Lecture 6.
Measurement of agricultural support

What we want to learn about this topic

The level of protection provided to an industry by tariffs is traditionally measured by indicators such as the nominal and effective rates of protection. Because of the complexity of agricultural support mechanisms, in which tariff protection is just one and not necessarily the most important component, there has been an attempt to define more comprehensive measures of support in the case of agriculture. In 1987 the OECD introduced two measures, the Producer Subsidy Equivalent (PSE) and the Consumer Subsidy Equivalent (CSE), which over the next decade became the de facto standard for the measurement of agricultural support levels. This was not least because the OECD published an annual series of these indicators calculated on the basis of official government data in its annual publication Agricultural Policies, Markets and Trade (now called Agricultural Policies in OECD Countries). In 1998, the OECD undertook a revision of these measures and relabelled them the Producer Support Estimate and the Consumer Support Estimate respectively. In this lecture, we look at:

• tariffs as a measure of agricultural support
• nominal and effective protection rate measures
• the definition and calculation of the original OECD PSE and CSE measures
• the changes to the PSE and CSE measures introduced in 1998
• the magnitude of the transfers generated by agricultural support as captured by the PSE and CSE measures
• some criticisms and limitations of the original PSE measure

Short introduction to the issues

Traditional measures of protection

The Nominal Protection Coefficient (NPC) is the ratio between the domestic price and the world price.

If the domestic price is 150 and the world price is 100, the NPC is 1.5.

An alternative way of expressing this is as the Nominal Protection Rate (NPR) which is the difference between the domestic price and the world price expressed as a percentage of the world price

In the above example, this would be 100 * (150 - 100) / 100 or 50%.

In either case, we are measuring the protection provided to the product. As a measure of the effect of government policy on the incentive to produce this commodity, the NPC can be criticised for not taking into account the impact of government policies on inputs. For example, even if there is no protection to the product itself, a government subsidy to an input used in the production of the product would provide an incentive to domestic production.

A measure which captures the effect of both product and input market interventions is the Effective Protection Coefficient (EPC). This is defined as the ratio of value added at domestic prices to value added at world prices. Analogously, we define the Effective Protection Rate (EPR) as the difference in value added at domestic and world prices expressed as percentage of value added at world prices. Note that relatively low NPC rates applied to a product can result in quite high EPCs particularly where value added is small relative to output value.

The Producer/Consumer Subsidy Equivalent

A key reason for introducing the PSE/CSE in the mid-1980s was to encompass as broad a range of support mechanisms as possible, including direct payments to farmers linked to either outputs or inputs which were not covered by the traditional measures. The original OECD PSE definition included four types of measures:

• market price support (measures which simultaneously affect producer and consumer prices)
• direct income support (measures which transfer money directly from taxpayers without affecting the market price)
• indirect income support (mainly reductions in input costs)
• other support (this included expenditure on general services to agriculture not paid directly to individual farmers, such as public support for agricultural R&D, structural policies, inspection services etc.) (Since 1998, this element is now reported separately as the General Services Support Estimate, see below).

An example of a PSE calculation is provided in class. Note the following:

• PSE calculations are done for a core group of 13 commodities but the importance of these commodities in total agricultural output varies across countries.
• transfers associated with particular policies, e.g. the underpricing of irrigation water and tax concessions, are often omitted for lack of data
• where expenditure is not commodity specific, it is allocated according to the share of each commodity in final agricultural output.
• PSEs exclude certain budgetary outlays (e.g. outlays on stock holding where their price-raising effect is already incorporated into the calculation of market price support, environmental payments etc.)
• in the case of livestock commodities, a feed adjustment is made to convert Gross PSEs into Net PSEs by subtracting the higher feed cost due to market price support given to cereal production.

