WP5 1 59 G
POLICY RESEARCH WORKING PAPER      1596
Agrarian Structure in Poland                                       PolicyrnakersvAio:rre t
promoto large, m. ,har z( 1
fairms ir Polanc tc the
The  Myth  of Large-Farm   Superiority                             cetrlment of sr-alf farn,
should reevalu-Ae Jhes(
Jo ban van Zy                                                      policies. Smaller f Ts <.-e
Bill R Miller                                                      more JaborAntuis:.e th. n
Andrew Parker                                                      large fafms, ar,i r les
efficienr
The World Bank
Agriculture and Natural Resources Department
Sector Policy and Water Resources Division
April 1996



| POLICY RESEARCH WORKING PAPER 1596
Summary findings
In Poland, present policies are aimed at promoting large,       between small and large farms. Then they used Data
mechanized farms over smaller family farms. These               Envelope Analysis to estimate scale efficiencies.
policies are based on the perception that large farms             The results show that, for the sample of farms
offer real economies of scale. But international evidence       analyzed:
indicates that such large, mechanized farms are generally          * Large farms are not more efficient than smaller
less efficient and use less labor than small family farms.      farms.
So, van Zyl, Miller, and Parker analyzed the                     * Smaller farms are more labor-intensive than larger
relationship between farm size and efficiency in Polish         farms.
agriculture.                                                      These results have important policy implications for
They used two different methods to do so. First, they         farm restructuring in Poland and other transition
determined differences in total factor productivity             economies facing similar issues and conditions.
This paper-a product of the Sector Policy and Water Resources Division, Agriculture and Natural Resources Department
- is part of a larger effort in the department to investigate appropriate policies to foster agricultural growth in transition
economies. Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please
contact Melissa Williams, room S8-222, telephone 202-458-7297, fax 202-522-1142, Internet address
mwilliams4@worldbank.org. April 1996. (47 pages)
The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about
development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The
papers carry the names of the authors and should be used and cited accordingly. The findings, interpretations, and conclusions are the
authors' ouw and should not be attributed to the World Bank, its Executive Board of Directors, or any of its member countries.
Produced by the Policy Research Dissemination Center



AGRARIAN STRUCTURE IN POLAND:
THE MYTH OF LARGE FARM SUPERIORITY1
Johan van Zyl
(World Bank)
Bill R. Miller
(University of Georgia)
Andrew N. Parker
(World Bank)
The issue of what is the most appropriate farm size to maximize agricultural
productivity has an extensiue literature. In Poland, present policies are aimed at
promoting large-scale, mechanized farm8 over smaller family farms. These policies
are based on the perception that there are real economies of scale present that favor
large farms. But these perceptions are contrary to international evidence, which
indicates that a large-scale, mechanized farm sector is generally inefficient and uses
less labor than sma<l-scale family farms. Thus, there is ample motivation for a
detailed analysis of the farm size-efficiency relationship in Polish agriculture.
Two different methodologies are used to determine the farm size-efficiency relationship
in two regions of Polan& first, total factor productivity (1FP) differences between small
and large farm8 are determined; and second, non-parametric Data Envelope Analysis
(DEA) is used to estimate scale efficiencies.
Despite a history of policies favoring relatively larger, mechanized farms in Poland,
the results show that, within the sample of farms analyzed, (i) large farms are not
more efficient than smaller farms; and (ii) smaller farms are relatively more labor
intensive in their mode of prod uction. These results have important policy
implications with respect to farm restructuring for Poland and other transition
economies facing similar issues and conditions.
1     The authors wish to thianlk Alija Dokurno, Edmund Lorencowicz, Henryk Luejan,
Jacek Mizerka, Jan Wojciechowski for their assistance. Ham Binswanger, Jean-Paul Chavas,
Paula Despins, Richard Lacroix, Jenny Piesse and Colin Thirtle also provided useful
contributions for which the usual dislaimers apply.






International evidence indicates that a large-scale, mechanized farm sector ia
generally less efficient, especially when compared to small-scale family farms.
Although there may exist real economies of scale, they are mostly 'false' because
they are usually the result of policies that favor larger farms over small farms. The
relationship between productivity and farm size in Polish agriculture has not been
adequately investigated. The issues of whether larger, mechanized farms are more
economically efficient than smaller farms has not been analyzed in detail. Yet
results of such an analysis could yield important findings for agricultural policy in
Poland.
This paper, therefore, seeks to address the farm size-productivity relationship in
Polish agriculture, first, by considering international empirical evidence on the farm
size-efficiency relationship in agriculture. This is followed by a discussion of Polish
farm structure and policies of farm restructuring. Third, an analysis of the farm
size-efficiency relationship for two agricultural regions of Poland is undertaken
using cross-section farm-level data. Both total factor productivity comparisons
between farm size categories and non-parametric Data Envelope Analysis are used
in the analysis. The paper concludes with implications of these findings for farm
restructuring in Poland.
International Experience On Economies Of Scale, Farm Size and
Productivity
Larger farms are often actively promoted over small farms because it is argued that
significant economies of scale exist in agriculture. However, the general consensue
2



in the literature is that economies of scale do not exist in agriculture, except under
specific, usually temporary, circumstances. Empirical evidence demonstrates that
family-type farms are generally more efficient and superior to other types of
farming because of the way in which labor relations are organized. Patterns of
resource utilization indicate that smaller farms use labor and land with greater
intensity than large farms. These findings suggest that agricultural policies should
not favor one form of agricultural production over another.
Economies of Scale in Agriculture
Most of the arguments for promoting larger farms over smaller ones rely on the
assumption that there are significant economies of scale in agriculture. The sources
of economies of scale are said to include: (i) lumpy inputs, such as farm machinery
and management skills, that can only be utilized on farms above a minimum size;
(ii) advantages in the credit market and in risk diffusion arising from collateral
associated with the ownership of large holdings; and (iii) processing plants that can
realize economies of scale, usually through wage plantations (Binswanger et al,
1995; Johnson and Ruttan, 1994).
However, many options are available to help smaller farms circumvent the negative
effects of economies of scale: (i) machine rental can permit small farmers to mitigate
the economies of scale advantage in all but the most time-bound of operations, such
as plowing and planting in dry climates, or harvesting where climatic risks are
high; (ii) management and technical skills can be contracted from specialized
consultants and advisory services, or can be provided by publicly financed extension
services; (iii) small farmers may be able to access credit through cooperatives where
3



many farmers may pool their land for use as collateraL and various sharecropping
arrangement, where the credit risks are shared by farmer and landowner; and (iv) the
coordination problem associated with processing in plantation crops can be solved at
a relatively low cost through contract farming.
Although lumpy inputs, credit and risk diffusion, and processing plants can,
therefore, be important temporary sources of economies of scale, market solutions
are generally available that can mitigate the effects of the scale economies. In any
case, the concept of economies of scale is problematic in practice. Theoretically,
economies of scale are defined by a production function that exhibits a more than
proportional increase in output for a given increase in magnitude of all inputs.
However, there is rarely a situation when an increase in magnitude of some inputs
does not imply a change in the utilization of other factors of production (Peterson
and Kislev, 1991).
The general consensus of researchers who have empirically investigated it is that
economies of scale do not exist in agriculture, except under specific circumstances
(Johnson and Ruttan, 1994; Peterson and Kislev, 1991). Chavas and Aliber (1993),
for example, found virtually no scale economies in dairy production in Wisconsin,
and the limited initial scale economies they observed were attributable to lumpiness
of certain inputs.
Public policies that focus resources on supporting large over small farms are
therefore unlikely to be successful in terms of realizing significant scale economies.
A more appropriate policy is one that facilitates the formation of markets to enable
4



small farmers to gain access to essential productive resources and to create a 'level
playing-field" for all agricultural producers.
Farm Size and Productivity
The apparent existence of economies of scale has led researchers to investigate more
closely the relative performance of small and large farms. Over the past three
decades, various studies have assessed the productivities of farms of different sizes.
In general, these studies demonstrate a higher level of factor productivity on small
farms than on large ones; however, the studies often suffered from methodological
shortcomings that have raised questions about their validity. More recent studies
have attempted to address these issues.
Based on Indian farm management data from the late 1950s, Sen (1962) observed
an inverse relationship between farm size and yields per unit area. Since then,
many articles have been written that have examined this relationship. Berry and
Cline (1979) carried out a detailed empirical investigation into the relationship
between farm size and productivity covering data from Brazil, Colombia, the
Philippines, (West) Pakistan, India, and Malaysia. They conclude that 'the evidence
presented... points to systematically higher land productivity on small farms than on
large ones, and to total factor productivities that are at least comparable" (Berry
and Cline, 1979: 4).
Other authors who found that an inverse relationship exists have included: Cornia
(1986), who found a strong negative correlation between farm size and factor inputs
and yields per hectare in 12 out of 15 developing countries for which he analyses
5



