomplete or zero values for some agricultural data, three composite territories were 
constructed by adding up available data for island states in the Lesser Antilles, Micronesia, 
and Polynesia. Altogether, the countries included in the analysis account for more than 
99.7% of FAO’s global gross agricultural output. The only areas not included in the analy-
sis that have significant agricultural production are the West Bank and Gaza.

In addition to individual countries, data are aggregated and TFP indexes estimated 
at the regional level. Input and output quantity aggregation is straight forward since they 
are all measured in the same units (although not adjusted for quality differences in the 
inputs). Regional cost shares are the weighted averages of the national cost shares for the 
countries in a region. Appendix Table A3 provides a complete list of countries included in 
the analysis and the regional groupings. 



216 K. Fuglie

3. Results: Growth in Agricultural Productivity

3.1 Sources of growth in global agriculture

Table 3 provides summary findings on productivity measures for the global agricul-
tural economy as a whole over the past five decades and for the entire 1961-2012 period. 
The first two columns of results show average annual growth rates of total agricultural 
outputs and inputs and the remaining columns indicate growth rates in four measures of 
productivity: changes in TFP, labor productivity, land productivity, and cereal grain yield 
per hectare harvested. The average growth rate of world agricultural output remained 
remarkably stable over time, rising by 2.8% per year in the 1960s and between 2.1% and 
2.5% per year in every subsequent decade. The source of output growth, however, shifted 
from being primarily input-driven to productivity-driven. Annual growth in total inputs 
fell from 2.8% in the 1960s, to between 1.5% and 1.7% in the 1970s and 1980s, and to less 
than 0.8% since 1991.

Table 3. Productivity Indicators for World Agriculture (average annual growth rate in percent).

Period Gross output Total input Total factor 
productivity

Output per 
worker

Output per 
hectare of land

Cereal yield 
per area 

harvested

1961-1970 2.79 2.79 0.00 1.13 2.44 2.88
1971-1980 2.29 1.74 0.56 1.55 2.12 2.08
1981-1990 2.10 1.49 0.62 0.59 1.76 1.88
1991-2000 2.17 0.63 1.54 1.92 2.06 1.56
2001-2012 2.52 0.84 1.68 2.83 2.59 1.53
1961-2012 2.23 1.30 0.93 1.25 2.01 1.94

Gross output: FAO gross production value in constant 2004-2006 international dollars. Total input: 
Author’s aggregation of agricultural land, labor, capital and material inputs (see text). TFP: The dif-
ference between output growth and total input growth, based on author’s estimation. Output per 
worker: FAO gross production value divided by number of persons working in agriculture. Output per 
hectare: FAO gross production value divided by total arable land and permanent pasture. Cereal yield: 
Global production of maize, rice and wheat divided by area harvested of these crops. The average 
annual growth rate in series Y is found by regressing the natural log of Y against time, i.e., the param-
eter B in ln(Y) = A + Bt.

Offsetting the declining growth in inputs to keep output growth from falling has been 
productivity. Annual TFP growth rose from a global average of 0% in the 1960s to about 
1.7% since 2001. The growth in global agricultural TFP has been generally lower than 
growth in either land or labor productivity. This reflects an intensification of capital and 
material inputs in agriculture, which raise land and labor productivity but may not affect 
TFP. Also, since the number of workers in agriculture expanded faster than agricultural 
land area, the growth rate in labor productivity tended to be lower than growth in land 
productivity. However, at the global level the agricultural labor force is now declining in 



217Accounting for growth in global agriculture

absolute terms (while agricultural land is still expanding), so the rate of labor productiv-
ity growth now exceeds the rate of land productivity growth. In the most recent period of 
2001-12, output per worker grew by 2.8% per year while output per hectare grew by 2.6% 
per year. 

The growth rate in cereal yield per area harvested, which has been used as a harbinger 
of slowing productivity growth, has actually remained fairly stable since 1990, averaging 
at least 1.5% in annual decadal growth rates. This is below the nearly 2.9% rate of yield 
growth achieved in the 1960s, but does not appear to indicate a persistent slowdown in 
yield growth. Note that land productivity (output per hectare of land in agriculture) has 
generally grown more rapidly that crop yield per area harvested. The main reason for this 
in increased land use intensity. While yield of individual crops is generally calculated on 
the basis of area harvested, land productivity is based on total output in a calendar year 
from the area designated as agricultural land. Increased land use intensity has come about 
from greater use of multiple cropping and less cropland in fallow or devoted to fodder 
crops. Globally, cropland intensity (total area harvested divided by total area designated as 
cropland) gradually increased from about 0.74 in the 1960s to 0.78 in the 1990s, but then 
grew more rapidly reaching 0.85 by 2012. 

The decomposition of global output growth into contributions from inputs and TFP 
is depicted in Figure 2. The height of each column gives the average annual rate of output 
growth by decade. Over the entire 50-year period, total inputs grew at about 60% the rate 
of output growth, implying that improvement in TFP accounted for about 40% of new 
output. However, the rate of input growth declined over time, and TFP’s contribution to 
output growth increased. By the 2001-12 period, TFP accounted for two-thirds of the 
growth in global agricultural production.

Figure 2, Panel A shows the contributions of various inputs to global agricultural 
growth according to their share of total costs (see equation 4), and the residual, or TFP. 
Increased use of material inputs, especially fertilizer, was a leading source of agricultur-
al growth in the 1960s and 1970s, when green revolution cereal crop varieties became 
widely available in developing countries (these crop varieties were more responsive to 
fertilizers, especially when grown under irrigation). Fertilizer and animal feed use also 
expanded considerably in the Soviet Union during these decades, where they were heav-
ily subsidized. The exceptionally low rate of input growth in global agriculture during 
the 1990s was due primarily to the rapid withdrawal of resources from agriculture in 
the countries of the former Soviet bloc when these countries underwent a transition 
from a centrally planned to market economies. By the early 2000s agricultural resources 
in this region had stabilized, and there was a recovery in the global rate of input growth 
compared with the 1990s. Also in the 2000s, the world’s agricultural labor force began 
to shrink for the first time in modern history. While the size of the agricultural labor 
had been falling for decades in industrialized countries, this is now also the case in Chi-
na and Latin America. South Asia may also soon enter into a period where the abso-
lute size of its farm labor force declines, if it hasn’t already (Rada, 2013). In low income 
countries (especially those in Sub-Saharan Africa), the number of persons primarily 
employed in agriculture continues to rise although the share of the labor force working 
in agriculture is falling.



218 K. Fuglie

Figure 2. Sources of Global Agricultural Growth.

Panel A: Input Cost Decomposition

Panel B: Resource Decomposition.



