Agriculture, Behaviour, Economics, Latest, Natural resource management, People, Policy

332. Farmer behaviour and agricultural policy

An understanding of farmers’ adoption of new practices is central to the design of effective and efficient agricultural policies. Aspects of agricultural policy that can be enhanced by good information about adoption include the design of the policy, the targeting of policy effort, and the assessment of additionality. 

In PD330 I advertised a new Special Issue on adoption of agricultural innovations in the journal Applied Economic Perspectives and Policy. There is an audio interview with me about the Special Issue available here.

One of the papers, by Roger Claassen and me, focuses on the relevance to agricultural policy of understanding farmers’ decisions about taking up new practices.

One simple reason for this relevance is that much agricultural policy is concerned with getting farmers to do something they are not already doing or would not otherwise do. Examples of such policies include the following:

  • Programs of agricultural extension to encourage farmers to adopt a new technology that is believed to be more productive than the existing technology farmers are using (e.g., a higher-yielding crop variety);
  • Programs that pay farmers to adopt a practice that generates public benefits (such as protecting or planting vegetation that provides habitat for wildlife);
  • Policies that fund agricultural research, with the intent of generating information or technologies that will be beneficial for farmers or the community; and
  • Policies that use regulations to constrain the behaviour of farmers (such as regulations on clearing of native vegetation).

Since all these policies are about influencing the behaviour of farmers, of course it makes sense that their design and implementation could be enhanced if the designers had a good understanding of what influences the behaviour of farmers. Sometimes policy makers do take this seriously, but not always. I’ve been critical of Australian agri-environmental policies, for example, for often being making overly optimistic assumptions about what farmers will do if we just provide them with some information or pay them a little bit.

In practice, some practices are more attractive to farmers than others. Zero till is used by about 90% of Western Australian farmers, but a practice like variable-rate precision agriculture is used by only a minority. Any one practice is more attractive to some farmers than to others due to varying local conditions, such as rainfall or soil types. Being able to predict variations in adoption would be very helpful to policy makers for targeting their resources. Having a sense of which practices can potentially be adopted, and where, is one of the factors that ought to influence where policies like extension or incentive payments are applied.

Another policy concept that is tangled up with farmer behaviour is additionality. As we say in the paper:

A conservation action (and the resulting environmental gain) that is supported by a payment is additional if the farmer would not have taken the action if he or she had not received the payment. Environmental gains that flow from nonadditional actions cannot be attributed to the incentive program.

If the additionality of a proposed agri-environmental payment scheme is too low, it’s not worth running the scheme. Most farmers were going to adopt the practice anyway, so any incentive payment to them is just a gift, making no difference to environmental outcomes. Policies to promote some practices have high additionality (e.g. filter strips or cover crops in the U.S.) while others have much lower additionality (e.g. conservation tillage in the U.S.).

Assessing additionality is essentially about predicting behaviour. In fact, for programs that aren’t in place yet, assessment of additionality required two predictions about behaviour: what will farmers do if the program does offer them the proposed incentive scheme, and what will they do if there is no incentive payment scheme. The difference between those two predictions tells you the additional change that is attributable to the scheme.

Even assessing an incentive scheme that is already well established requires a sort of prediction. You can observe what farmers are doing with the scheme in place, but to assess additionality you still need to estimate what they would have done in the absence of the scheme.

For all these reasons, an ability to understand and predict farmers’ adoption of new practices is critically important to agricultural policy makers if they want their policies to be effective and efficient.

Further reading

Pannell, D.J. and R. Claassen. 2020. The Roles of Adoption and Behavior Change in Agricultural Policy. Applied Economic Perspectives and Policy 42(1), 31-41.