AGRICULTURAL POLICY REFORM IN TURKEY:  A SOCIAL  ACCOUNTING MATRIX PERSPECTIVE

Jonathan BROOKS, Ayşen TANYERİ

This study examines the implications for agricultural and other households in Turkey of a switch away from existing agricultural policy mechanisms (principally support prices, input subsidies and trade restrictions) towards a system of direct income payments to households. 

The cornerstone of the analysis is the development of a Social Accounting Matrix (SAM) for Turkey.  The SAM emphasises the agricultural aspects of the economy and the linkages from agriculture to the rest of the economy.

A SAM describes the circular flow of income in the economy.  For example, in a simple closed economy setting, production generates value added, which is paid as wages and rent to factors of production.  These incomes are then payable to institutions including households and the government.  The incomes of institutions are then saved or spent on domestic production or imports.  The circle is closed when domestic consumption leads to further production.  The linkages become more complicated as greater detail on the economy is provided, but the principle of a circular flow of income still holds. 

The SAM is first and foremost a framework for organising economic data.

The first contribution of the SAM is that it describes the role of agriculture in the Turkish economy.  In particular, the SAM brings together data on the structure of the agricultural sector, its linkages to the broader economy, and the connections to the distribution of household incomes and expenditures.  It consolidates three important data sets: an input-output survey, a survey of household incomes and a parallel survey of household expenditures.

The SAM can also be used to consider the implications of alternative reform proposals.  The SAM developed in this study is designed specifically to address some of the key issues concerning Turkish agricultural policy.  Thus, in the disaggregation of productive sector accounts, detail is provided on the agricultural sector and agriculture-related industries both upstream and downstream from the farmgate.  Similarly, households are distinguished according to whether they are engaged in agricultural activities or not, whether they are located in rural or urban areas and according to their level of income.  This exercise in data organisation is of value in its own right.  The data may be used both as a source for future analysis and to address specific policy questions. 

The shortcomings of Turkish agricultural policy have been well documented, and there is considerable evidence to suggest that ‘decoupled’ direct income payments can deliver assistance more efficiently and more equitably.

Under the agricultural policies adopted in Turkey, the level of support provided to producers is linked to the volume of production.  The net economic costs of such policies are well known and there is specific evidence regarding the extent of these losses in Turkey (including that presented by the World Bank, the United Nations’ Food and Agriculture Organisation and the Organisation for Economic Cooperation and Development).

There is considerable evidence to suggest that decoupled agricultural policies, such as direct payments, can deliver income assistance more effectively than market based measures such as price supports, input subsidies and trade restrictions.  Not only do direct income payments induce fewer economic distortions than policies under which support is tied to production, they also mean that support can be targeted to deliver the desired distributional effect. 

Although policy reform is likely to lead to an overall improvement in economic welfare, any change from the status quo will mean that there are both winners and losers.

From a practical standpoint, it is important to have information on the distribution of agricultural incomes, and the distributional implications of policy reform.  Such information can assist in the design of a system of direct payments and the development of compensatory payments (to the extent that these are deemed to be necessary).

The distributional consequences of any change in policy depend fundamentally on the structure of the agricultural sector, and on the linkages between agriculture and the rest of the economy.  These factors are considered by the SAM in order to establish the pattern of winners and losers that is likely to be generated by policy reform.  The SAM may also serve as a first step towards the development of a database according to which eligibility for support can be determined. 

It is notable that the problem of low incomes in Turkey is more a rural issue than a specifically agricultural one.  This raises the question of whether agriculture-specific policies are the best way of supporting household incomes.

The income distribution data (which are incorporated into the SAM) show that, in rural areas, the incomes of agricultural and non-agricultural households are very similar, not just in terms of average incomes, but also in terms of the distribution.  The same is true in urban areas.  These similarities are evident despite basic differences in the sources from which agricultural and non-agricultural households derive their income.  The key distinction is between rural and urban households, with urban households likely to earn more than rural households, irrespective of whether they derive their income from agriculture or not.

If a goal of social policy is to support incomes at some minimum level, or to redress problems of income inequality, the data therefore suggest that policies should be targeted at rural households in general rather than just agricultural households. 

