Association for Public Policy Analysis & Management

Co-production emphases the active role of citizens, government agencies, and professionals in the delivery of public goods and services (Ostrom 1996; Parks 1981). The nature of the good, and the incentives for citizens to participate in co-production are critical elements for success, and thus the focus of substantial literature on the subject (Jakobsen 2013, Bovaird 2016, Parrado 2013, Loeffler 2008, Riccucci 2015). Co-production processes, however, can face limitations, including free-riding, public accountability, blurring of the public, private, and voluntary sectors, and differences in values among co-producers (Bovaird 2007, Mayor and Moore 2002). In this paper, we present an experimental test for the effect of market-based incentives and program stringency on the level of collective or public goods co-production. The experimental context is the co-production of climate mitigation benefits from forest-based activities. Forest carbon offset programs are market-based strategies to incentivize the co-production of public goods (carbon storage and sequestration), with participants earning marketable credits for activities resulting in carbon emission reductions. Carbon offset programs are a particularly interesting example, where the producers and beneficiaries of co-production are not necessarily collated in their geographic location, and where local co-production activities result in globally-relevant collective benefits.

Both offset program standards and market factors have been found to shape willingness to enroll forestland in a carbon offset program (Markowski-Lindsay et al. 2011, Wade and Moseley 2011). In this study, we hypothesize that the level of inputs to co-production (amount of land enrolled in the program) will be: (1) positively related to expected income from the carbon market, and (2) negatively related to program standards for permanence of offsets, operationalized as contract length. Further, we expect these relationships to be mediated by preferences for the sector administering the program (public, private, or non-profit organization).

Methodologically, we use a 3x4 between-subjects survey experiment to examine the effect of expected income (less, same, more) and contract length (none, 15, 40, 100) on amount of land enrolled in the program. Additionally, we test how preferences for sector administering the program mediates this relationship. Data come from two survey experiments: Study 1 based on a large 2016 Mturk sample (n=1,672); and, Study 2 using a 2018 nationally representative sample (n=1,000). Results from Study 1 suggest that participants are willing to enroll half or more than half of their land under the same level of expected income (as that received under a baseline timber management scenario) when contract terms are short. Non-profit and public sector agency preferences increase the mean level of co-production, while preference for private organizations has a negative effect. The paper will summarize and compare the results from the two studies, and offer a set of recommendations. The findings have relevance for decision-makers and administrators interested in increasing levels of participation in carbon offset programs, as well as other policies that target co-production of environmental public goods.