farm management data from FAQ; Bardhan (1973), who found that the elasticity of
output per acre with respect to acreage in selected Indian villages is mostly
negative; Khusro (1968), who similarly found that if acreage, uncorrected for its
level of fertility is taken, then as acreage increased, gross output per acre decreased;
Carter (1984), who concluded that the inverse relationship is due to specific
characteristics of small farms; and Kanel (1967), who discovered that it is not only
output per acre, but net return per manday, and the ratio of values of outputs to
inputs that were negatively correlated with farm size.
In Kenya, a World Bank study (World Bank, 1983) that investigated the higher
efficiency of small versus large farms, found that output per hectare was 19 times
higher and employment per hectare was 30 times higher on holdings under 0.5
hectare than on holdings over 8 hectares. At the national level, this meant that a 10
percent reduction in average farm size would increase output by 7 percent and
employment by over 8 percent. Binswanger et al (1995)report similar results for
many other countries. Evidence is also available at the macro-level, but only in
terms of physical yields-an imperfect indicator of efficiency. Prosterman and
Riedinger (1987) using data from 117 countries, show that 11 of the top 14 countries
in terms of grain yields per hectare are countries in which small-scale, family
farming is the dominant mode of production.
Based on the observation that there are no significant economies of scale in
agriculture, most studies that consider the relationship between farm size and
productivity assume that returns to scale in agriculture are constant (Barnum and
Squire 1978; Berry and Cline 1979; and Johnson and Ruttan 1988). In the absence
6



of decreasing returns to scale some other explanation of the inverse farm size-
productivity relationship must be sought. Berry and Cline argue that "if returns to
scale for inputs actually utilized are constant, then the crucial determinant of the
farm size-productivity relationship becomes the behavioral pattern of resource
utilization by farm size" (Berry and Cline, 1979: 7; italics added).
Resource Use and Farm Size
Closer analysis of the pattern of resource utilization by farms indicates that smaller
farms use factors of production with greater intensity than large farms. On the one
hand, it is observed that the apparent benefits large farms enjoy over small farms
with respect to land and capital are due mainly to market imperfections. These
imperfections can be addressed through the promotion of markets to serve small
farms. On the other hand, it is observed that small farms use labor resources more
intensively than larger farms. For example, Cornia (1985) finds a negative elasticity
for resource inputs with respect to land, and higher land use intensity on smaller
farms. For land, it is argued that large farms underutilize the total land area at
their disposal by, for example, having an output composition that is more geared to
land extensive enterprises (e.g., cattle or other livestock projects), while smaller
farms practice multi-cropping, which increases land utilization and raises the total
value of output for a given area of land.
Under conditions of perfect competition, we would expect these differences to be
equalized over time. The fact that they continue to persist many years after being
identified suggests that there are underlying economic factors that go beyond
7



essentially technical, descriptive explanations. The basis for this economic analysis
revolves around the existence of imperfections in the markets for factors of
production that result in small farms facing a lower effective price for labor and
higher effective prices for land and capital. This leads to small farms using more of
the relatively abundant factor-labor, and less of the relatively scarce factors-land
and capital, whereas large farms treat land as an abundant resource, and substitute
capital for labor.
Land. Larger farms face an implicit cost for land that is lower than its opportunity
cost. This is due to: (i) the existence of a limited land market in many developing
countries, where the ownership of land changes only slowly over time, and in some
areas where the market for land is virtually nonexistent; (ii) large landowners
sometimes being absent from their land, so having little direct relationship with the
land, and as a consequence undervaluing its merits as a productive resource; (iii)
large farmers having better access to credit, so they can buy land more easily at
relatively low, sometimes negative, real rates of interest; (iv) "distress" sales that
enable large landowners to purchase land from small farmers at lower prices, where
land prices do not fully reflect the risks of parting with land for small farmers, who
if they sell must seek some risky alternative; and (v) land being held as an asset or
for prestige purposes that will tend to reduce its productivity. Barnum and Squire
(1978: 198) present results from their analysis that show land is consistently
underused by large farms, and they "attribute this result to imperfections in the
land market, especially the tendency to treat land as a non-economic factor in intra-
family transactions".
8



Capital. Larger farms also face an implicit cost for capital that is often lower than
its opportunity cost. They are often able to gain easier access to official subsidized
credit schemes, and obtain capital at cheap rates of interest. These imperfections
are often exacerbated by macroeconomic policies that maintain a strong exchange
rate, meaning that large farmers can import capital equipment at effectively
subsidized prices. Small farmers, by contrast, lack access to formal credit markets,
and depend on informal markets that have high real rates of interest. Part of
Feder's (1985) model relates to capital market imperfections, where obtaining
working capital depends on the farm size, i.e., on the collateral offered. His model
shows that under these conditions a systematic relationship between farm size and
productivity exists. This is confirmed by the research of Carter and Kalfayan (1989)
and Carter (1994).
Labor. Small farms commit more labor to production than large farms. This is due
to the existence of a dual labor market. On small, family labor-based farms, each
member receives a share of the total value product of the farm. This means that an
individual's wage will be less than the marginal product, and closer to the average
product. Consequently, an individual on a family farm will supply more labor at a
lower marginal value product (with a lower implicit wage rate), than would be the
case if the marginal value product of labor were being equated to the market wage.
In fact, even if marginal value product equalization were to take place, family
workers would discount the market wage due to the uncertainty of finding work.
None of these situations apply in the large farm sector, where the marginal value
9



product of labor will be at least equal to, and in the case of monopsony power,
greater than the market wage rate (Chattopadhyay and Rudha 1976).
The literature demonstrates that family farms are generally more efficient and
superior to other types of farming because of the way in which labor relations are
organized (Berry and Cline, 1979; Binswanger et al, 1995; Deininger, 1993; Johnson
and Ruttan, 1994; Lipton, 1993). Family farms are by definition farms where the
owner is the operator and where the family provides the large bulk of the regular
labor requirements throughout the year. While the definition of family farms does
not exclude the hiring of other people, especially in a part-time capacity when
related to seasonal labor, they tend not to rely too much on such activity and can
avoid moral hazard problems because family farm members typically work side-by-
side with hired labor and can monitor work effort. Hired labor on large farms is far
harder to monitor and may result in shirking. The use of less labor as farm size
increases is well documented (e.g. Ghose, 1979; Van Zyl et al, 1995).
Also underlying the establishment and maintenance of large-scale farmrs is the
misguided perception that there is a relationship between modernization, specifically
mechanization, and large farms. This has been clarified in the literature (Johnson and
Ruttan, 1994). Capital intensity is explained by the substitution of capital for labor
because of high wages. This substitution process, brought about by changes in relative
factor prices (Peterson and Kislev, 1991), indirectly encouraged the formation of
larger farms. It is argued that machinery permits farmers to work progressively
larger units of land (Hayami and Ruttan, 1985). However, a closer examination of the
application of capital and machinery in agriculture suggests that although the
10



number of workers can be increased substantially without increasing labor
supervision costs with stationary machinery, in agriculture, labor and machines are
both mobile, making supervision expensive and increasing management costs. Also,
agricultural tasks are sequential in nature due to the annual cycle of production; this
limits the opportunities for specialization and division of labor, which creates few
advantages to expansion beyond the size of owner-operator (Brewster, 1950).
Criticsms of the Findings
Although there appears to be strong empirical evidence supporting the existence of
a farm size-productivity relationship, the basis for this finding has been criticized.
Barbier (1984) has criticized the use of pooled data. He finds no set relationship in
Hoogly district in West Bengal, and concludes that 'there does not exist one form of
relationship but multiple forms -inverse or not -holding for different places in
India, along different size distributions and for different types of crops". Carter
(1984) has also pointed out the possibility of multicollinearity among the
independent variables, e.g., soil quality and farm size.
Patnaik (1972) has criticized the assumption that scale and size are closely and
positively correlated. Using data from the Farm Management Survey in India, she
shows that inputs and- outputs per acre both decline as farm size increases, but the
opposite is true when scale is considered. The importance of this finding relates
mainly to issues of policy, as it can be seen from her analysis that size in itself is not
an adequate measure on which to base agricultural policy. Closely related to scale is
the issue of technical change. Deolalikar (1981), using cross-section regional data
11