219Accounting for growth in global agriculture

The line shows the average annual growth rate in gross agricultural output during the 
period specified. The shaded components of the bar show the contribution of an input or 
productivity to total output growth. In Panel A, the growth rate of an input is weighted 
by its cost share. Panel B shows the growth rate in agricultural land (and the contribution 
of irrigation to raising effective land area) and the growth rate in yield, which is further 
decomposed into growth due to input intensification (inputs per area) and total factor 
productivity (TFP). 

Figure 2 Panel B shows the resource decomposition of global agricultural growth 
slightly differently. Instead of by input cost as in Panel A, it shows the relative contribu-
tion of growth in land and water (irrigation), input intensification and TFP to growth (see 
equation 9). The rate of expansion in natural resources (land and water for irrigation) has 
diminished over time while the rate of growth in resource yield has risen. The underlying 
source of resource yield gain has shifted markedly from input intensification to improve-
ment in TFP. 

As expected, the inclusion of animal feed as an explicit input in production raised the 
growth rate of inputs and reduced the growth rate of TFP. However, the changes were not 
substantial and did not affect the general pattern of TFP growth acceleration world-wide 
over the study period. Over the entire 1961-2012, TFP growth without animal feed data 
(where feed use was assumed to grow at the same rate as the size of the livestock herd) 
averaged 1.01% per year, compared with 0.94% per year with animal feed inputs meas-
ured directly. The most significant effect of including animal feed was on TFP growth 
rates estimated for the former Soviet Union (FSU), especially the pre-transition era (1961-
1991). Without including animal feed inputs, agricultural TFP in FSU remained virtual-
ly unchanged between 1961 and 1991, but with animal feed inputs it regressed by 23%. 
It would imply that during the Soviet era, agricultural growth was largely achieved by 
increasing levels of inputs but with declining marginal (and average) productivities in the 
use of those inputs. 

3.2 Sources of agricultural growth by world region

The same kind of growth analysis shown above for global agriculture can be carried 
out at the regional or country level. Figure 3 shows the contribution to agricultural growth 
from land, labor, capital, material inputs, and TFP for industrialized market economies, 
developing countries, and transition economies of the former Soviet Union and Eastern 
Europe. In industrialized economies (Panel A), the average annual rate of output growth 
fell from around 2% in the 1960s and 1970s to less than 1% in the last three decades. 
This slowdown in agricultural growth partly reflects Engel’s Law, where per capita food 
demand is satiated and the growth in food demand reflects the growth in population, 
which is declining in these countries. Labor and land (and in recent decades, capital and 
materials as well) are being withdrawn from the agricultural sector. The fact that output is 
able to continue to expand in the face of these resource withdrawals is entirely due to TFP. 
The increase in the productivity of the resources remaining in the sector has been rapid 
enough to offset the decline in the amount of resources used. The high rate of TFP growth 
enabled the agricultural sectors of these countries to remain internationally competitive, 
and developed countries as a whole were net exporters of food. 



220 K. Fuglie

Figure 3, Panel B indicates that improved productivity performance in developing 
countries was the proximate cause of the acceleration in global agricultural TFP growth 
after 1990. During the 1960s and 1970s, annual TFP growth averaged less than 1% for 
these countries, but since 1990 their agricultural TFP growth doubled to nearly 2% per 
year. For developing countries as a group, agricultural labor declined in absolute numbers 
over the 2001-12 period. This was primarily due to the exit of nearly 100 million Chinese 
farm workers to the non-farm sector. This trend is likely to accelerate in the coming dec-
ade not only in China but in other developing countries as well, as structural transforma-
tion moves workers out of agriculture. Sub-Saharan Africa, however, is expected to con-
tinue to experience growth in the size of its agricultural labor force at least through 2020, 
according to FAO projections.

During the era of central planning (pre-1991), today’s transition economies (Figure 
3 Panel C) experienced agricultural TFP regression. All agricultural growth achieved in 
this region during this period was due to resource expansion, especially the rapid growth 
in material inputs like fertilizers and feed. Inputs supplied to agricultural were often at 
highly subsidized rates. When the Soviet bloc broke apart in 1991 and these countries 
began to move toward market economies, agriculture underwent a sharp contraction as 
subsidies were withdrawn from the sector. Agricultural output growth resumed in the 
2000s, and most of this renewed growth can be attributed to improvement in TFP. More 
region-specific information on agricultural output TFP growth is provided in Table 4. 
These estimates show considerable heterogeneity in agricultural TFP growth rates among 
regions, which is even more pronounced if TFP is compared at the national level – not 
shown but available from Economic Research Service (2015). The outstanding produc-
tivity performers over the past few decades have been Brazil and China. Both are large 
agricultural producers (China has by far the largest agricultural sector in the world, 
accounting for 24% of gross agricultural output in 2013 according to FAO, and Brazil 
was the fourth largest, after India and the United States). TFP growth over the past sev-
eral decades enabled China to remain largely self-sufficient in foodstuffs during a peri-
od of rapidly rising domestic food demand (due to population and per capita income 
growth) despite virtually no new land for agriculture. For Brazil, rapid TFP growth since 
the 1980s enabled it to move from a food deficit country to become a major exporter of 
agricultural commodities. Besides these countries, Southeast Asia, South Asia, and North 
Africa have also accelerated their agricultural TFP growth, achieving an average annual 
growth rate of over 2% since 2001. 

A number of industrialized regions (Southern Europe, South Africa, Northeast Asia, 
and North America) have maintained agricultural TFP growth rates averaging at least 
1.9% per year since 2001. These estimates are generally higher than those reported by 
country studies of agricultural TFP growth in these regions (Ball et al., 2010). But recall 
that the TFP estimates reported in national studies typically make adjustments for qual-
ity changes in inputs, particularly labor, while the estimates reported in Table 4 do not. 
Adjusting an input for quality changes usually increases the share of output growth that 
is “accounted for” by growth in that input (e.g., adding one more skilled worker to the 
agricultural labor force raises output by more than adding one more unskilled worker). 
So, the TFP estimates in Table 4 should be interpreted as including not only the effects of 
technical change, but also the effects of using inputs of higher quality. 



221Accounting for growth in global agriculture

Figure 3. Sources of Agricultural Growth in Industrialized, Developing, and Transition Economies.

Panel A: Industrialized Countries.

Panel B: Developing Countries.

Panel C: Transition Economies.



222 K. Fuglie

Table 4. Agricultural Output and Total Factor Productivity Growth in Global Regions.

