Variable compliance incentives in conservation policies in a dynamic setting

Abstract

The analysis jointly considers the dynamic aspects of conservation programmes with incomplete compliance and allows land users' behaviour to change over time. A distinction is made between compliance and participation incentives. Under conditions that are frequently observed in practice, we find that compliance incentives of programme participants are variable over time, which can have a significant impact on the efficacy of conservation policies. We match existing enforcement policies with the rational actor model to demonstrate the effects of different kinds of penalty and monitoring schemes for a range of programme participant characteristics. Land users are shown to have sizeable incentives to participate in compensation schemes, but their incentives for compliance with programme requirements are considerably lower.

Keywords:

• compliance
• monitoring and enforcement
• policy instruments
• conservation policy
• penalty schemes

Notes

1. In reality, start-up costs tend to be higher than continuing compliance costs, because they include learning and conversion costs, changes in suppliers or fixed investment costs such as building fences or planting trees. Note that these costs do not include the potential loss or gain in land revenues. The changes in land revenues are modeled separately to make the distinction between the two types of regulatory effects clear: the first is the direct costs of implementing the regulation and the second is an indirect effect on potential land revenues.

2. If the regulator would perfectly know each land user's compliance cost, he would be able to perfectly predict which land users would be compliant and which not. Thus, if the available budget is sufficiently large, monitoring would be perfect and all potential violators would be deterred.

3. This is of course not the only way of modelling the changes in land values over time and we use a normalisation. However, it is the difference between the values of the implementing conservation measures or not that is crucial to the model.

4. We do not incorporate targeting inspection strategies into the model since would not significantly change the results with respect to the impact of having cumulative penalties and the option of banning detected violators from the program. Harrington (1988) showed that state dependent monitoring increases deterrence compared to random monitoring under a binding budget constraint. Also, for a discussion of the potential of targeting strategies in conservation policy see Fraser (2004). We assume that in each period, inspections are randomly executed among all participants. So land users inspected in a previous period and found in compliance can be inspected again in the next period. However, this random inspection probability p_t can vary between periods.

5. Note that we assume that the agricultural yield from the land is an exogenously given function. Obviously, this is a simplification since in practice land yields are uncertain, risky outcomes depending on any number of factors such as whether conditions. In this paper, we focus on the direct impact of having a cumulative penalty and do not look at interactions occurring as a result of introducing two types of uncertainty (i.e. agricultural yield and fines) in the land users' utility function. However, this is certainly an interesting extension to study.

6. The second and third conditions can be met since the possible fine is limited to the cumulative sum of past payments.

7. DCP stands for the Direct and Counter-Cyclical Payment Program and it falls under the 2002 US Farm Bill.

8. Previously, the selection of policy instruments for conservation has been studied by, among others, Babcock et al. (1997), Wätzold and Schwerdtner (2005) and Moons and Rousseau (2007).

9. Fines are limited in practice because firms' wealth is limited, measurement and judicial errors occur and due to justice considerations (see among others, Polinsky and Shavell 2000).