from agricultural censuses in India, shows that with technical change the inverse
relationship breaks down. He points out that much of the empirical work on farm
size dates from before the green revolution. In West Godavari, for example, in 1958-
60 there was an inverse relationship between size and output per acre, but by 1969-
70 this relationship had become positive, which he suggests is due to more intensive
use of fertilizer. He argues that although labor market dualism is important, with
technical change the share of labor in total output will decline, and other current
inputs, e.g., fertilizer, will become more important.
Other writers (Binswanger et al, 1995) have argued that many of the studies on
farm size suffer from analytical shortcomings. For example, studies often do not
account for differences in land quality or labor productivity; they use physical
yields; and they do not account for differences in size and ownership holding size.
More accurate measures of efficiency are based on the difference in total factor
productivity between small and large farms, and the difference in profits per unit of
capital invested, net of the cost of family labor. However, if these more appropriate
productivity measures are used, the earlier findings are reinforced rather than
contradicted. Binswanger et al (1995) find that: (i) the productivity differential
favoring small farms over large farms increases with the differences in size, implying
that it is largest where inequalities in landholdings are the greatest, in the
relatively land-abundant countries of Latin America and Africa, and smallest in
land-scarce Asian countries where farm size distributions are less equal; and (ii) the
highest output per unit area is often achieved not by the smallest farm size category
12



but by the second smallest farm size class, suggesting that the smallest farms may
suffer from diseconomies of scale, e.g., due to severe credit constraints.
Any observed positive relationship between farm size and productivity is the result
of market imperfections, and then only when more than a single market is
imperfect. For example, if credit is rationed according to farm size, but all other
markets are perfect, land and labor market transactions will produce a farm
structure that equalizes yields across farms of different operational size. But if
there are imperfections in two markets, land rental and insurance, or credit and
labor, a systematic relationship arises between farm size and productivity
(Binswanger et al, 1995).
In countries, such as Poland, where markets facing small farmers for any
combination of factors of production (e.g., labor, land, credit, land rental, insurance)
are often imperfect or missing-at least for some farmers and in general those who
are small, this may give rise to apparent economies of scale over the short term.
However, these economies of scale are 'false" in the sense that they are only
temporary, and the result of deliberate elimination of, or restrictions on factor
markets.' Where market development is promoted, economies of scale diminish and
eventually disappear. The issue, thus, is not to pursue a farm structure that over
the short term captures these benefits, but over the longer term locks a country into
an inefficient and inequitable structure centering on large-scale, mechanized farms,
but rather to encourage more competitive markets. Over-emphasizing
modernization, restructuring, mechanization and other similar concepts that imply
13



the use of more capital to labor than that dictated by economic realities should be
discouraged.
Structure and Size in Polish Agriculture
Polish farm structure is characterized by three distinct sectors: state farms,
cooperatives and private farms (Table 1).
Table 1: Agricultural Landuse by Major Sector
(Thousand hectares, with proportion of total in brackets)
Sector                 1985         1990          1992
Private farms      14,425 (77%)  14,228 (76%)  14,264 (77%)
State farms         3,509 (19%)   3,490 (20%)   3,304 (19%)
Co-operative farms   770 (40%)     650(4%)      678 (4%)
Total                 18,844       18,720        18,644
Source. Central Statistical Office (CSO, 1993)
State Farms
The concentration of state farms varies substantially between regions. It is higher
in the west and north (40 percent of all farmland) and lower in the central and
southern regions (10-15 percent). State farms are generally very large-5,000 to
6,000 hectares is not uncommon-and tend to be horizontally integrated in groups
of several large farms, as well as vertically integrated with feed mills, seed farms,
and poultry processing units (World Bank, 1990). More than half of state farms are
over 1,000 hectares in size. In 1988, state farms produced 19 percent of farm output
and marketed more than 50 percent of the total production for some crops, e.g.,
14



wheat and other grains (CSO, 1988). The level of mechanization on state farms is
higher than on private farms. Although the average number of tractors per area was
similar at 6.5/100 hectares in 1991 (Gomula et al, 1991), tractors on state farms
generally were much more powerful. State farms' contribution to rural employment
is, however, relatively lower than private farms. While accounting for roughly 19
percent of the agricultural land, they employed only 10 percent of the agricultural
workers in 1992 (Wojciechowski, 1994).
At the start of the economic program of transition from a centrally planned to a
market economy during the late 1980s, the agricultural sector in Poland was highly
protected and subsidized. Subsidies to producers in the form of price support ranged
from 3.4 to 4.8 percent of GDP in the period 1986 to 1989. This was supplemented
by input subsidies of 1.3 percent of GDP (Wojciechowski, 1994). The state farm
sector, however, benefited disproportionately from the subsidies (Table 2).
Table 2: Percentage Producer Subsidy Equivalents, by Sector and
Commodity, 1987 (percent)
Sector   Wheat  Rye   Barley  Beet  Seed   Pork  Beef  Milk
State     67.4   69.3   75.2   49.7  69.0   36.6   3.9   41.1
Private    47.1   48.1   58.7   31.0  43.5   28.1   3.9   22.7
Source: Foreign Trade Institute (1989)
State expenditures were reduced dramatically and farm subsidies were abruptly
terminated when the state experienced a fiscal crisis with the collapse of central
planning. Subsidies for price support to farmers declined from more than 4.0 to 0.2
percent of GDP from 1989 to 1990. Similarly, agricultural input subsidies fell from
16



1.3 to 0.3 percent of GDP. State farms, the major beneficiaries of agricultural
subsidies, were thus left without support and had to adjust to the new conditions of
the market (Wojciechowski, 1994).
This situation, as well as a deterioration of overall economic conditions, e.g., high
inflation, the limited availability of capital and increased costs of inputs, resulted in
many state farms being unable to meet their financial commitments. The level of
inputs on state farms almost halved between 1990 to 1992, while the use of
fertilizer fell even further (CSO, 1993). Lower levels of investment and input use
were accompanied by extensification of crop and animal production. For example,
cereals replaced more intensive crops, such as sugar beet or potatoes
(Wojciechowski, 1994). The drought of 1992 worsened the financial plight of state
farms (Institute of Agriculture and Food Economics, 1994).
As a result of the fiscal squeeze, the state farm sector and most individual state
farms have run at a loss since 1991. A change in the accountancy system allowed
this loss to be explicit in farms' financial statements (Kondraszczuk et al, 1992). By
June 1990, 70 percent of state farms had lost their credit-worthiness. By the end of
1993, state farm liabilities were in excess of 18 billion PZL (APA, 1994). There was,
thus, a clear need for restructuring the state farm sector, and the Agriculture
Property Agency (APA) was therefore established on January 1, 1992 to privatize
state-owned agricultural land.
16



State Farm Restructuring
The Polish Government's financial capacity is presently limited, and funds are not
available to continue to subsidize agriculture. On the contrary, the Government
expects to receive revenues from the proceeds of the privatization of state
agricultural land and assets. While the privatization process was perceived to be
well on track in most sectors by the early 1990s, the privatization of state farms had
encountered major difficulties (APA, 1994).
APA was set up as an agency of the Treasury, but acts as an independent entity. It
thus has characteristics of both a bureaucratic and private organization. It does not
receive direct government assistance, but has to rely on loans from the financial
sector and proceeds (rents and sales) from assets under its control. All state
agricultural land is being transferred to APA with the objective of privatizing it as
soon as possible. However, a major impediment to this process is that when APA
takes over a farm, it also assumes responsibility for all outstanding debt, which
averages nearly 50 percent of the market value of all farm assets. This debt has to
be serviced from the proceeds of the farm, through either the privatization of assets
or the returns to farming, or some combination (APA, 1994).
Despite this difficulty, APA has achieved considerable progress in its first two years
of operation (1992-1993): (i) an action program has been prepared to restructure,
transfer, privatize or close down agro-processing enterprises on state farms-
liquidation of non-agricultural property has proceeded faster than liquidation of
farms; (ii) it has completed the take-over of the majority of farms, and began the
17



transfer process for the remaining ones; and (iii) it has privatized more than 35
percent of the former state agricultural land through lease, sale or give-away.
APA uses four basic processes to privatize state land it acquires (APA, 1994; World
Bank, 1994):
•  Land sale is regarded as the preferred manner of privatization.
Reprivatization claims of former owners present problems with land sales, as
does the lack of finance on the part of buyers, which restricts the number of
buyers.
*  Leasing of land requires less money and is quick to execute. It also allows
for easy sub-division of large farms. In general, however, farmers need
outside financial help to meet obligations as proceeds from the farms do not
provide sufficient revenue. Individual farmers are used to having ownership
rights and do not easily accept leasing as a contractual form governing
landuse. However, APA does its best to make the lessees feel secure on the
land.
*  Company agencies are limited special agencies formed in cooperation with
the Ministry of Agriculture to manage both animal and plant breeding farms.
In general, the operational assets are transferred to the company, but not the
. land, which is leased. These companies are wholly owned by APA.
* Management contracts are similar to leasing. However, the manager does
not farm for his own account, but works for the APA. In general,
management contracts are used when there is no possibility of leasing.
18



