9 Treatment of Agriculture

Liam Ebrill, Michael Keen, and Victoria Perry
Published Date:
November 2001
  • ShareShare
Show Summary Details

The agricultural sector often receives special treatment under the VAT, not only in developing countries. Why? This chapter considers the proper treatment of this important sector.


As the results of FAD’s VAT survey showed, practice in the treatment of agriculture varies relatively little, with the large majority of countries surveyed exempting agriculture and a minority, all transition economies, taxing it (though generally with an exemption for small traders). Many countries also exempt agricultural inputs—meaning, typically, fertilizers, pesticides, seeds, farm machinery—and two zero-rate them.91 In contrast to practice, the advice offered through FAD technical assistance has varied, with both exemption and taxation being recommended.

The survey picture appears broadly typical of the more general one in developing countries, with various forms of special treatment extended to the agricultural sector. In developed countries also, the sector has often received special treatment under the VAT (as described, for instance, in Tait (1988)). We have seen too that the revenue increase associated with the presence of a VAT is lower, all else equal, in countries where agriculture accounts for a larger share of national output. The interpretation of this result is not entirely clear. A high agricultural share may proxy for a relatively low level of administrative development. But the result is also consistent with explicit exclusion of agriculture from the VAT system. Whatever the reason, it seems a common feature of practice that value added in agriculture is taxed relatively lightly.

Distinctive Features of the Sector

The specific difficulties that the agricultural sector raises for VAT design come from its combining two features:

  • In many developing countries, most agricultural producers are outside the formal sector. Even among farmers who do work within the formal system, it may be only the largest who keep records sufficient for an accurate measurement of annual turnover. Physical remoteness may further increase the difficulties of monitoring tax compliance by farmers. Agriculture is also marked by issues of seasonality and mismatched timing between inputs and outputs that complicate both measurement and payment procedures.92 In short, compliance and administration costs for agricultural producers are likely to be high.

  • The agricultural sector is often a particular concern in the pursuit of wider distributional objectives. A tax on food is either borne by consumers in the form of a higher price or by producers in the form of reduced real income. The former would be widely perceived as regressive, at least in relation to basic foods (such as unprocessed vegetables, meat, and fruit); so too would the latter, insofar as producers are thought to have low incomes.

The first of these points is commonly agreed. But similar collection problems are associated with small traders, raising the question of whether there should be any special measures for agriculture beyond the setting of a high threshold applicable to all enterprises.

Views differ on the importance of the second consideration. The distributional argument on the consumption side clearly only applies to basic foods. Moreover, expenditure policies may be better targeted to poverty relief than departures from uniform taxation: as seen in Chapter 10, the amount of redistribution that can be achieved through indirect taxation is limited. The distributional argument on the production side may also be disputed. In the absence of a properly functioning income tax—even (in some cases, especially) in relation to large and prosperous farmers—implicit taxation of agricultural producers through the indirect tax system may be the only prospect for ensuring that they bear a reasonable share of the tax burden.

Irrespective of one’s view on this latter issue, including whether one wants to tax agricultural products at regular rates or at low ones (possibly zero), the implication is that agricultural products should be fully within the VAT system: either in order to charge them at regular rates or, on the contrary, to exclude them from tax altogether (by ensuring that tax on inputs is refunded). In this sense the long-term objective in the treatment of agriculture is clear: tax agriculture as any other good, subject to the normal threshold.

In the interim, however, high collection costs may validate other methods. Specifically, the exemption of agricultural products is a potential compromise measure. Exemption excludes farmers from the administration of the tax altogether while ensuring that agricultural products are taxed at a reduced but positive rate because of the absence of relief for any tax paid on inputs, notably, seeds, fertilizers, pesticides, and animal fodder. In efficiency terms, this will distort production decisions away from these inputs and toward socially excessive use of untaxed inputs. However, there may be environmental grounds for imposing tax on the use of some agricultural inputs, such as fertilizer and pesticides to achieve exactly this result. A more obvious implication of unrelieved tax on inputs is simply the implied tax burden on final agricultural products. This could result in some foodstuffs being more heavily taxed than they otherwise would be, as in the case of agricultural products subsequently used as inputs into taxable activities (perhaps, the production of processed foodstuffs). Exports too will suffer from some tax burden.

Move Toward Full Taxation

There are two broad responses to these difficulties. One is to remove the exemption on basic agricultural outputs. The other is to seek to alleviate the tax burden arising from input taxes. In practice, the latter seems to be the preference of policymakers. There are several ways in which the burden of input tax might be mitigated.

  • Intellectually, the cleanest is to zero-rate agricultural inputs (as recommended by both Tait (1988) and Due (1990)). It is important to strictly limit the zero-rating to items truly specific to agriculture. Zero-rating machinery, equipment, or spare parts that have multiple uses, or conditioning the zero-rating on final use rather than the nature of the item in question, creates opportunities for avoidance and evasion. For example, zero-rating machinery invites attempts to redesignate as such other machinery likely to be used by exempt traders. When tax administration is weak, the danger can be significant. In order to target the relief at inputs to small farmers, and to restrict revenue losses, it is best to restrict any zero-rating of inputs to fertilizers, seeds, and perhaps pesticides. Farm-specific machinery is likely to be little used in a culture of subsistence agriculture, and so best taxed.

  • Special schemes might be adopted that enable some credit to be claimed by farmers or by their taxable clients. Under the former heading, for example, refunds could be granted for large purchases. Schemes of the latter kind include the flat rate scheme used by a number of EU countries, under which purchasers from exempt farmers are able to show payment of tax on the tax invoice they receive from that farmer, and then credit the tax shown against their own liabilities, even though none was withheld by the farmer. The latter scheme provides the small farmer relief from input taxes without requiring registration. However, the scheme is open to abuse by overstating values traded on invoices; and it only relieves input tax in relation to sales to registered traders.

  • Agricultural inputs might be exempted. As with zero-rating, it is important to delimit the exemptions carefully. This approach is quite widespread in developing countries, and is a prime instance of the danger of “exemption creep” discussed in Chapter 8: the presence of one exemption (agricultural products) creates pressure for the exemption of its inputs.

Technically, the last of these is likely to be the best approach when tax administration is weak: zero-rating inputs and the special schemes both create greater opportunities for fraud. The danger is that exemption contains no impetus for movement toward full taxation; indeed the danger of exemption creep points in the opposite direction.


There is nothing inherent in agriculture that implies that the sector should not be subject to VAT. Collection difficulties, distributional concerns, and political considerations, however, mean that exemption has proved an irresistible option for many countries. Against such a background, it is important to develop a well-articulated strategy for this sector. Elements of the strategy should include levying VAT on the inputs to agriculture and applying a high threshold to bring large farmers within the VAT net. Small farmers otherwise placed at a disadvantage should be allowed to register voluntarily.

The apparent absence of any tendency for countries to progress toward full taxation is disconcerting. If anything, the tendency is for some degree of exemption creep. Moreover, the experience in developed countries is not entirely encouraging: many EU countries, in particular, continue to operate special schemes for agriculture. Developments in New Zealand and Denmark to fully tax agriculture in the normal manner show, however, that the special treatment of agriculture is not inevitable.

    Other Resources Citing This Publication