The PSE and CSE 'subsidy equivalent' was originally defined by the OECD as 'the monetary value that would be required to compensate farmers or consumers for the loss of income resulting from the removal of a given policy'. In other words, it was defined as the lump sum payment which would leave farmers or consumers as well off as they were at present in the absence of current policies. This was, in fact, an error as what the indicator measures is the transfer to producers or consumers but not the change in their producer or consumer surplus. More correctly defined as:

• the PSE measures the value of the monetary transfers to producers from consumers of agricultural products and from taxpayers . The formal defintion is An indicator of the annual monetary value of gross transfers from consumers and taxpayers to agricultural producers, measured at the farm-gate level, arising from policy measures that support agriculture, regardless of their nature, objectives or impacts on farm production or income.
• the CSE measures the value of monetary transfers from domestic consumers to producers and taxpayers . It tells us how much more consumers are paying for food as a result of agricultural policy compared to buying at existing world market prices.

Most attention focuses on comparisons of PSE levels across commodities, across countries and over time and the rest of the lecture will focus on PSEs rather than CSEs.

PSEs are expressed in a number of ways:

• total PSE: the total value of transfers to producers
• percentage PSE: total value of transfers as a percentage of the total value of production (valued at domestic prices), adjusted to include direct payments and to exclude levies
• unit PSE: the total value of transfers per tonne
• PSE per farm or per labour unit

Two additional indicators of support are also reported in the OECD's annual monitoring exercise:

• the Nominal Protection Coefficient (NPC): this is the ratio between producer and border prices, and corresponds to the traditional measure of protection described above.
• the Nominal Assistance Coefficient (NAC): this is defined as the value of total production at farm gate prices plus budgetary support to the value of total production at world prices. (Alternatively, it can be defined as the ratio of the border price plus the unit PSE to the border price). It is thus analogous to the Nominal Protection Coefficient except that the top line of the ratio, the numerator, encompasses all forms of assistance and not just that provided through market price support.

The Producer Support Estimate

Changes introduced by the OECD in 1998:

• most obviously, the change in nomenclature to Producer Support Estimate to reflect the fact that the indicator is measuring transfers and not the 'subsidy equivalent' of the support provided.
• the separation of general services support from the Producer Support Estimate measure - thus expenditure on research, infrastructure, taxation concessions etc. are no longer included in the PSE but are referred to as the General Services Support Estimate (GSSE). The GSSE is also removed when calculating the NAC measure. The effect is to lower slightly percentage PSE and NAC measures compared to the old definitions.
• the sum of the Producer Support Estimate, the General Services Support Estimate and the transfers from taxpayers to consumers (in the CSE) is now called the Total Support Estimate (TSE). Therefore, the TSE includes all transfers included in the three other indicators (excluding double-counting). It is the overall monetary cost of the transfers arising from policy measures in a given country. As defined, it thus overestimates the total support received by producers alone.
• on the other hand, the total PSE indicator is now published to represent the whole agricultural sector and not just the 13 commodities (even though the national percentage PSE for each country continues to be based on these 13 commodities). The reason given for this change is that, under the previous method, total budget support to the agricultural sector as a whole (i.e. the non-market price support bit) had to be arbitrarily scaled back to the 13 commodities to provide a consistent PSE measure. This is at least as likely to cause errors as 'inflating' the PSE estimate based on the 13 commodities to the agricultural sector as a whole, and over time as budgetary expenditure grows in importance relative to market price support, the relative error using the new procedure should decrease.
• a new classification scheme for measures included within the PSE based on implementation criteria defined as the conditions under which the associated transfers are provided to farmers, or the conditions of eligibility for the payment. The purpose of this exercise is to enable a much better analysis of the impact of different measures to be made; for example, payments linked to output are separated from payments based on area planted/animal numbers and from payments based on historical entitlements based on the different degrees of decoupling involved.