Region
1961-1970 1971-1980 1981-1990 1991-2000 2001-2012

Output TFP Output TFP Output TFP Output TFP Output TFP

Developing Countries 3.15 0.61 2.97 0.85 3.42 1.06 3.61 2.00 3.42 1.96
Sub-Saharan Africa 3.01 0.17 1.07 -0.12 3.17 0.90 3.21 1.11 3.25 0.60
Latin America and Caribbean 3.05 0.80 3.32 1.33 2.27 0.90 3.15 2.02 3.19 2.00

Caribbean 1.70 -0.93 1.98 0.35 0.62 -0.52 -0.51 -0.26 0.39 -0.09
Central America 4.64 3.00 3.72 1.82 1.36 -1.79 2.96 2.68 2.21 1.90
Andean 2.97 1.44 2.82 1.08 2.79 0.47 3.20 1.78 2.66 1.46
Northeast (Brazil, mainly) 3.56 0.24 3.86 1.07 3.45 2.95 3.58 2.38 4.10 3.23
Southern Cone 1.80 0.49 2.87 2.57 1.13 -0.88 3.17 1.26 2.76 0.79

Asia (except West Asia) 3.28 0.74 3.11 1.02 3.64 1.41 3.71 2.55 3.49 2.61
Northeast Asia (China, 
mainly) 4.79 0.93 3.32 0.69 4.45 1.79 5.04 3.94 3.52 3.09

Southeast Asia 2.63 0.48 3.92 1.85 3.34 0.42 2.96 1.36 4.00 2.53
Pacific 2.52 -0.04 2.34 0.21 1.58 -0.65 2.06 0.51 2.16 0.92
South Asia 2.02 0.57 2.66 0.81 3.32 1.21 2.66 1.03 3.63 2.04

West Asia-North Africa 2.87 1.33 3.02 1.52 3.56 1.35 2.79 1.47 2.49 2.13
North Africa 2.62 1.28 1.57 0.34 4.19 2.52 3.26 1.58 3.49 2.70
West Asia 2.98 1.13 3.65 2.09 3.30 0.75 2.60 1.50 2.01 1.86

Industrialized Countries 2.06 0.76 1.94 1.62 0.72 1.14 1.36 1.94 0.56 2.00
Europe, Northern 1.55 0.93 1.36 1.26 0.51 1.40 0.37 1.38 0.11 1.44
Europe, Southern 2.11 1.43 1.96 1.87 0.69 1.13 1.32 1.88 -0.40 1.92
Japan-S. Korea-Taiwan 3.52 1.67 2.45 1.95 1.19 1.34 0.06 2.06 -0.28 2.02
Australia-New Zealand 2.90 0.91 1.69 1.59 1.49 1.18 3.22 2.79 0.67 1.35
Canada-USA 2.06 0.47 2.29 1.55 0.68 1.00 1.96 1.95 1.10 1.96
South Africa 3.18 -1.12 2.55 0.95 1.21 2.97 1.54 3.01 2.55 2.62

Transition countries 3.54 -0.69 1.29 -0.56 0.80 0.25 -3.61 0.41 1.39 1.30
Eastern Europe 2.67 -0.16 1.73 0.32 -0.03 0.60 -1.33 0.04 -1.18 0.08
Russia-Ukraine-Belarus-
Moldova-Kazakhstan 0.76 -1.33 1.41 0.44 -5.43 1.00 2.35 2.39

Central Asia & Caucasus 4.71 1.93 0.55 -1.20 0.11 1.60 3.98 2.02
Baltic countries 0.93 -0.97 1.09 0.49 -6.00 -1.75 1.87 1.90

World 2.79 0.00 2.29 0.56 2.10 0.62 2.17 1.54 2.52 1.68

The average annual growth rate in series Y is found by regressing the natural log of Y against time, i.e., 
the parameter B in ln(Y) = A + Bt.

Regions experiencing persistent low growth in agricultural TFP include Sub-Saharan 
Africa, Eastern Europe transition economies, Southern Cone countries of South America, 
the Caribbean and Pacific island nations. All of these regions show a growth trend in agri-
cultural TFP of substantially less than 1%. Sub-Saharan Africa is the most critical case, 
given its large population, rapid population growth, and heavy dependence on agriculture 
as a source of livelihood. The fact that agricultural TFP growth has remained low for this 



223Accounting for growth in global agriculture

region means that it has remained poor and food insecure, and increasingly dependent on 
food imports. 

4. Conclusions

The principal advantage of a TFP measure of productivity growth is that it clearly dis-
tinguishes between resource expansion, resource substitution, and technical or efficiency 
improvements in resource utilization as sources of economic growth. Growth in TFP is 
more likely to be associated with lower unit costs of production, and, in long-run equilib-
rium, changes in market prices of output, than partial productivity indexes. The limitation 
of TFP is that it is subject to error if outputs and/or inputs are not fully or appropriately 
measured or if procedures for aggregation are biased. 

This paper seeks to move toward plausible indexes of international agricultural TFP 
by constructing a more complete accounting of the inputs employed in the sector. Spe-
cifically, the paper develops an explicit measure for animal feed inputs, something which 
most previous studies of international agricultural productivity have ignored. Animal feed 
inputs are composed of much more than the portion of crops retained on farms and fed 
to animals. It includes many by-products of food manufacturing, such as oilseed cakes, 
distiller grains, milling brans, sugar and molasses, whey, animal slaughter waste, and fish 
meal. The paper proposes three ways of aggregating these diverse feed sources into a sin-
gle quantity measure of feed input: dry-matter weight, metabolizable energy in Mcal, and 
tonnes of crude protein. It turns out that the growth in dry-matter weight and metaboliz-
able energy are highly correlated, while the growth of crude protein has been more rapid, 
implying quality improvement in the overall animal feed mix over time. One direction for 
future work could be to develop a quality-adjusted measure of feed input that combines 
energy and protein (and perhaps other nutrients). 

As expected, inclusion of animal feed in the measure of total agricultural inputs led 
to higher growth in measured agricultural inputs and thus lower growth in agricultural 
TFP for the world economy over 1961-2012, although the differences were not substan-
tial. Results did not alter the central finding from my previous analyses using this approach 
(Fuglie, 2008, 2010b, 2012) that there has apparently been a significant acceleration in global 
agricultural productivity growth since the 1990s. This is in contrast to evidence – based pri-
marily on partial productivity indexes like crop yield – that since around 1990 the rate of 
agricultural productivity has significantly slowed in most of the world (Alston et al., 2009; 
Alston and Pardey, 2014). The evidence for accelerated productivity arises from the fact that 
the real growth of agricultural output has not fallen while growth rates for most agricultural 
inputs have declined. It suggests that much of the rapid growth in crop yield and land and 
labor productivity observed in the 1960s and 1970s was due to factor substitution (especially 
fertilizer for land and capital for labor), and once the growth of other factors is taken into 
account, the real gains in efficiency during these decades were not exceptional. 

The present analysis suggests, though, that the global trend is hardly uniform. At least 
three general patterns of agricultural growth are evident:
1. In industrialized market economies, the agricultural output growth rate is slowing 

while input growth has turned negative. TFP growth offset the decline in resources to 
keep output from falling in absolute terms. 