Thus, a satisfactory procedure already exists for transferring ownership of state
farms. However, the privatization process has been handicapped by a number of
practical and policy issues. More than four-fifths of the land already privatized is
being leased, while only five percent has been sold (Miller and Wojciechowski,
1994). Reasons for this imbalance include: insufficient private resources to buy land;
low profitability of agriculture; administrative constraints with respect to land
registration; land reprivatization claims; and institutional and political constraints
to more rapid land sales-strong pressure to sell state assets not too cheaply with
prices based on non-market or book values rather than on market values;
perceptions that future competitiveness of Polish agriculture requires large areas
under single control so as to allow for economies of scale; and internal concerns at
APA regarding its long term role and the tenure of personnel. As a result,
transformation is not popular across the entire political spectrum in Poland,
particularly as it reveals hidden unemployment.
Moreover, the farm size issue has also not been resolved satisfactorily. In general,
APA is in favor of smaller farms and acknowledges the superiority of family farms
(Tanski, 1994). However, the prevailing opinion is that specific circumstances in
Poland, enable large, mechanized farms to cope better financially, especially with
respect to input and output marketing. Reasons given for this advantage include:
the quality of existing infrastructure on farms; the policy environment; the lack of
access to European Community markets; the concentration in agro-processing; and
the collapse of agro-services and transportation difficulties.2
19



The belief in the superiority of large-scale mechanized farms under these conditions
is also borne out by the reluctance to restructure state farms by creating farms of
less than 200 hectares in size. State farms average about 2,300 hectares in size,
while the average parcel of land offered for privatization by APA is generally 600
hectares-far higher than the average private farm size. Thus, although the
privatization of state farms is accompanied by some restructuring into smaller
units, these new farms maintain the general characteristics of large-scale
mechanized farms. Moreover, criteria for business plans on which the restructuring
is based accentuate modernization and mechanization (APA, 1994).
Private Farms
Polish agriculture never underwent complete collectivization. The dominant private
sector mainly consists of small farms. Only 8 percent of private farms are larger
than 15 hectares, yet they cultivate 24 percent of the total of private farmland
(Table 3).
Table 8: Private Farm structure in Poland, 1991
Category     Size           Number        |       Area          Average size
(Hectares)  (Thousand)  (Percent)  (Million.  (Percent)  (Hectares)
hectares)
Small        < 5        1,108       48.7      2.7       19.9        2.4
Medium       5-10        692        30.4      4.4       32.4        6.4
Large        > 15        183        8.0       3.3       24.3        18.0
Source: Institute of Agricultural and Food Economics (1990); World Bank (1991)
The Polish Government has sought to promote a long-term policy of land
consolidation in an attempt to restructure agriculture. Between 40,000 to 60,000
20



hectares of privately-owned farmland were consolidated annually. While the
principal effect has been on middle-sized commercial farms that depend largely on
farming for a living, the smallest farms have also been targeted. Farms, ranging in
size from 5-15 hectares, are declining in number as they are consolidated either
with official financial support, bought by larger farms (over 15 hectares), or
purchased by the State in exchange for a pension. Until 1989, land consolidation
was also fostered by restricting access to inputs, operating and development credit,
and extension services; by encouraging older farmers to retire and sell their land in
exchange for a pension; and by deliberately not producing the appropriate size
machinery for smaller farms. The overpowered equipment used on medium-size
Polish farms is particularly noticeable (Report of the Polish Task Force, 1990).
Analyzing the Farm Size -Efficiency Relationship in Polish
Agriculture
The issue of optimal farm size remains unresolved in Poland. Promotion of larger
farms is based on the perception that there are real economies of scale present in
Polish agriculture that favor large farms over smaller family farms. But these
perceptions are contrary to international evidence, which indicates that a large-scale,
mechanized farm sector is generally inefficient, especially when compared to small-
scale family farms. No study has, however, been carried out to date that adequately
analyzes the relevance of the farm size-productivity relationship in Polish agriculture.
There is, thus, ample motivation for a detailed analysis that should allow for an
evaluation of the present policies of small farm consolidation and state farm
restructuring.
21



There is a paucity of accurate farm-level data on the performance of large-scale,
mechanized farms, both state-owned and private. The old state farms are in a state of
neglect with many facing bankruptcy. Newly privatized farms have not been in
operation long enough to justify analysis. Therefore, the analysis in this paper utilizes
data on existing private farms to investigate the farm size-efficiency relationship.
Two different methodologies are used to determine the farm size-efficiency
relationship in two regions of Poland: first, total factor productivity (I'FP) differences
between small and large farms are determined; and second, non-parametric Data
Envelope Analysis (DEA) is used to estimate scale efficiencies. These two approaches
to determining efficiency are complementary. TFP provides a measure of overall
productivity, but does not analyze the contribution of different sources of efficiency.
By contrast, DEA isolates scale efficiency from technical and allocative efficiency.
Both approaches meet the requirements for appropriate methodologies to measure
farm productivity covered in Binswanger et al (1995).
Data
Surveys conducted by the Institute of Agricultural and Food Economics (IAFE) in
Warsaw, Poland provided the data for the analysis. IAFE conducts these surveys
every year, but in this paper only 1993 data were used. In 1993, 736 private farms
were surveyed. For these surveys, Poland was divided into nine relatively
homogeneous regions. However, the analysis in this paper uses data from only two
of these regions: Region 1 is represents the area around the capital of Warsaw-
22



Stoleczny; and Region 8 represents the mid-west of Poland-Srodkowo-Zachodni
(Table 4).
Table 4: Surveys Included in the Farm Size-Efficiency Analyses
Region                  Predominant   Year Covered   Number of
Type of Farming   by Survey     Farms
Surveyed
Region 1:             Mixed farming:      1993          97
Stoleczny             Grain, livestock
Region 8:             Mixed farming:      1993          151
Srodkowo-Zachodni     Grain, livestock
The sample of farms in each region was selected according to specific criteria, for
example farms smaller than three hectares were excluded and only progressive
farmers were selected. The selection within these criteria is random. However, due
to this procedure the sample does not represent the actual structure of farming and
the size distribution of farms in the regions surveyed. Typical farms included in the
sample are generally more market oriented and have higher productivity than
average farms. In 1993, the productivity of those farms was on average more than
40 percent higher than for the average farm.
Data were collected on-farm by using a Farm Record Book that covers the following:
(i) cash transactions (trade); (ii) credit transactions (trade) and reconciliations; (iii)
trade of crops, animal and processed products; (iv) monthly work of family and hired
labor (hours); (v) disposal of own machinery and custom hire (hired machinery); (vi)
farm inventory and changes; and (vii) description of farm (personal data, location of
farm, cropping structure, taxes and insurance paid, farmer's intentions and goals,
23



etc.). This information was then processed following internationally accepted
accounting practices to develop balance sheets, income and cash-flow statements.
These data are generally of good quality and have been used in several analyses
(Miller et al, 1994; Polish-American Extension Project, 1995).
The final data for each farm in the different samples used in the analysis involve
inputs and outputs. These were aggregated to give two output series-crops and
livestock-and seven input series-land, buildings, livestock and machinery
representing the stock inputs, and crop-specific (seed, fertilizer), livestock-specific
(feed, dips) and other inputs (electricity, etc.) representing flows (Table 5).
Table 5: Summary Size Characteristics of Farms Analyzed
Variable                     Repon 1: Stolecn            Regio 8: Srodkowo-Zachodni
___Average   Maximum   Minimum   Average  Maximum   Minimum
Farm Size:
Actual size (ha)       12.87     63.88       3.56     14.69     101.53      3.19
Quality-adjusted (ha)  13.87     44.66       1.44     14.71     82.44       2.36
Output:
Crops (PLN)           8,366      19,671     1,365     13,664    82,990     2,215
Livestock (PLN)        7,906     24,512      134      12,472    54,140       0
Stock Inputs:
Land (PLN)            16,737     52,278     2,311     20,746    131,901    3,769
Buildings (PLN)       29,076    102,076     4,076     22,324    86,983     1,475
Livestock (PLN)        5,505     20,371      12       6,714     26,467       0
Machinery (PLN)       29,353    118,851     2,412     20,814    116,896     399
Flow Inputs:
Crop inputs (PLN)     1,701      6,514       310      3,286     24,338      747
Livestock (PLN)       4,099      13,255      93       7,158     30,310       2
Other inputs (PLN)     1,474     4,072       201      2,144     7,062       367
The data from these surveys specifically allow for the elimination of the problems
with earlier studies. In particular, farm size has been adjusted for differences in land
quality within regions by using the land classification system and land tax
coefficients.3 This methodology is commonly used in Poland to normalize land
24



