security is essential to avoid dependence on other countries for food supply. However, recent decades’ objectives

have been focused on farmers’ income stability as well as on the environmental impacts of agriculture. While there

is a consensus on the depressing effects on consumers' and taxpayers’ welfare, the discussions on the public policy

impacts on the agricultural outcome are of a wider range. Empirical studies at the farm level doubt the positive

effect of farm support on their technical efficiency. This paper provides an analysis of the role of Producer Support

Estimate (PSE) as a source of assistance on a commodity basis in a group of OECD and other big agricultural

traders. With a special focus on the Producer Single Commodity Transfer (PSCT) effect on the countries’

commodity production levels, the general finding is that the government intervention in specific commodities

investigated here may not be efficient.

1 Introduction

The agriculture sector has steadily enjoyed

government support for a relatively long period,

especially in developed economies. Considerations

relate to the strategic behavior of countries'

leadership, in that ensuring food security is essential

to avoid dependence on other countries for food

supply. However, recent decades’ objectives have

been focused on farmers’ income stability as well as

on the environmental impacts of agriculture. While

agricultural support policies on farms’ economic

performance in terms of production levels, although

policies that support agriculture, regardless of their

objectives and their economic impacts”, [2].

production levels. The PSCT tool falls under the

Producer Support Estimate (PSE), which is an

instrument of the support to producers’ policy.

Considering components of PSE and PSCT which

include support and payments based on commodity

outputs and input use, one would theoretically

suggest that the effect of the measure and its

empirical hints on the direction of policy impacts

mostly at the farm level, while implying complex

theoretical impact pathways. In this paper, it is

argued that PSCT has an impact on the total product

level of specific commodities when estimated for a

mix of developed and emerging economies for a

period of about 20 to 30 years, although the direction

might not be as positive as it could be expected. This

analysis contributes to the knowledge of agriculture

related support policies by investigating on the

efficiency of such policies, in terms of the production

level impacted. Most of the existing literature

calculations and interpretation of the estimates of

interest that fall under the PSE category. The third

section surveys existing empirical literature that

evaluates the role of support policies and also the

possible changes in their regimes on the farmers’

performance. The methodology for estimation of the

PSCT component of PSE on commodity-specific

production in an aggregate level through a production

function approach that will follow in this paper is

also introduced. Section four provides the data

developed to monitor and evaluate performance in

agricultural policy. The purpose would be to

to substitute for the Producer Subsidy Equivalent.

The change was to reflect the fact that the indicator is

measuring transfers and not the 'subsidy equivalent'

of the support provided. An investigation of the

estimation of such assistance by [6], explores the five

categories of agricultural policy measures to be

included in the OECD calculations of PSEs: (i)

Payments which indicate measures that transfer

money directly to producers without raising prices to

consumers; (iii) Reduction in Input Costs that involve

measures which lower input costs (capital or other

inputs); (iv) General Services which indicate for

money not directly received by producers but that

contribute in the long term reduction of costs; (v)

Other indirect support such as taxation concessions

and other subnational/subregional subsidies.

Commodity Transfer (PSCT) represents the “annual

monetary value of gross transfers from consumers

from policies that have been attributed to single

commodities, as the following equation (1) shows:

where the Σ BOT represents the national aggregate

budgetary and other transfers to producers from

policies that have been labeled as based on a single

commodity (SC). Because the empirical analysis in

this paper will be performed on this measure (PSCT),

there is a rationale to provide some detailed

market price is Pd, the world price is Pw, direct

payments are D, levies on producers are L and all

other budgetary-financed support is B.

the PSE is based on a comparison of world market

prices with the domestic ones, and most criticisms

existence of support to the production level through

the PSE and hence the calculated measure could

overestimate the amount of support provided.

Moreover, a variation in the world price could be

reflected in the PSE value even when there is no

explicit change in domestic agricultural policy.

Furthermore, there are concerns regarding the

exchange rate (fluctuations and not equilibrium rate

usage) in the comparison of the world to domestic

prices in the PSE calculations.

aggregation of the measures related to real market

programs are explicitly set up to insure against risk in

the first place, thus suggesting that the programs tend

to exist in the markets where the insurance effect

may be largest.

subsidy-free equilibrium world prices as a

benchmark for calculating subsidies would hardly

command agreement among policymakers, as it

would anyway require a model using the PSE as

calculated by the OECD, [19]. Overall, there is a

technologies. Direct payments based on output or

variable input use were also found to be highly

can generate a negative effect on productivity;

coupled output support generally produces a negative

effect on technical efficiency and productivity. The

rationale is that the existence of supply control

measures might severely constrain the output-

increasing effect of higher support payments.