There are few linkages from agricultural households to rural non-agricultural households.

A 1 million TL increase in the expenditures of the average rural agricultural household causes the incomes of all households to rise by an estimated 1.48 million TL over and above the initial payment.  Of this increase, 0.33 million TL accrues to rural agricultural households and 0.25 million TL to rural non-agricultural households.  This contrasts with a benefit of 0.85 milllion TL to urban non-agricultural households.  The implication is that agricultural support cannot be relied upon to support the broader rural community through induced linkage effects. 

Agricultural households would benefit more from direct income payments than they do from price supports and input subsidies.

Although the total income effect of agricultural support exceeds the income effect of support to households (the multipliers being higher in the former case), households gain more from direct support because they benefit from the initial injection as well as from linkage effects.  Indeed, support to the poorest agricultural households leads to an increase in aggregate household income that is 50% higher than that achieved by an equivalent amount of support to the cereals sector. 

If support is restricted to low-income households, the economic costs of intervention can be cut dramatically.

The richest 20% of agricultural households receive nearly ten times more income than the poorest 20%.  The broad correspondence between farm income and farm size implies that the cost of using price supports to establish a minimum level of income is several times higher than the cost of a targeted direct payment scheme would be. 

The limitations of the SAM approach can be overcome by a more sophisticated model in which prices are flexible and by the modelling of household level decisions.

The SAM framework suffers from a number of limitations.  The most significant of these is the assumption that the economy is purely demand driven and there are no relative price effects.  This limitation can be overcome by a CGE model.  The SAM provide the basic database for a CGE model and the current framework could be adapted to consider the shift in resources from agricultural to non-agricultural activities when agricultural prices are reduced.

The accumulation of survey information on the behaviour of agricultural households may also be used to consider how agricultural households adapt in the face of structural change and policy change.  As agricultural prices fall, either as a result of policy reform or fundamental pressures under which the growth of supply tends to outpace increases in demand, some agricultural households and enterprises are likely to remain profitable, while others are not. Farm households which are not competitive are likely to respond either by diversifying their sources of income, or by exiting the sector altogether.  Household level analysis can provide important insights into the likely path of adjustment.

The data are now available both for the development of a CGE model and the specification of a range of household level models.  Together these may yield further insight into the behaviour of agricultural households in Turkey and how they are likely to respond to a more liberalised policy environment. 

A key policy recommendation is that the objective of income redistribution needs to be made more specific.

If existing policies are to be removed, then it is legitimate to worry about who will win and who will lose.  Indeed, a main worry about the removal of price supports and other output-related measures is that, whilst aggregate incomes may improve, many agricultural households will suffer a sharp drop in their incomes and there will be attendant social problems (such as accelerated rural-urban migration).  Our study suggests that such effects can be forestalled by the design of an appropriate method of compensation.

In general, the objectives of policy need to be reformulated in terms that are consistent with legitimate economic and social objectives and sufficiently clearly articulated that the performance of policy can be measured and compared with possible alternatives.  In particular, the objective of income support needs to specify the criteria under which households would qualify for support.  Given that, in rural areas, the incomes of non-agricultural households are generally no higher than those of agricultural households, it may not make sense for such criteria to be tied to agricultural production at all. 

A pre-requisite for a direct income payment scheme is a registry of households.

If a system of direct income payments is to be introduced, its viability will depend largely on the functioning of a household registry.  If the goal of policy is to compensate farm households for the removal of price supports, input subsidies and other subsidies, then payments based on historic farm size are one way of ensuring that the distribution of the benefits from change corresponds broadly to the pattern of losses.  In the long term, however, payments based on the level of household income can be targeted to low-income households.  Unfortunately, it is more difficult to build a registry under which selected agricultural (and perhaps non-agricultural) households provide full reports of their incomes, than it is to establish one based on farm size.

Neither the SAM, nor the income distribution survey around which it is partially constructed, provide enough information on which to base a system of direct payments.  However, they do provide information on the structure of the agricultural and rural economy, and this may inform decisions regarding the criteria under which households would qualify for income support.

Yayının tamamını alabilmek için lütfen bizimle irtibata geçiniz