areas.4 In this paper all land is quality-adjusted and expressed in terms of Land
Class IV equivalents.
Another important point is that, within a specific region, all farmers face essentially
the same sets of prices because they buy from the same input suppliers and output
markets for most commodities were controlled. This implies that monetary values of
outputs and inputs (revenues and costs in the relevant categories) can be treated as
quality adjusted quantities,5 which greatly enhances the reliability of the analysis
as it also normalizes input and output quantities by eliminating the effect of quality
differences.
All analyses were conducted separately for each region or survey. Because the
analysis implicitly neglects possible production uncertainty, e.g., due to weather
effects, the underlying assumption is that all farmers within each survey face
similar production uncertainty. This seems to be appropriate given that the analysis
is conducted for a given production year and covers one relatively homogeneous
region at a time.
Total Factor Productivity Measures
Total factor productivity (TFP) for different farm size categories is a superior
indicator of the farm size-efficiency relationship compared to partial indicators,
such as physical output or value of agricultural output per unit of operated area, as
it accounts fully for differences in labor and other input use. All quantity
measurements used in the TFP analysis were annual flow variables. Stock variables
were transformed into flow variables by calculating the equivalent annuities based on
25



the relevant interest rate for that period and region, the average useful life of the
particular assets, and the applicable tax rate.6 Thus, the analysis presented measures
all inputs and outputs as annual flows expressed in monetary terms.
Private Total Factor Productivity Analysis
TFP values for different farm size categories have been calculated and compared for
each of the farm surveys (Table 6). The most efficient farm was used as the
reference point for calculating the Tornquist-Theil Index, which yields a
comparative TFP index.7 The methodology for constructing the TFP index is
described in detail in Thirtle et al (1993).
Table 6: Relative Total Factor Productivity and Labor/Machinery Indices,
by Farm Size
Region / Item                       Farm Size  ectares|
<5    5-10   10-15    >15
Region 1: Stoleezny
Total Factor Productivity*    138.4  140.2  147.0   100.0
Labor/Machinery Ratio*        118.3  127.1  139.7   100.0
Region 8: Srodkowo-Zachodni
Total Factor Productivity*    111.6  110.7  121.7   100.0
Labor/Machinery Ratio*        126.8  131.4  125.5   100.0
Notes: * TFP index and labor/machinery ratio of large farms (>15 ha) are the norms
(100) against which the other size categories were compared.
Analysis indicates that the differences in both TFP and labor/machinery ratios
between the farm size categories smaller than 15 hectares and larger than 15
hectares are statistically significant at the 10 percent level, while differences
between the farm size categories smaller than 15 hectares are not statistically
26



significant at the 10 percent level (Table 6). The finding yields a downward sloping
curve for TFP with respect to farm size (Figure 1).
Within the sample of farms analyzed, the results are clear: (i) large farms are not
more efficient than smaller farms, particularly those within the 10-15 hectares
range; and (ii) smaller farms have a larger labor/machinery ratio than large farms,
indicating that they are relatively more labor intensive in their mode of production.
Figure 1: Relative Total Factor Productivity per farm size in the Stoleczny
Region of Poland, 1993
110                      -
100
90
30               I 
0      5      10     15     20     25      30     35     40      45
Adjusted Farn Size (ha)
Social Total Factor Productivity Analysis
Using market prices to measure productivity assesses differences in private
efficiency, while the use of social opportunity costs as a measure eliminates the
impact of distortion and measures differences in social efficiency. Few studies, none
of them in Poland, has made this distinction in the analysis of the farm size-
efficiency relationship. In the period under consideration, the price of capital was
distorted by several factors, including tax benefits and interest rate subsidies. This
27



contributed, amongst other things, to over-capitalization of larger farms relative to
smaller ones (Table 6); output prices were also distorted due to protection and
market price support,8 with larger farms benefiting disproportionately. Accounting
for these distortions is thus important when looking at farm structure and
production relations from a social perspective.
To account for these distortions, social efficiency estimates were calculated for
farms in the two samples. Social opportunity costs for capital, labor, variable farm
inputs and farm outputs (both crops and livestock) were obtained from previous
studies and verified through discussions with government officials and World Bank
staff. The TFP analysis of each farm was repeated using these social opportunity
costs rather than private costs (Table 7).
The results obtained from the social TFP analyses suggest that: (i) average social
TFP is lower than average private TFP in both regions; and (ii) large farms (> 15
hectares) are more inefficient relative to smaller farms when social opportunity
costs are taken into account.
The reason large farms are more inefficient stems mainly from differences in the
relative importance of labor and capital in the input mix of large and small farms.
The value of output of small and large farms is generally affected in a similar
manner because the ratio of livestock to crops does not differ significantly between
these groups, but the input mix varies considerably, with large farms being
relatively more capital intensive and small farms being relatively more labor
intensive. Because the social opportunity cost of labor is lower than the actual wage
28



rate due to high rural unemployment, but the social opportunity cost for capital is
higher than the actual prices farmers face, the total value of inputs increases more
for large farms than for small farms.
Table 7: Social Relative Total Factor Productivity by Farm Size
Region                               Farm Size                Average
(Htares _)            Social/Private
< 5   6 5-10   10 - 15   > 15   TFP Ratio
;Reegn 1: Stoleczny         145.8  150.4   159.6   100.0        0.87
Reion 8: Srodkowo-Zachodni  120.4  121.9   131.2   100.0        0.89
Notes:    *     Four farm size categories were defined for each dataset.
**    TFP index and labor/machinery ratio of large farms (>15 ha) are the
norms (100) against which the other size categories were compared.
The results of the social TFP analysis should be interpreted with care. Farmers
react to the incentive structure facing them, and if capital is relatively cheaper, they
use more of it, and vice versa. For this reason the social TFP calculations are more
indicative of the distortions than the actual social costs or efficiency losses.
Changing the values from private to social prices strictly does nothing to the
physical input and output ratio, and TFP stays essentially the same, although the
weighting of the inputs and outputs change. However, the point here is to determine
to what an extent farm size influences the farmer's ability to capture benefits and
utilize the incentive structure.
29



Non-parametric EiTiciency Estimation Using Data Envelope Analysis
(DEA)
The analysis of efficiency has fallen into two broad categories: parametric and non-
parametric. The first approach relies on a parametric specification of the production
function, cost function or profit function (Forsund et al, 1980; Bauer, 1990).
Alternatively, production efficiency analysis can rely on non-parametric methods
(Seiford and Thrall, 1990). Building on the work of Farrell (1967) and Afriat (1972),
the non-parametric approach has the advantage of imposing no a priori parametric
restrictions on the underlying technology (Fare et al, 1985). It can also easily handle
disaggregated inputs and multiple output technologies. As the non-parametric
approach develops, its applications for production analysis have become more
refined (Chavas and Aliber, 1993), thus providing new opportunities for empirical
analysis of economic efficiency.
Non-parametric scale efficiency (SE) measures were developed in response to the
earlier work on technical efficiency (TE) and allocative efficiency (AE) (Baumol et al,
1982). While TE and AE take the output level as given, SE is concerned with
choosing the output level itself. The key question becomes whether firms are
operating under decreasing, increasing, or constant returns to scale (CRTS). SE
takes on values between 0 and 1, where SE=1 identifies scale efficiency under (local)
CRTS. Finding SE<l means that the firm is not scale efficient, i.e., does not produce
at a scale.exhibiting local CRTS. In this context, (1-SE) can be interpreted as the
relative decrease in average cost obtainable from rescaling outputs to the point of
(locally) CRTS.
30