The magnitude of support

There is no substitute for looking at the latest OECD Agricultural Policies in OECD Countries volume to look at the trends in support over time and across commodities and countries. Key points highlighted in the 2004 highlights include:

• a slow downward trend in the level of support to producers, as measured by the percentage PSE, in most OECD countries, but more markedly in countries with low overall levels of support
• stability in the support to general services provided to agriculture, as measured by the percentage GSSE (the percentage of GSSE in TSE)
• a slow long run decline in total support to agriculture (measured by the percentage Total Support Estimate which is calculated as the ratio of TSE to national GDP ) and a shift in the financing of support from consumers to taxpayers;
• but despite reductions in market price support, it is still the main source of support and, together with commodity-linked payments, represents over 75% of support to producers.

Issues in the calculation of PSEs

• Are actual world market prices a proper reference point for the PSE? Measurement of the market price support element requires the use of a world reference price as a benchmark. Taking the existing world market price ignores the fact that this price may be depressed by the very existence of support and the PSE calculated on this basis overestimates the amount of support provided. OECD argues that the PSE is a measure of policy effort which should e kept separate from any analysis of policy effect.
• Do world market conditions distort the PSE? Because the market price support element is based on a comparison of domestic with world market prices, the PSE value can vary even if there is no explicit change in domestic agricultural policy (i.e. domestic prices remain unchanged) because there is a change in world market prices. OECD argues that, if policy makers insulate their domestic markets from world markets they implicitly provide levels of support which vary over time, and thus it is correct to take border price fluctuations into account in comparing PSE levels from year to year.
• Role of exchange rates. As PSEs require a comparison of world and domestic prices, they need an exchange rate to convert one into the other. But the market exchange rate may not be an equilibrium one and may thus exaggerate or underestimate the true level of support. This was an important issue in the OECD measurement of agricultural support in the CEECs. Another argument is that exchange rates fluctuate, thus causing the level of the PSE to change for reasons which have nothing to do with agricultural policy per se. However, if the underlying exchange rate is an equilbrium rate, these fluctuations represent real changes in competitiveness and are appropriately taken into account in the measurement of agricultural support. Note, however, that if, for example, the US$/Euro exchange rate changes over time, then a PSE series measured in US$ will evolve differently than a PSE series measured in Euro, not only in magnitude but also sometimes in direction.
• Does the PSE properly measure agricultural policy reform? Some of the criticisms of the PSE measure arose because of its use as a measure of trade distortion rather than simply support. Developing countries and NGOs often argue that 'OECD countries are spending $300 billion on support to farmers' implying that all of this is somehow damaging to producers in other countries. Critics worry that the aggregate measure takes no account of the degree of decoupling from output of different payment mechanisms, for example, a similar expenditure on market price support and R&D would not be expected to have the same effect on trade. Nor does it take into account the existence of supply control measures which might severely constrain the output-increasing effect of higher support payments. OECD argues that it is necessary to look at the composition of the PSE as well as the overall size to correctly interpret its impact on trade. A variety of PSE-like measures which tried to 'adjust' the basic PSE measure for the different distortive 'strength' of different individual policy measures have been suggested, but none have become as popular as the basic PSE.

Reading suggestions

Legg, W., 2003. Agricultural subsidies: measurement and use in policy evaluation, Presidential Address to the UK Agricultural Economics Society (forthcoming in the Journal of Agricultural Economics).
(this overview paper is written by one of the OECD economists involved in developing the PSE/CSE concept; it reviews the origins, concepts and trends of the PSE indicator)

OECD, 2004. Agricultural Support: How is it Measured and What Does It Mean?, OECD Policy Brief, Paris

Tangermann, S., 2005. Is the concept of the Producer Support Estimate in need of revision? OECD Food, Agriculture and Fisheries Working Papers, No. 1, OECD Publishing.

Supplementary references

The OECD PSE database can be accessed as an Excel file.

OECD, 2005, Agricultural Policies in OECD Countries: Monitoring and Evaluation, Paris, 2005.
(this extract summarises tables on recent trends in PSE indicators for OECD countries. Read only the first half 'Evaluation of Support Policy Developments', the second half consists of country chapters where you might like to look at the chapter on the EU)