224 K. Fuglie

2. The dissolution of the Soviet Union in 1991 imparted a major shock to agriculture 
in transition economies as they began the adjustment from centrally-planned to mar-
ket-oriented economies. In the 1990s, agricultural resources sharply contracted and 
output fell. Agricultural inputs stabilized in the early 2000s and agricultural growth 
resumed in former Soviet states but not yet in Eastern European. Agricultural TFP 
growth, which was negative during the Soviet era, turned positive following economic 
reforms.

3.  In developing countries, agricultural productivity growth doubled from around 1% 
per year during 1960-1990 to around 2% per year during 1991-2012. Two large devel-
oping countries in particular, China and Brazil, have sustained exceptionally high TFP 
growth in recent decades. Several other developing regions, including Southeast Asia, 
Central America, and North Africa, also registered accelerated TFP growth in the 
1990s and/or 2000s. Very recently, agricultural TFP growth in India has also acceler-
ated. The major exception is the developing countries of Sub-Saharan Africa, where 
long-run TFP growth has remained below 1% per year.

Despite these generally optimistic findings on agricultural productivity growth, the 
next several decades present major challenges to maintaining present rates of improve-
ment. The prospects for a general slowdown, even though it may not have yet occurred, 
are probably inevitable. One source of a slowdown, as Alston and Pardey (2014) argue, is 
if global investments in agricultural R&D are not sufficiently robust to create new produc-
tivity-enhancing technologies and offset technological obsolescence. Stagnant or declin-
ing spending on public agricultural R&D in industrialized countries, which has been a 
key source of major scientific advances for world agriculture, may put future productiv-
ity growth in agriculture at risk. Another source of a slowdown is likely to emerge from 
natural resource degradation, particularly from the warming of the climate. The effects on 
agriculture from climate change may be positive in some areas in the short term, but are 
likely to turn increasingly negative over time. Having a robust measure of agricultural TFP 
growth as outlined in this paper provides a promising means of tracking these develop-
ments and for analysing their causes and consequences.

Nonetheless, the measurement of world agricultural TFP continues to suffers from 
some serious limitations, so caution is warranted in its interpretation. Information on 
farm investments in new capital is incomplete, leading to deficiencies in the measurement 
of agricultural capital stock and capital services. Heterogeneous quality of inputs, especial-
ly land, may introduce serious measurement errors when aggregating across national and 
regional frontiers. The labor input, which is always hard to measure in agriculture where 
much of the work is done by unpaid family members, could be mismeasured if hours 
worked per capita changes, not to mention skill levels. 

A broader issue in agricultural productivity measurement is the consumption of unpaid 
(but socially valuable) environmental resources. Agriculture imposes significant costs on 
the environment in the form of greenhouse gas emissions, soil and water quality degrada-
tion, consumption of scarce and non-renewable water resources, and loss of biodiversity. 
How agricultural productivity growth affects these costs is not well understood (although 
evidence being assembled by the OECD (2014) suggests that in many cases agricultural pro-
ductivity growth is conserving of environmental as well as market resources). Insufficient 



225Accounting for growth in global agriculture

understanding of the environmental inputs and outputs associated with agricultural produc-
tion (and how to value them) represents a serious limitation to using any standard produc-
tivity index to judge questions of long-term sustainability of agricultural systems.

The important question about the usefulness of TFP is not so much whether it is 
complete as a productivity index, but rather does it convey more meaningful information 
than commonly used alternatives like crop yield? The same informational deficiencies that 
plague TFP also affect the interpretation of other available measures as indicators of the 
rate of technical change. The growth accounting approach proposed in this paper seems 
to provide additional and useful insights on the nature and sources of economic growth 
that partial productivity indexes lack. Efforts to construct TFP also point the way to what 
needs to be done to strengthen them.

Acknowledgements

The author would like to thank Ornella Maietta, Nicholas Rada and Davide Viaggi 
for their comments on an earlier draft of this paper. Nicholas Rada also contributes to 
building the ERS database on international agricultural productivity. The views expressed 
in this paper are the author’s own and should not be attributed to the U.S. Department of 
Agriculture or the Economic Research Service. 

References

Alston, J., Babcock, B. and Pardey, P. (eds.) (2010). The Shifting Patterns of Agricultural 
Production and Productivity Worldwide. Midwest Agribusiness Trade and Research 
Ames: Information Center, Iowa State University, Ames.

Alston, J., Beddow, J. and Pardey, P. (2009). Agricultural Research, Productivity, and Food 
Prices in the Long Run. Science 325(5945): 1209-1210.

Alston, J., and Pardey P. (2014). Agriculture in the Global Economy. Journal of Economic 
Perspectives 28(1): 121-146.

Arnade, C. (1994). Using Data Envelopment Analysis to Measure International Agricul-
tural Efficiency and Productivity. Technical Bulletin No. 1831. Washington, DC: 
Economic Research Service, U.S. Department of Agriculture.

Asian Development Bank. On-line Statistical Database System. Manila, The Philippines. 
https://sdbs.adb.org/sdbs/index.jsp. Accessed 16 March 2015.

Avila, A.F.D. and Evenson, R. (2010). Total Factor Productivity Growth in Agriculture: 
The role of technology capital. In Pingali, P. and Evenson, R. (eds.), Handbook of 
Agricultural Economics, Volume 4. Amsterdam: Elsevier, pp. 3769-3822.

Ball, V.E. (1985). Output, Input and Productivity Measurement in U.S. Agriculture, 1948-
79. American Journal of Agricultural Economics 67: 475-486.

Ball, V.E., Bureau, J-C, Nehring, R., and Somwaru, A. (1997). Agricultural Productivity 
Revisited. American Journal of Agricultural Economics 79: 1045-63.

Ball, V.E., Butault, J., Mesonada, C. and Mora, R. (2010). Productivity and International 
Competitiveness of Agriculture in the European Union and the United States. Agri-
cultural Economics 41: 611-627. 



226 K. Fuglie

Boserup, E. (1965). The Conditions of Agricultural Growth: The Economics of Agrarian 
Change Under Population Pressure. London: Allen & Unwin.

Butzer, R., Mundak, Y. and Larson, D. (2012). Measures of Fixed Capital in Agriculture. 
In: Fuglie K., Wang, S.L. and Ball, V.E. (eds.), Productivity Growth in Agriculture: 
An International Perspective. Wallingford, UK: CAB International, pp. 313-34.

Cahill, S., and Rich, T. (2012). A Production Account for Canadian Agriculture, 1961-
2006. In: Fuglie K., Wang, S.L. and Ball, V.E. (eds.), Productivity Growth in Agricul-
ture: An International Perspective. Wallingford, UK: CAB International, pp. 33-72.

Caves, D., Christensen, L., and Diewert, W. E. (1982). The Economic Theory of Index 
Numbers and the Measurement of Input, Output, and Productivity. Econometrica 
50: 1393-1414.

CISSTAT. CIS CD-ROM. Official Statistics of the Countries of the Commonwealth of 
Independent States Moscow: Interstate Statistical Committee of the Commonwealth 
of Independent States. 