However, in the long term, there may be a positive

improving innovations. He also acknowledges the

direct incentive to increase production when

assuming the environmental effects of the A1 and A2

components of the PSE (which are also part of

PSCT).

measures have been mostly used to estimate technical

efficiency, trade distortions, and environment

impacts at the farm level, [1], [10], [23], [24], the

empirical literature built on agriculture aggregate

data makes almost no use of the above-discussed

measures. There is limited evidence on how

concept of the agricultural sector, [26], [27] and

hence the multidimension (or even absence of well-

defined) objectives of agricultural policy as argued

by [28], produce various (contradictory) performance

effects of the support policies. The following section

contributes with a discussion on how the support

policies could be estimated to impact agricultural

production function.

on the level of production, γ is the coefficient for the

PSCT variable and u is the error term. The panel data

more degrees of freedom and more efficiency” (p.

637).

investigation, the fixed and random effects methods

simulate unmeasured time-invariant factors as

country-specific intercepts. The intercept varies for

each country but still, the slope coefficients are

intercepts are treated as fixed effects to be estimated,

equivalent to including dummy variables for the N-1

number of countries. On the other side, instead of

treating the intercept as fixed and assuming that it is

a random variable with a mean value of a, the

intercept value for one country can be expressed as:

aj = a + εj j = 1, 2, …, N countries.

Equation (3) would be re-written as follows:

informs about the data used and the estimated

Commodity Transfer, measured as the ratio of the

total sum of the components A1, A2, B, C2, and D

described earlier in section 2 over the farmers’ gross

revenue, expressed in percentage. Given that the

monetary indicators are influenced by the size and

structure of the country’s agricultural sector, as well

as the country’s rate of inflation, percentage

indicators allow for comparisons of support levels

between countries, assess the level of support

provided within a country to different commodities

phosphate content of various fertilizers consumed,

measured in thousand metric tons.

empirical estimations, are presented for the whole

sample as well as for the subsample for which the

indicator does not take negative values (is either 0 or

positive). In Table 3 (Appendix), the subsamples are

represented by a smaller number of observations for

each of the commodities. As can be observed, there

are discrepancies between the countries observed in

geographical area and hence employment in the

sector and area harvested. There are differences also

in terms of machinery and fertilisers used. For

example, for wheat production, the minimum values

for all the variables (except for fertilizers which is

South Africa and PSCT) are for Israel, and the

biggest stands for either China or India. The lowest

(negative) value of PSCT percentage is for Argentina

and the largest for Japan.

percentage PSCT indicator in the countries under

investigation for wheat commodities. Other

effectively “taxed” by government policies.

have enjoyed higher support, although that has been

reduced. The same is observed for Korean soybean

growers. It can also be observed from the graphs of

PSCT indicators that India, Russia, and Ukraine farm

policies have followed a very changing trend from

positive to negative support and vice versa. Referring

to other countries, the trends are in line with what

[35] also confirms, that overall producer support as a

shown in Appendix. The relationships observed are

those expected, production is positively related to

aggregate input coefficients are not significant, and

since these are indeed not commodity specific, it is

deemed rational to drop them from the equation and

run the regressions maintaining area harvested and its

interaction term with the aggregate inputs. The

degrees of freedom would also increase as the

aggregate input variables are dropped off with the

efficient estimation.

coefficient results, the land input is significant for

Maize and Soybean, showing a concave relationship

between production level and with area harvested.

That indicates that there is a decreasing marginal

productivity of the area harvested, which could be

due to the use of marginal land as the crop area (for

these two commodities) increases. On the contrary,

the relation between output and fertilizers observed is

indicative of increased marginal productivity of their

usage (negative coefficient of the input and positive

with employment and machinery (when coefficients

show to be significant). Reminding that these are

agricultural sector aggregate input levels, it could be

that intensity of labor and technology in limited

cropland areas make these inputs less efficient. The

interactive term of area harvested with fertilizers is

still significantly positive, indicating that the intense

use of this input positively affects commodity

products. This finding is in line with what [36]

highlights, that agricultural fertilizer use is one of the