The fDrmulation of the non-parametric programming problem for this analysis is
based on the work of Chavas and Aliber (1993) and is therefore not repeated in its
entirety. Essentially, data envelope analysis (DEA) uses a linear programming
procedure to minimize inputs per unit of output to determine the frontier of best-
practice farms, and then to determine the efficiency of all the production units relative
to the frontier. This estimation approach is preferred to econometric modeling where
the techniques impose a functional form, and having zeros for some inputs can cause
problems (Nunamaker, 1985; Ali and Seiford, 1993).
The non-parametric Farrell technical efficiency measure is defined as the locus of the
efficient points using the minimum required inputs to produce the unit level of output.
The efficiency of the other farms is measured relative to this isoquant (Figure 2).
Figure 2: Farrel Efficiency Measurement
'i' A;~~~~
°                               input X,
The efficiency frontier unit isoquant is determined by the linear combination of just
two efficient farms, B and C, and is labeled Y*. The efficiency of a farm such as A,
that is not on the frontier, is measured by the ratio OD/OA, since OD is the vector
31



representing the lowest mix of inputs that farm A could use and still reach the
isoquant, using its own factor combination.
Efficiency measures for the Polish data provide an assessment of aggregate, total
efficiency, and include both technical and scale effects. As our interest is primarily
on the farm-size-efficiency relationship, it is important to separate out the effects
of farm size on efficiency (Figure 3). The effect of this decomposition is covered by
Fare et al (1985), while Piesse et al (1995) and Van Schalkwyk et al (1993) provide
empirical applications.
CRTS technology is denoted by the linear total product curve, OP, from the origin,
through the efficient production units B and C. Units A and D, in this example, are
inefficient as they are below the CRTS frontier. When non-constant returns to scale
are allowed for, the frontier (ABCD) is concave, and the input-output combinations
A, B, C and D are all technically efficient.
Figure 3: Decomposition of Technical and Scale Efficiency
i                              P..
A  ~         ,*
o    x    x          x*.   Xe**   Iput x
32



When technical efficiency is extracted from total efficiency, only the scale effect
remains. Thus, farm A is scale inefficient by OX/OX', due to being too small, but is
technically efficient. Farm D is similarly technically efficient, but is too large and is
scale inefficient by OX*'/OX". Finally, farm E is technically inefficient by OX'/OX'
and scale inefficient by OX/OX', giving a total level of inefficiency, relative to the
CRTS frontier, of OX/OX'.
Results
DEA analysis has been used to estimate the farm size-efficiency relationship in
Poland for the data described earlier.9 In particular, the scale efficiency of each of
the farms is determined relative to that of all other farms in the same region
(Table 8).
Table 8: Average Technical and Allocative Efficiency Results
by Different Farm Size Categories
(1.00 = Efficient)
Efficiency     Region      < 6ha     5-lOha  |10- 15ha    > 16 ha
Technical     Region 1     1.000       0.971      0.981      0.939
CIE)          Region 8     1.000      0.980       0.980      0.955
Scale         Region 1     0.996       0.982      0.983      0.971
(SE)          Region 8     0.994      0.966       0.999      0.982
AMocative     Region 1     0.735       0.728      0.741      0.724
(AE)          Region 8     0.788       0.815      0.804      0.790
Total         Region 1     0.732       0.694      0.689      0.660
(TE*SE*AE)   Region 8      0.783      0.771       0.787      0.741
33



The results are, to some extent, similar to those obtained with the TFP analyses.
Large farms (>15 hectares) are in general less efficient than smaller farms (<15
hectares). However, the results of the different efficiency measure are mixed.
Insignificant differences between mean farm sizes of scale efficient (SE-eff) and
inefficient (SE-ineff) farms (p<0.10) were obtained for both regions, as well as for
the different farm size categories. The results for allocative efficiency (AE) were
similar, yielding no significant differences (at the 10 percent level). On the other
hand, large farms (>15 hectares) were significantly (p < 0.10) less technically
efficient than smaller farms.
The reason for the perceived inconsistency between the results obtained with the
TFP and the DEA methodology may be attributed to the inherent differences in the
methodology. DEA isolates scale efficiency from technical and allocative efficiency,
while TFP measurements do not differentiate between them.
The conclusions from the DEA analysis with respect to scale are: (i) differences
between scale efficiency for large farms (>15 hectares) and smaller farms (< 15
hectares) are not significant; (ii) large farms (> 15 hectares) are technically
significantly less efficient than smaller farms; and (iii) total efficiency (SE*TE*AE)
of large farms (>15 hectares) is does not differ significantly from that of smaller
farms.
The inverse of the scale efficiency measure (1/SE) has been investigated by plotting
it against quality-adjusted farm size to present visually scale efficiency (Figure 4).
Following Chavas and Aliber (1993), this inverse can be interpreted as something
34



akin to an average cost function, i.e., it is a declining function of outputs under
increasing returns to scale (IRTS) and an increasing function of outputs under
decreasing returns to scale (DRTS). Initially, efficiency increases as farm size
increases, depicted by the down-sloping section of the curve for farm sizes less than
10 hectares. This is followed by horizontal section in the curve where scale efficiency
is at its highest. Eventually scale efficiency decreases again as farm size become
larger than 20 hectares. From the figure it appears that the optimal farm size in the
Srodkowo-Zachodni Region of Poland seems to be in the range of 10-20 hectares.
Figure 4: Inverse Scale Efficiency Per Farm Size in the Srodkowo-
Zachodni Region of Poland, 1993.
1.15
'-1.10- ,                  .  
J105
£100                4
! 1 oo t " **   CI*.  Cv    *   CC
0.95
0.90                     i                               I
0         5         10         15        20         25         30
Adjusted Farm Size (ha)
Conclusions and Implications for Farm Restructuring
Despite a history of policies favoring relatively larger, mechanized farms in Poland,
the results show that, within the sample of farms analyzed, (i) large farms are not
more efficient than smaller firms, particularly those within the 10-15 hectares
range; and (ii) smaller farms have a larger labor/machinery ratio than large farms,
indicating that they are relatively more labor intensive in their mode of production.
35



These results are in accordance with empirical evidence on the farm size-efficiency
relationship elsewhere in the world. It is somewhat surprising, however, that the
range of scale-efficient farms is relatively small. General experience indicates that
inefficiencies usually set in after a much wider range of optimal farm sizes.
These results have important policy implications with respect to farm restructuring
for Poland and other transition economies facing similar issues and conditions. The
results imply that there will be no efficiency losses if large mechanized farm8 are
restructured into smaller units. On the contrary, there may well be some efficiency
gains. Moreover, such a process will lead to more labor intensive processes that will
be beneficial in an environment where unemployment and capital scarcity are
severe constraints to production. It also illustrates that the consolidation present
policy of consolidating small farms into larger units should be reconsidered. The
results suggest that the most appropriate policies are those that 8upport a flexible
farm structure with a range of farm sizes where market factors and managerial
ability determine individual size preferences.
An important element in the promotion of such a flexible structure is to get the
system of incentives right. This implies the removal of policies and distortions
favoring larger farms over smaller farms. The basic principle should be to make
markets work by removing distortions and privileges favoring large farms, and
creating markets to service small farmers in areas where they are missing. This
should be done without entrenching new privileges. Imperfect markets should be
encouraged to work better.
36



Notes
l     Under certain circumstances, such as those in Poland, there are external ecdnomies
of scale (Johnson and Ruttan, 1994). These occur when firms or farms, as they increase in
size, experience advantages in terms of access to inputs, credit, services, storage facilities, or
marketing and distribution opportunities relative to smaller farms. This gives large farms
real advantages relative to small farms due to pecuniary economies or policy distortions
rather than to greater efficiency. On the other hand, diseconomies of scale may also occur, for
example when the labor market fails or do not exist, when transaction coats in the labor
market are high, or when the effort of hired labor is significantly affected by supervision (De
Janvry, 1987).
2     See Guba et al (1995) and Wei et al (1995) for recent developments in input and
output markets.
a     The land classification system distinguishes between six different classes of land for
arable land and pastures. The land tax coefficients, which were used to compare and
normalize the different land dasses are: Class I = 1.8; Class II =1.5; Class III = 1.3, Illa and
-IIb have values of 1.25 and 1.35, respectively; Class IV = 1.0, IVa and Vb have values of
1.05 and 0.95, respectively; Class V = 0.8; and Class VI = 0.5. The normalized land area
adjusted for quality is expressed in land class IV equivalents.
4     The normalized, quality adjusted land area is expressed in Land Class IV
equivalents. However, there has been some recent criticism against the use of the land tax
equivalents based on the land dassification system as this does not in all cases accurately
reflect the real value and production potential of land (Kazimierz, 1986). However, for lack of
any other acceptable methodology, this was applied in this study to adjust for land quality
37