Coelli, T. and Rao, D.S.P. (2005). Total Factor Productivity Growth in Agriculture: A 
Malmquist Index Analysis of 93 Countries, 1980-2000. Agricultural Economics 32: 
115-134. 

Craig, B., Pardey, P. and Roseboom, J. (1997). International Productivity Patterns: 
Accounting for Input Quality, Infrastructure, and Research. American Journal of 
Agricultural Economics 79: 1064-1076.

Cungu, A. and Swinnen, J. (2003). Transition and Total Factor Productivity in Agricul-
ture, 1992-1999. Working Paper 2003/2. Leuven, Belgium: Research Group on Food 
Policy, Transition and Development, Katholieke Universiteit.

Economic Research Service (2014). Agricultural Productivity in the United States. Agri-
cultural Productivity Briefing Room. Washington, DC: Economic Research Service, 
U.S. Department of Agriculture, Washington, DC. http://www.ers.usda.gov/data-
products/agricultural-productivity-in-the-us.aspx. Accessed 1 June 2015.

Economic Research Service (2015). International Agricultural Productivity. Agricultur-
al Productivity Briefing Room. Washington, DC: Economic Research Service, U.S. 
Department of Agriculture. http://www.ers.usda.gov/data-products/international-
agricultural-productivity.aspx. Accessed 15 October 2015.

EUROSTAT. Database, Statistics, Employment and Unemployment (LFS). European Com-
mission. http://epp.eurostat.ec.europa.eu/ . Accessed 3 March 2015.

Evenson, R., Pray, C. and Rosegrant, M. (1999). Agricultural Research and Productiv-
ity Growth in India. Research Report Number 109. Washington, DC: International 
Food Policy Research Institute.

Fan, S. and Zhang, X. (2002). Production and Productivity Growth in Chinese Agricul-
ture: New National and Regional Measures. Economic Development and Cultural 
Change 50: 819-38.

FAO. FAOSTAT Database. Rome: Food and Agriculture Organization of the United 
Nations. http://faostat.fao.org/ . Accessed 15 June 2015. 

Fernandez-Cornejo, J. and Shumway, C. (1997). Research and Productivity in Mexican 
Agriculture. American Journal of Agricultural Economics 79: 738-753.

Fuglie, K. (2008). Is a Slowdown in Agricultural Productivity Growth Contributing to the 
Rise in Commodity Prices? Agricultural Economics 39: 431-441. 



227Accounting for growth in global agriculture

Fuglie, K. (2010a) Sources of Growth in Indonesian Agriculture. Journal of Productivity 
Analysis 33: 225-240.

Fuglie, K. (2010b) Total Factor Productivity in the Global Agricultural Economy: Evi-
dence from FAO data. In: Alston, J., Babcock, B. and Pardey, P. (eds.), The Shifting 
Patterns of Agricultural Production and Productivity Worldwide. Ames: Midwest 
Agribusiness Trade and Research Information Center, Iowa State University, pp. 
63-95.

Fuglie, K. (2011). Agricultural Productivity in Sub-Saharan Africa. In: Lee, D.L. and 
Ndulo, M.B. (eds.), The Food and Financial Crises in Sub-Saharan Africa: Origins, 
Impacts and Policy Implications. Wallingford, UK: CAB International, pp. 122-153.

Fuglie, K. (2012). Productivity Growth and Technology Capital in the Global Agricultur-
al Economy. In: Fuglie K., Wang, S.L. and Ball, V.E. (eds.), Productivity Growth in 
Agriculture: An International Perspective. Wallingford, UK: CAB International, pp. 
335-368.

Fuglie, K., and Rada, N. (2013). Resources, Policy and Agricultural Productivity in Sub-
Saharan Africa. Economic Research Report 145. Washington, DC: Economic 
Research Service, U.S. Department of Agriculture.

Fuglie K., Wang, S.L. and Ball, V.E. eds. (2012). Productivity Growth in Agriculture: An 
International Perspective. Wallingford, UK: CAB International.

Hassan, M.S. (2013). Present Status, Prospects and Challenges of Farm Mechanization in 
Bangladesh. Beijing, China: UN ESCAP Centre for Sustainable Agricultural Mecha-
nization.

Hayami, Y. and Ruttan, V.W. (1985). Agricultural Development: An International Perspec-
tive, 2nd edition. Baltimore: Johns Hopkins University Press.

Hayami, Y., Ruttan, V.W. and Southworth, H. (eds.) (1979). Agricultural Growth in Japan, 
Taiwan, Korea and the Philippines. Honolulu: University Press of Hawaii.

International Fertilizer Association. IFADATA. Washington, DC: International Fertilizer 
Association (IFA). http://www.fertilizer.org/ifa/ifadata/search.Accessed 30 March 
2015.

Kwon, O.S. (2010). Agricultural R&D and Total Factor Productivity of Korean Agriculture. 
Korean Journal of Agricultural Economics 51(2): 67-88 (in Korean).

LABORSTA. Database, Bureau of Statistics. Geneva: International Labour Organization. 
http://laborsta.ilo.org/. Accessed 6 April 2015. 

Larson, D., Butzer, R., Mundlak, Y. and Crego, A. (2000) A Cross-Country Database for 
Sector Investment and Capital. World Bank Economic Review 14(2): 371-391.

Lerman, Z., Kislev, Y., Biton, D. and Kriss, A. (2003). Agricultural Output and Productiv-
ity in the Former Soviet Republics. Economic Development and Cultural Change 51: 
999-1018. 

Ministry of Agriculture, Forestry and Fisheries (2012). Main Indicators Relating to Agri-
culture, Forestry and Fisheries. Tokyo: Ministry of Agriculture, Forestry and Fisher-
ies.

National Agricultural Statistical Service (2014). 2012 Census of Agriculture. Washington, 
DC: U.S. Department of Agriculture, Washington, DC.

National Bureau of Statistics of China. China Statistical Yearbook. Beijing: China Statistics 
Press. http://www.stats.gov.cn/tjsj/ndsj/2014/indexeh.htm. Accessed 27 April 2015.



228 K. Fuglie

National Research Council (1982). United States-Canadian Tables of Feed Composition: 
Nutritional Data for United States and Canadian Feeds, Third Revision. Washington, 
DC: National Academy Press.

Nin, A., Arndt, C., Hertel, T., and Preckel, P. (2003). Bridging the Gap Between Partial 
and Total Factor Productivity Measures using Direction Distance Functions. Ameri-
can Journal of Agricultural and Applied Economics 85: 928-942.

Nin-Pratt, A. and Yu, B. (2012). Getting Implicit Shadow Prices Right for the Estima-
tion of the Malmquist Index: The Case of Agricultural Total Factor Productivity in 
Developing Countries. Agricultural Economics 41: 349-360.