differences. Testing the results by applying the Delphi technique shows that this
methodology is appropriate for the two cases/regions under consideration.
5     This amounts to assuming that the corresponding implicit price indexes are unity.
This approach has the advantage of being empirically tractable. Although it allows for price
variation across years and areas, it has the disadvantage of neglecting price variations across
farms within any particular survey. While the intuition is that these variations are small or
even negligible, they cannot be ruled out. The "rule of one price" (Chavas and Aliber, 1993)
does, for example, not take into account different transaction costs or market failures. The
assumption that all farmers within a survey face the same prices seem to reasonable given
the nature of the farm support system in these areas. An additional, but related point is that
the "rule of one price" implicitly accounts for commodities that are not of homogeneous
quality. Different farmers may face different prices because they purchase inputs or sell
outputs of different quality. By using the monetary values of input and output as quantities,
there is an adjustment for these quality differences.
6     To convert the stocks, namely land, buildings, livestock and machinery, into annual
flows, discount rates for these inputs based on the economic rate of depreciation (10 years for
machinery and 25 years for buildings), the interest rate on the relevant annuities, and the
pertinent tax rate were all calculated and multiplied by the market value of each asset. The
Jorgensen approach was used to determine the cost of capital for the conversions of the
stocks to annual flows.
7     See Ball, Bureau and Butault (1994) for a review  of the properties and
recommendations on the selection of different index numbers based on the axiomatic and
economic approaches. Following from this, the Tornquist-Theil methodology is appropriate
for this analysis.
38



8     See World Bank (1990; 1991) for a synopsis of these policies and their effects.
9     The non-parametric analysis of efficiency benefited from discussions with and
suggestions by Jean-Paul Chavas from the University of Wisconsin-Madison. He developed
the inital GAMS code for the analysis.
39



REFERENCES
Afriat, S.N. (1972). "Efficiency estimation of production funetions". International
Economic Review, 13: 568-598.
Ali, A. and Seiford, L. (1993). 'The Mathematical Approach to Efficiency Analysis.
In: Fried, H., Lovell, K. and Schmidt, S. (eds.) The Measurement of Productive
Efficiency, Techniques and Applications, Oxford University Press, 120-159.
APA. (1994). 'Report on the Activities of the Agricultural Property Agency of the
State Treasury in 1993". Agricultural Property Agency of the State Treasury,
Warsaw, May 1994.
Ball, V.E., Bureau, J.C., and Butault, J.-P. (1994). Intercountry Comparisons of
Prwes and Real Values. Unpublished Paper, Economic Research Service, USDA,
Washington, D.C.
Barbier, P. (1984). "Inverse Relationship between Farm Size and Land
Productivity." Economic and Political Weekly, XIX (52 &53): A-189-A-198.
Bardhan, P.K (1973). 'Size, Productivity, and Returns to Scale: An Analysis of
Farm-Level Data in Indian Agriculture". Journal of Political Economy, 81: 1370-
86
Barnum, H.N., and Squire, L. (1978). "Technology and Relative Economic
Efficiency". Oxford Economic Paper8, 30: 181-98
40



Bauer, P.W. (1990). "Recent developments in the econometric estimation of
frontiers". Journal of Econometrics, 46: 39-56.
Baumol, W.J., Panzar, J. C. and Willig, R.D. (1982). Contestable Markets and the
Theory of Industry Structure. Harcourt Brace Jovanovich, Inc., New York.
Berry, R.A. and Cline, W.R. (1979). Agrarian Structure and Productivity in
Developing Countries. Baltimore, Johns Hopkins University Press.
Binswanger, H.P., Deininger, K and Feder, G. (1995). "Power, Distortions, Revolt
and Reform in Agricultural Land Relations". In: J. Behrman and T.N.
Srinivasan eds., Handbook of Development Economics, Volume HI, ch. 42, pp.
2659-2772. Elsevier, New York.
Brewster, J.M. (1950). "The Machine Process in Agriculture and Machinery."
Journal of Farm Economics, 32(1): 69-81.
Carter, M.R. (1984). "Identification of the Inverse Relationship between Farm Size
and Productivity: An Empirical Analysis of Peasant Agricultural Production".
Oxford Economic Papers, 36: 131-45
Carter, M.R. (1994). Sequencing Capital and Land Market Reforms for Broadly
Based Growth. Paper No. 379, Department of Agricultural Economics,
University of Wisconsin, Madison.
Carter, M.R. and Kalfayan, J. (1989). A General Equilibrium Exploration of the
Agrarian Question. Unpublished Research Paper, Department of Agricultural
Economics, University of Wisconsin, Madison.
41



Chattopadhyay, M., and Rudra, A. (1976). 'Size-Productivity Revisited". Economic
and Political Weekly, XI: A- 104-A- 123
Chavas, J-P. and Aliber, M. (1993). "An analysis of economic efficiency in
agriculture: A nonparametric approach." Journal of Agricultural Resource
Economics, 18 (1): 1-16.
Cornia, G.C. (1985). 'Farm Size, Land Yields and the Agricultural Production
Function: An Analysis for Fifteen Developing Countries". World Development, 13
(4): 513-34
CSO. (1988). Unpublished Data, Central Statistical Office, Government of Poland,
Warsaw.
CSO. (1993). Statistical Yearbook of Agriculture (in Polish). Central Statistical
Office, Government of Poland, Warsaw.
De Janvry, A. (1987). Farm Structure, Productivity and Poverty. Working Paper No.
432, Department of Agricultural and Resource Economics, University of
California, Berkeley.
Deininger, K W. (1993). 'Cooperatives and the Breakup of large Mechanized
Farms". World Bank Discussion Papers, No. 218: 60 pp.
Deolalikar, A. (1981). 'The Inverse Relationship between Productivity and Farm
Size: A Test Using Regional Data from India". American Journal of Agricultural
Economics, 63 (2): 275-279.
42



Fare, R., Grosskopf, S. and Lovell, C.A.K. (1985). The Measurement of Efficiency of
Production. Kluwer-Nijhoff, Boston.
Farrell, M.J. (1957). 'The Measurement of Productive Efficiency". Journal of the
Royal Statistical Society A 120, Part 3: 253-81.
Feder, G. (1985). "The Relation Between Farm Size and Productivity: the Role of
Family Labour, Supervision, and Credit Constraints". Journal of Development
Economics, 18: 297-313.
Foreign Trade Institute. (1989). Annual Report. Foreign Trade Institute, Warsaw.
Forsund, F.R., Lovell, C.A.K. and Schmidt, P. (1980). "A survey of frontier
production functions and their relationship to efficiency measurement." Journal
of Econometrics, 13: 5-25.
Ghose, A.J. (1979). "Farm Size and Land Productivity in Indian Agriculture: A
Reappraisal". Journal of Development Studies, 16: 27-49
Gomula, M., Kondraszuk, T., Pyrgies, J., Krawczyk, I. and Rozwadowski, R. (1991).
'Current Status of State Farms". Land Policy and Privatization and
Restructuring of State Farms. Report No. P-5685-Pol, World Bank, Washington
DC.
Guba, W., Wei, A., Krzyzanowska, Z., Dabrowski, J., Chmielewska, W. and Burcroff,
R. (1995). "The Emerging Trends in Agricultural Marketing in Poland--A Report
of a Field Investigation". Unpublished Paper, Agriculture and Natural Resources
Department, World Bank, Washington, DC.
43



Hayami, Y. and Ruttan, V. (1985). Agricultural Development An International
Perspective. Johns Hopkins University Press, Baltimore.
Institute of Agriculture and Food Economics. (1990). Unpublished Data, Institute of
Agriculture and Food Economics, Warsaw.
Institute of Agriculture and Food Economics. (1994). Agricultural Input and
Services Market (in Polish). Institute of Agriculture and Food Economics,
Warsaw.
Johnson, N.L. and Ruttan, V.W. (1994). "Why are Farms so Small?' World
Development, 22(5): 691-706.
Kanel, D. (1967). "Size of Farm and Economic Development". Indian Journal of
Agricultural Economics, 22 (2): 26-44.
Kazimierz, B. (1986). 'Prpozycja nowego ujecia bonitacji gruntow ornych"
Zagadnienia Ekonomiki Rolnej.
Khusro, A.M. (1964). "Returns to Scale in Indian Agriculture". Indian Journal of
Agricultural Economics, 19 (3 & 4).
Kondraszczuk, T., Olko-Bagenska, Pyrgies, J., Runowski, H and Zientara, W.
(1992). Ownership Changes of State Farms. Unpublished Document, Warsaw.
Lipton, M. (1993). "Land reform as Commenced Business: The Evidence Against
Stopping'. World Development, 21(4): 641-658.
44