OECD (2014). Monitoring Progress Towards Green Growth in Agriculture. Joint Working 
Party on Agriculture and the Environment. Paris, France: Organisation for Econom-
ic Cooperation and Development.

Peterson, W. (1987). International Land Quality Indexes. Staff Paper P87-10. St. Paul, MN: 
Department of Applied Economics, University of Minnesota.

Rada, N. (2013). Agricultural Growth in India: Examining the Post-Green Revolu-
tion Transition. Paper presented at the Annual Conference of the Agricultural and 
Applied Economics Association, August 4-6, 2013, Washington, DC.

Rao, D.S.P. (1993). Intercountry Comparisons of Agricultural Output and Productivity. 
FAO Economic and Social Development Paper. Rome: United Nations Food and 
Agriculture Organization.

Ravn, M. and Uhlig, H. (2002) On Adjusting the Hodrick-Prescott Filter for the Frequen-
cy of Observations. Review of Economics and Statistics 84(2): 371-376.

Russian Federation Federal State Statistics Service (2015). Russia in Figures: Agriculture. 
Moscow: Russian Federation Federal State Statistics Service.

Schimmelpfennig, D., Thirtle, C., van Zyl, J., Arnade, C. and Khatri, Y. (2000). Short and 
Long-Run Returns to Agricultural R&D in South Africa, or Will the Real Rate of 
Return Please Stand Up? Agricultural Economics 23: 1-15.

Shend, J. (1993). Agricultural Statistics of the Former USSR Republics and the Baltic 
States. Statistical Bulletin No. 863. Washington, DC: Economic Research Service, 
U.S. Department of Agriculture.

Singh, R.S., Singh, Surendra, Singh, S.P. 2015. Farm Power and Machinery Availability on 
Indian Farms. Agricultural Engineering Today 39(1): 45-56. 

Statistics New Zealand (2003). Agricultural Statistics 2002. Wellington: Statistics New Zea-
land. 

Swinnen, J., Dries, L., and Macours, K. (2005). Transition and Agricultural Labor. Agricul-
tural Economics 32: 15-34. 

Thirtle, C., Piesse, J. and Schimmelpfennig, D. (2008). Modeling the Length and Shape of 
the R&D Lag: An Application to UK Agricultural Productivity. Agricultural Econom-
ics 39: 73-85.

Van der Meer, C. and Yamada, S. (1990). Japanese Agriculture: A Comparative Economic 
Analysis. London: Routledge.

Zhao, S. Sheng, Y. and Gray, E. (2012). Productivity of the Australian Broadacre and Dairy 
Industries: Concept, Methodology and Data. In: Fuglie K., Wang, S.L. and Ball, V.E. 
(eds.), Productivity Growth in Agriculture: An International Perspective. Walling-
ford, UK: CAB International, pp. 73-108.



229Accounting for growth in global agriculture

Appendix

Table A1. Composition and Price of Animal Feedstuffs.

Feed type Price  
(2005 $/T)

Dry matter 
(%)

Crude protein  
(%)

Dietary energy 
(Mcal/kg)

Wheat 158 89 14.2 3.45
Rice (milled equivalent) 279 89 7.6 3.47
Barley 119 88 11.9 3.27
Maize 142 89 9.6 3.40
Rye 112 88 12.1 3.24
Oats 114 89 11.8 3.02
Millet 181 90 11.6 3.34
Sorghum 154 90 11.1 3.40
Cereals, other 142 89 9.6 3.40
Brans 158 89 15.2 2.74
Potatoes 30 23 2.2 0.84
Cassava 30 37 1.5 1.31
Yams 30 23 2.2 0.84
Sweet potatoes 30 37 1.5 1.31
Roots & tubers, other 30 37 1.5 1.31
Sugar cane 33 15 1.2 0.41
Sugar beet 43 11 1.5 0.38
Molasses 440 94 9.7 2.91
Sugar (raw equivalent) 440 100 0.0 4.30
Pulses (beans, peas) 556 89 22.6 3.31
Oilcrops 274 90 39.2 3.67
Vegetable oils 274 100 0.0 8.85
Soybean meal 274 90 42.9 3.37
Groundnut meal 274 93 48.1 3.39
Sunflowerseed meal 274 90 23.3 1.75
Rape & mustard meal 274 92 35.6 3.08
Cottonseed meal 274 93 37.9 2.46
Palm kernel meal 274 90 42.9 3.37
Copra meal 274 92 20.7 3.34
Sesameseed meal 274 93 45.5 3.15
Oilseed meal, other 274 90 42.9 3.37
Vegetables 188 92 21.6 2.35
Fruits 349 89 4.6 2.67
Cocoa beans 1,038 87 11.8 2.18
Meat, meat meal & offal 1,322 94 51.4 2.93
Animal fats, including butter 274 100 0.0 9.92
Whey 312 93 13.3 3.33
Milk, excluding butter 312 12 3.3 0.70
Eggs 892 94 51.4 2.93



230 K. Fuglie

Feed type Price  
(2005 $/T)

Dry matter 
(%)

Crude protein  
(%)

Dietary energy 
(Mcal/kg)

Fish 411 50 32.7 1.86
Fish meal 411 92 65.5 3.20

Oilcrops include soybean, cottonseed, groundnuts, rapeseed, sesame, sunflower, coconuts, palm ker-
nels and other.
Sources: Feed composition from National Research Council (1982); Prices are FAO global average com-
modity prices from 2004-2006, except for the following: prices for roots and tubers have been adjust-
ed downward to reflect feed quality; wheat price is used for bran; soybean price is used for all oilcrops 
and meals, vegetable oils, and animal fats; fish and fish meal price are assumed to be 1.5 times the 
price of soybean.

Table A2. Agricultural Input Cost Shares.

Source Study - country/
region and period of study Input 

Input Cost Shares Input shares 
applied to

1961-70 1971-80 1981-90 1991-00 2001-10 2011-12

Industrialized countries              

USA (1948-2011) 
 
Economic Research Service 
(2014), based on Ball (1985)

Labor 0.219 0.170 0.155 0.196 0.203 0.133

USA

Land 0.190 0.212 0.187 0.171 0.147 0.245
Livestock capital 0.111 0.114 0.115 0.102 0.091 0.077
Fixed capital 0.112 0.117 0.155 0.113 0.112 0.100
Crop materials 0.175 0.193 0.222 0.274 0.298 0.274
Livestock materials 0.192 0.194 0.166 0.143 0.150 0.170

Canada (1961-2006) 
 
Cahill and Rich (2012)

Labor 0.345 0.406 0.303 0.431 0.349 0.349

Canada

Land 0.035 0.023 0.022 0.016 0.016 0.016
Livestock capital 0.009 0.007 0.005 0.004 0.005 0.005
Fixed capital 0.146 0.147 0.162 0.087 0.085 0.085
Crop materials 0.223 0.211 0.279 0.262 0.328 0.328
Livestock materials 0.242 0.206 0.229 0.200 0.217 0.217