Miller, B.R. and Wojciechowski, J. (1994). "Status of State Farm Privatization in
Poland". Policy Note No. 1, Polish-American Extension Project, Warsaw.
Miller, B.R., Lublinska, B. and Wojciechowski, J. (1994). "Strategies for Farm
Growth in Poland". Unpublished Paper, Polish-American Extension Project,
Warsaw.
Nunamaker, T.R. (1985). 'Using Data Envelopment Analysis to Measure the
Efficiency of Non-Profit Organisations: A Critical Evaluation". Managerial and
Business Economics, 6(1): 50-58.
Patnaik, U. (1972). 'Economics of Farm Size and Farm Scale". Economic and
Political Weekly, VII (31, 32 and 33): 1613-24.
Peterson, W. and Kislev, Y. (1991). Economies of Scale in Agriculture. A Re-
examination of the Bvidence. Staff Paper P91-43, Department of Agricultural
and Applied Economics, University of Minnesota, St. Paul.
Piesse, J., Sartorius von Bach, H., Thirtle, C. and Van Zyl, J., (1995). "Agricultural
Efficiency in the Northern Transvaal Homelands". Journal of International
Development, (7): (forthcoming).
Polish-American Extension Project. (1995). Unpublished Research Papers. Polish-
American Extension Project, Warsaw.
Prosterman, R.L. and Riedinger, J.M. (1987). Land Reform and Democratic
Development. Baltimore and London, The Johns Hopkins University Press.
45



Report of the Polish Task Force. (1990). "An Agricultural Strategy for Poland".
Report of the Polish-European Community-World Bank Task Force. World
Bank, Washington, DC.
Seiford, L.M. and Thrall, R.M. (1990). 'Recent developments in DEA: The
Mathematical Programming approach to frontier analysis". Journal of
Econometrics. Vol 46: 7 - 38.
Sen, A. (1962). 'An Aspect of Indian Agriculture". Economic Weekly.
Tanski, Adam. (1994). Personal Communication. Agricultural Property Agency of
the State Treasury, Warsaw.
Thirtle, C., Sartorius von Bach, H.J. and Van Zyl, J. (1993). "Total Factor
Productivity Growth in South African Agriculture, 1947-1991". Development
Southern Africa, 10(3): 301-318.
Van Schalkwyk, H.D., Van Zyl, J. and Sartorius von Bach, H.J. (1993).
'Management and Returns to Farm Size: Results of a Case Study using
Parametric and Non-Parametric Methodology to Measure Scale Efficiencies".
Agrekon, 32(4): 252-256.
Van ZyL J., Binswanger, H.P. and Thirtle, C. (1995). "The Relationship Between
Farm Size and Efficiency in South African Agriculture". Policy Research
Working Paper No. 1548. World Bank, Washington DC.
Wei, A., Guba, W., Dabrowski, J., Chmielewska, W., Opalka, M. and Burcroff, R.
(1995). 'The Emergence of Integrated Agricultural Markets: Price Co-Movement
46



since Liberalization in Poland's Hog/Pork Sector". Unpublished Paper,
Agriculture and Natural Resources Department, World Bank, Washington, DC.
Wojciechowski, J. (1994). "The Implications of the Privatization Process of State
Farms in Poland with Particular Reference to Purchased Farms". Unpublished
MSc (Agricultural Economics) dissertation, Department of Agriculture,
University of Aberdeen.
World Bank. (1983). Kenya: Growth and Structural Change. Basic Economic Report,
Africa Region, Washington, D.C.
World Bank. (1990). "An Agricultural Strategy for Poland". Report of the Polish-
European Community-World Bank Task Force. World Bank, Washington, DC.
World Bank. (1991). Land Policy and Privatization and Restructuring of State
Farms. Report No. P-5685-Pol, World Bank, Washington DC.
World Bank. (1994). Poland: The Agricultural Property Agency. World Bank,
Wa3shington DC.
47



Policy Research Working Paper Series
Contact
Title                            Author                   Date              for paper
WPS1570 Protecting the Old and Promoting   Estelle James           January 1996       S. Khan
Growth: A Defense of Averting the                                            33651
Old Age Crisis
WPS1571 Export Prospects of Middle Eastern   Alexander Yeats       February 1996      S. Lipscomb
Countries: A Post-Uruguay Round                                              33718
Analysis
WPS1572 Averting the Old-Age Crisis:       Robert J. Palacios      February 1996      M. Pallares
Technical Annex                                                              30435
WPS1 573 North-South Customs Unions and    Eduardo Fernandez-Arias  February 1996     S. King-Watson
International Capital Mobility   Mark M. Spiegel                             31047
WPS1574 Bank Regulation: The Case of the    Gerard Caprio, Jr.     February 1996      D. Evans
Missing Model                                                                38526
WPS1575 Inflation, Growth, and Central Banks: Jose de Gregorio     February 1996      K. Labrie
Theory and Evidence                                                          31001
WPS1576 Rural Poverty in Ecuador-A         Jesko Hentschel         February 1996      E. Rodriguez
Qualitative Assessment           William F. Waters                           37873
Anna Kathryn Vandever Webb
WPS1 577 The Peace Dividend: Military      Malcolm Knight          February 1996      R. Martin
Spending Cuts and Economic        Norman Loayza                              31320
Growth                           Delano Villanueva
WPS1578 Stock Market and Investment:       Cherian Samuel          March 1996         C. Samuel
The Governance Role of the Market                                            30802
WPS1579 Different Strategies of Transition  Marek Dabrowski        March 1996         C. Rollison
to a Market Economy: How Do                                                  84768
They Work in Practice?
WPS1580 Indonesia's Cocoa Boom: Hands-    Takamasa Akiyama         March 1996         G. Ilogon
Off Policy Encourages Smallholder   Akihiko Nishio                           33732
Dynamism
WPS1581 Where Has All the Education Gone?  Lant Pritchett          March 1996         S. Fallon
38009
WPS1582 Stock Market Development and       Ross Levine             March 1996         P. Sintim-Aboagye
Long-Run Growth                  Sara Zervos                                 38526
WPS1583 Trade Preferential Agreements in   Michael Michaely        March 1996         D. Papageorgiou
Latin America: An Ex-Ante Assessment                                         31910
WPS1 584 Price Support at Any Price? Costs    Witold M. Orlowski   March 1996         W. Orlowski
and Benefits of Alternative Agricultural                                     37270
Policies for Poland



Policy Research Working Paper Series
Contact
Title                            Author                  Date              for paper
WPS1585 Public Finances and Economic      Luca Barbone            March 1996         C. Pelegrin
Transition                       Hana Polackova                             85087
WPS1586 Did External Barriers Cause the   Azita Amjadi            March 1996         S. Lipscomb
Marginalization of Sub-Saharan   Ulrich Reinke                              33718
Africa in World Trade?           Alexander Yeats
WPS1587 Payments and Finance Problems    Constantine              April 1996         M. de la Puente
in the Commonwealth of            Michalopoulos                             31206
Independent States
WPS1588 Social Insurance in the Transition  Deborah Mabbett       April 1996         L. Biely
to a Market Economy: Theoretical                                            36280
Issues with Application to Moldova
WPS1589 The Analysis of Emerging Policy   Sudarshan Gooptu        April 1996         R. Vo
Issues in Development Finance                                               31047
WPS1590 The Latvian Banking Crisis:       Alex Fleming            April 1996         S. Coffey
Lessons Learned                  Samuel Talley                              32535
WPS1591 Competition in Network Industries    Michael Klein        April 1996         S. Vivas
82809
WPS1592 The Stockmarkets as a Source of    Cherian Samuel         April 1996         C. Samuel
Finance: A Comparison of U.S. and                                           30802
Indian Firms
WPS1593 Rural Finance for Growth and Poverty Saeed Qureshi        April 1996         C. Anbiah
Alleviation                      Ijaz Nabi                                  81275
Rashid Faruqee
WPS1594 Why Have Some Indian States Done  Gaurav Datt             April 1996         P. Sader
Better Than Others at Reducing   Martin Ravallion                           33902
Rural Poverty?
WPS1595 In Search of Owners: Lessons of    Cheryl W. Gray         April 1996         M. Geller
Experience with Privatization and                                           31393
Corporate Governance in Transition
Economies
WPS1596 Agrarian Structure in Poland: The    Johan van Zyl        April 1996         M. Williams
Myth of Large-Farm Superiority   Bill R. Miller                             87297
Andrew Parker