Australia (1978-2009) ^ 
 
Zhao et al. (2012) with  
Butzer et al. (2012) 
decomposition of total  
capital stock

Labor 0.176 0.176 0.093 0.088 0.099 0.099

Australia  
and New Zealand

Land 0.349 0.349 0.600 0.661 0.541 0.541
Livestock capital 0.182 0.182 0.110 0.085 0.136 0.136
Fixed capital 0.137 0.137 0.096 0.065 0.081 0.081
Crop materials 0.115 0.115 0.074 0.076 0.105 0.105
Livestock materials 0.041 0.041 0.026 0.025 0.039 0.039

Japan (1880-1985) 
 
Van der Meer and Yamada 
(1990)

Labor 0.388 0.351 0.313 0.313 0.313 0.313

Japan

Land 0.288 0.224 0.200 0.200 0.200 0.200
Livestock capital 0.024 0.028 0.026 0.026 0.026 0.026
Fixed capital 0.113 0.165 0.195 0.195 0.195 0.195
Crop materials 0.077 0.107 0.117 0.117 0.117 0.117
Livestock materials 0.110 0.125 0.149 0.149 0.149 0.149



231Accounting for growth in global agriculture

Source Study - country/
region and period of study Input 

Input Cost Shares Input shares 
applied to

1961-70 1971-80 1981-90 1991-00 2001-10 2011-12

Korea-Taiwan (1914-1971; 
1971-2007) 
 
1961-70 is average for Korea 
and Taiwan from Hayami et 
al. (1979); 1970+ from Kwon 
(2010) using Korea data

Labor 0.374 0.558 0.349 0.208 0.156 0.156

South Korea and 
Taiwan

Land 0.417 0.227 0.392 0.506 0.519 0.519
Livestock capital 0.020 0.004 0.009 0.010 0.012 0.012
Fixed capital 0.010 0.016 0.040 0.080 0.122 0.122
Crop materials 0.130 0.097 0.105 0.098 0.096 0.096
Livestock materials 0.049 0.097 0.105 0.098 0.096 0.096

United Kingdom (1952-2005) 
 
Thirtle, Piesse and 
Schimmelpfennig (2008)

Labor 0.327 0.164 0.136 0.137 0.137 0.137

United Kingdom

Land 0.084 0.126 0.179 0.216 0.216 0.216
Livestock capital 0.031 0.052 0.050 0.060 0.060 0.060
Fixed capital 0.183 0.199 0.202 0.204 0.204 0.204
Crop materials 0.220 0.281 0.235 0.176 0.176 0.176
Livestock materials 0.155 0.178 0.199 0.209 0.209 0.209

Europe, Northern except UK 
(1972-2002) ^ 
 
Ball et al. (2010); capital 
decomposition from  
Butzer et al. (2012)

Labor 0.339 0.339 0.251 0.243 0.229 0.229

Northern Europe 
except United 
Kingdom

Land 0.043 0.043 0.082 0.082 0.080 0.080
Livestock capital 0.020 0.020 0.026 0.017 0.014 0.014
Fixed capital 0.075 0.075 0.111 0.141 0.143 0.143
Crop materials 0.243 0.243 0.254 0.251 0.265 0.265
Livestock materials 0.280 0.280 0.276 0.265 0.268 0.268

Europe, Southern (1973-
2002) ^ 
 
Ball et al. (2010); capital 
decomposition from  
Butzer et al. (2012)

Labor 0.539 0.539 0.403 0.388 0.443 0.443

Southern Europe

Land 0.073 0.073 0.112 0.136 0.089 0.089
Livestock capital 0.019 0.019 0.022 0.016 0.012 0.012
Fixed capital 0.072 0.072 0.094 0.130 0.121 0.121
Crop materials 0.141 0.141 0.207 0.148 0.139 0.139
Livestock materials 0.155 0.155 0.161 0.182 0.195 0.195

South Africa (1947-1992) 
 
Schimmelpfennig et al. (2000)

Labor 0.232 0.210 0.166 0.161 0.161 0.161

South Africa

Land 0.129 0.143 0.169 0.144 0.144 0.144
Livestock capital 0.043 0.018 0.010 0.035 0.035 0.035
Fixed capital 0.252 0.230 0.237 0.239 0.239 0.239
Crop materials 0.246 0.279 0.275 0.274 0.274 0.274
Livestock materials 0.098 0.120 0.143 0.147 0.147 0.147

Developing countries & regions            

Sub-Saharan Africa  
(1961-2008) 
 
Fuglie (2011)

Labor 0.248 0.248 0.248 0.248 0.248 0.248

Sub Saharan Africa

Land 0.315 0.315 0.315 0.315 0.315 0.315
Livestock capital 0.308 0.308 0.308 0.308 0.308 0.308
Fixed capital 0.024 0.024 0.024 0.024 0.024 0.024
Crop materials 0.055 0.055 0.055 0.055 0.055 0.055
Livestock materials 0.049 0.049 0.049 0.049 0.049 0.049

Mexico (1960-1991) 
 
Fernandez-Cornejo and 
Shumway (1997)

Labor 0.257 0.240 0.119 0.115 0.115 0.115

Central America  
& Caribbean

Land 0.505 0.352 0.179 0.225 0.225 0.225
Livestock capital 0.089 0.161 0.315 0.263 0.263 0.263
Fixed capital 0.089 0.161 0.315 0.263 0.263 0.263
Crop materials 0.031 0.027 0.017 0.045 0.045 0.045
Livestock materials 0.029 0.059 0.056 0.090 0.090 0.090



232 K. Fuglie

Source Study - country/
region and period of study Input 

Input Cost Shares Input shares 
applied to

1961-70 1971-80 1981-90 1991-00 2001-10 2011-12

Brazil (1970, 1985, 1996, 
2006) 
 
Unpublished estimates 
provided by Nicholas Rada, 
calculated from Brazilian 
Agricultural Censuses

Labor 0.434 0.434 0.443 0.415 0.373 0.373

South America, 
North Africa and 
West Asia

Land 0.342 0.342 0.159 0.115 0.083 0.083
Livestock capital 0.096 0.096 0.090 0.070 0.053 0.053
Fixed capital 0.071 0.071 0.110 0.177 0.161 0.161
Crop materials 0.027 0.027 0.120 0.112 0.255 0.255
Livestock materials 0.030 0.030 0.078 0.111 0.076 0.076

China (1952-1992) 
 
Fan and Zhang (2002)

Labor 0.443 0.396 0.413 0.333 0.333 0.333

China, Mongolia, 
and North Korea

Land 0.250 0.209 0.178 0.258 0.258 0.258
Livestock capital 0.210 0.222 0.207 0.190 0.190 0.190
Fixed capital 0.021 0.021 0.087 0.074 0.074 0.074
Crop materials 0.038 0.064 0.084 0.121 0.121 0.121
Livestock materials 0.038 0.039 0.031 0.023 0.023 0.023

India (1956-1987; 1980-
2008) * 
 
Evenson et al. (1999);  
Rada (2013)

Labor 0.406 0.419 0.564 0.554 0.505 0.505

South Asia

Land 0.314 0.210 0.173 0.181 0.267 0.267
Livestock capital 0.213 0.269 0.123 0.115 0.052 0.052
Fixed capital 0.003 0.010 0.024 0.043 0.065 0.065
Crop materials 0.014 0.042 0.066 0.047 0.044 0.044
Livestock materials 0.050 0.050 0.050 0.060 0.067 0.067

Indonesia (1961-2006) 
 
Fuglie (2010a)

Labor 0.370 0.538 0.476 0.388 0.392 0.392

Southeast Asia  
and Pacific

Land 0.219 0.195 0.188 0.306 0.329 0.329
Livestock capital 0.327 0.166 0.221 0.160 0.120 0.120
Fixed capital 0.018 0.020 0.004 0.010 0.015 0.015
Crop materials 0.033 0.048 0.054 0.045 0.046 0.046
Livestock materials 0.033 0.033 0.057 0.091 0.098 0.098

Transition countries & regions            

USSR, European (1965-1990; 
1992-1999) 
 
Lerman et al. (2003);  
Cungu and Swinnen (2003)

Labor 0.104 0.104 0.104 0.190 0.190 0.190
European states of 
the former Soviet 
Union and formerly 
communist 
countries of eastern 
Europe

Land 0.257 0.257 0.257 0.230 0.230 0.230
Livestock capital 0.183 0.183 0.183 0.170 0.210 0.210
Fixed capital 0.043 0.043 0.043 0.090 0.090 0.090
Crop materials 0.143 0.143 0.143 0.070 0.070 0.070
Livestock materials 0.270 0.270 0.270 0.250 0.210 0.210

USSR, Asia (1965-1990; 1992-
1999) 
 
Lerman et al. (2003);  
Cungu and Swinnen (2003)

Labor 0.194 0.194 0.194 0.190 0.190 0.190

Irrigation-
dependent Asian 
states of the former 
Soviet Union

Land 0.210 0.210 0.210 0.230 0.230 0.230
Livestock capital 0.054 0.054 0.054 0.300 0.270 0.270
Fixed capital 0.113 0.113 0.113 0.090 0.090 0.090
Crop materials 0.379 0.379 0.379 0.070 0.070 0.070
Livestock materials 0.050 0.050 0.050 0.120 0.150 0.150

* Evenson et al. (1999) and Rada (2013) do not report a cost share for animal feed for India. To derive 
the feed cost share for India, I estimated total feed costs from FAO commodity balance sheets on feed 
utilization and divided this by FAO gross agricultural output (both valued at FAO international prices 
for 2004-2006). I then subtracted the feed cost share from the livestock capital cost share reported in 
the Evenson et al. (1999) and Rada (2013) studies so that the input shares sum to 1.00.
^ When studies did not report fixed capital separately from livestock capital, average capital compo-
nent shares for high-income, middle income, and low-income countries from Butzer et al. (2012) were 
used to divide total capital into these components.
Cost shares in italics are extrapolations using estimates from the nearest period available.



233Accounting for growth in global agriculture

Source: Compiled by author from sources listed. Eldon Ball, Shenggen Fan, Jorge Fernandez-Cornejo, 
Oh-Sang Kwon, Nicholas Rada, David Schimmelpfennig and Colin Thirtle kindly provided additional, 
unpublished data.

Table A3. Countries and Regional Groupings Included in the Productivity Analysis.

Sub-Saharan Africa (SSA) (developing)
Central Eastern Horn Sahel Southern Western Nigeria
Cameroon Burundi Djibouti Burk. Faso Angola Benin  
CAR Kenya Ethiopiab C. Verde Botswana Côte d’Ivoire
Congo Rwanda Somalia Chad Comoros Ghana  
Congo, DR Seychelles Sudanb Gambia Lesotho Guinea  
Eq. Guinea Tanzania Mali Madagascar G. Bissau  
Gabon Uganda Mauritania Malawi Liberia  
Sao Tome Niger Mauritius Sierra Leone  
 & Principe Senegal Mozambique Togo  
  Namibia  
  Swaziland  
  Zambia  
        Zimbabwe    

Latin America & Caribbean (LAC) (developing)  N. America Africa
Northeast Andes S. Cone C. America Caribbean Developed Developed
Brazil Bolivia Argentina Belize Bahamas Canada South Africa
Fr. Guiana Colombia Chile Costa Rica Cuba USA  
Guyana Ecuador Paraguay El Salvador Dom. Rep.    
Suriname Peru Uruguay Guatemala Haiti    
  Venezuela Honduras Jamaica    
  Mexico L. Antillesa    
  Nicaragua Puerto Rico    
      Panama Trin. & Tob.    

Asia-Pacific West Asia & North Africa
Developed South Asia SE Asia Pacific NE Asia West Asia North Africa
Japan Afghanistan Brunei Fiji China Bahrain Algeria
Korea, Rep. Bhutan Cambodia Micronesiaa Korea, DPR Iran Egypt
Taiwan Nepal Indonesia N. Caledonia Mongolia Iraq Libya
Singapore Sri Lanka Laos PNG   Israel Morocco
  Bangladesh Malaysia Polynesiaa   Jordan Tunisia
  India Myanmar Solomon Is.   Kuwait
  Pakistan Philippines Vanuatu   Lebanon Syria
  Thailand   Oman Turkey
  Viet Nam   Qatar UAR
          S. Arabia Yemen



234 K. Fuglie

Oceania Europe Developed Europe Former Soviet Union (transition)

Developed Northern Southern Transition Baltic E. Europe CAC
Australia Austria Cyprus Albania Estonia Belarus Armenia
N. Zealand Belgium-Lux. Greece Bulgaria Latvia Kazakhstan Azerbaijan
  Denmark Italy Czecho- Lithuania Moldova Georgia
  Finland Malta  slovakiab   Russia Kyrgyzstan
  France Portugal Hungary   Ukraine Tajikistan
  Germany Spain Poland   Turkmenistan
  Iceland   Romania   Uzbekistan
  Ireland Sweden Yugoslaviab    
  Netherlands Switzerland      
  Norway UK        

CAC = Central Asia & Caucasia.
a Composite territories composed of several small island nations.
b Statistics from the successor states of Ethiopia (Ethiopia and Eritrea), Sudan (Sudan and South 
Sudan), Czechoslovakia (Czech and Slovak Republics), and Yugoslavia (Slovenia, Croatia, Bosnia, Mac-
edonia, Serbia and Montenegro) were merged to form continuous geographical coverage since 1961.