The problem of free riding in voluntary generic advertising: parallelism and possible solutions from the lab.

Most agricultural commodities in the United States have programs assessing producers for generic advertising and promotion, a public good for producers. Examples of generic advertising campaigns include "Beef, It's What's For Dinner;" "Got Milk?;" "Dancing Raisins;" and "The Incredible Edible Egg." Designed to increase the overall market demand for firms within an industry, generic advertising and promotion programs are usually found in industries producing homogeneous commodities with little potential for product differentiation. In the United States, the budgets for these programs total more than $1 billion annually (Forker and Ward 1993). Originally, participation in many of these programs was voluntary, in that revenue came from donations from willing participants using a version of the Voluntary Contributions Mechanism (VCM). Consistent with theory and past experimentation, free riding increased dramatically over time as contributions decreased, thereby creating a significant policy problem for producers.

The eventual failure of the VCM to provide a sufficient level of support for generic advertising led many programs to hold referenda with relevant producers to establish mandatory programs. Consequently, nearly all programs in operation today are mandatory. However, the constitutionality of mandatory programs has been under attack with three cases heard by the U.S. Supreme Court in the last decade. These challenges have led to a variety of decisions handed down. In 1997, the Court ruled in Glickman v. Wileman Brothers and Elliott, Inc. that the California peaches, plums, and nectarines checkoff program did not violate the plaintiff's right to free speech since generic advertising was part of a larger set of regulations designed to help producers. In 2001, the Court ruled in United States v. United Foods, Inc. the mushroom checkoff program unconstitutional because it was a stand-alone checkoff advertising program and not part of a broader set of economic regulations. In 2004, three of the most successful and visible mandatory generic advertising programs--for beef, pork, and milk--were ruled unconstitutional by federal appeals courts. In 2005 in Johanns v. Livestock Marketing Association, et al., the Court ruled the beef advertising checkoff program constitutional because the advertising was considered "government speech." Despite the recent ruling, these programs remain vulnerable to legal challenges on new grounds, such as arguments related to freedom of association (for a review, see Crespi and McEowen 2006).

One might question the social importance and magnitude of under-provision of advertising for generic commodities. However, contrast the public health impacts from the types of foods associated with the majority of branded advertising, such as soda, beer, chips, and candy, to the types of foods that now benefit from mandatory generic advertising, such as fruits, vegetables, nuts, chicken, pork, beef, and milk. Not only do the generic commodities comprise the key nutritional elements of the United States Department of Agriculture food pyramid but these commodities also tend to be low in fat and salt (in comparison to branded snack foods and restaurant meals) and represent the bulk of what might be called the components of a healthy diet. If generic advertising for agricultural commodities collapses because mandatory programs are declared unconstitutional, the "Dancing Raisins" will be gone and the vast majority of ads for snacks will be for chips, cookies, and candy. Given important health problems such as obesity, juvenile diabetes, and osteoporosis, the under-funding of generic commodity advertising has serious public health consequences. (1)

The purpose of this experimental study is to investigate the effectiveness of alternative mechanisms for voluntarily funding generic advertising. Four research questions are examined. First, to what extent can field conditions be replicated in a laboratory setting using student subjects? This experimental research attempts to capture the institutional details of the egg program from both an economic and a psychological perspective. This careful calibration allows for an assessment of the degree to which parallelism holds by testing whether an experiment can replicate actual field participation rates over time (rounds) in a voluntary contribution (assessment) program. While contribution similarities have been demonstrated between lab subjects and subjects recruited from the field (List 2004), these similarities were observed in situations that lacked key context effects such as communication and the donation mechanisms that occur in the real world. This parallelism question is important, because if parallelism holds, then the laboratory results for other funding mechanisms will have good predictive power for what might be observed in the field.

Second, are there viable voluntary mechanisms to replace the VCM or current mandatory ones? The dilemma is that mandatory programs generally are highly effective in generating positive net benefits to producers, but voluntary programs using the VCM have been plagued by free riding. One obvious alternative is the provision point mechanism (PPM), which has been shown to have desirable theoretical properties by Bagnoli and Lipman (1989) and has been tested extensively in the laboratory and in the field (Dawes et al. 1986; Isaac, Schmidtz, and Walker 1989; Bagnoli and McKee 1991; Marks and Croson 1998, 1999; Krishnamurthy 2001; Poe et al. 2002; Rondeau, Poe, and Schulze 2005). Under this mechanism, the commodity promotion program would still be voluntary but generic advertising would occur only if at least a certain percentage of producers (the provision point) made contributions to it. If the percentage of producers making contributions was less than the provision point, all participants would receive a complete refund (a money-back guarantee) and no advertising program would be implemented. In laboratory experiments as well as in the field, PPMs have been shown to significantly lessen free riding compared to voluntary contributions. Using the PPM for voluntary programs may be advantageous because this funding mechanism would not be subject to the current legal challenges and it could reduce the degree of free riding seen in previous voluntary programs.

The third and fourth questions are whether creating a status quo of giving significantly increases voluntary contributions for generic advertising in the VCM and PPM. Contributions to generic advertising programs have typically started at a much higher level than the traditional VCM. For example, producer contributions to fund the American Egg Board's generic advertising started at 90.7 % in the first year of its voluntary program and declined to 49.0% after eleven years. This pattern is significantly higher than in public good economics experiments, where initial contributions are usually around 50% and quickly decay over time, especially when the marginal per capita rate of return (MPCR) is significantly less than one (see, for instance, Isaac, Walker, and Thomas (1984); Andreoni (1988); Messer et al. (2007); and others described in Ledyard (1995)). It is possible that the relatively high contributions resulted from the peculiar version of the VCM used to fund generic advertising, where all producers were assessed a per unit amount on sales but producers could request their money back by submitting a refund application. The contextual shift of requiring a refund by request changes the status quo from not contributing, as in the standard VCM, to contributing. Status quo bias suggests that the contribution rate should be increased by this contextual shift since decision makers are reluctant to leave the status quo even in the face of economic incentives (Samuelson and Zeckhauser 1998; Kahneman, Knetsch, and Thaler, 1990, 1991). A number of additional studies have demonstrated status quo bias in insurance choice (Johnson et al. 1993), pension savings (Madrian and Shea 2001), Internet privacy (Johnson, Bellman, and Lohse 2002), and organ donation (Johnson and Goldstein 2003).

Three related studies have recently examined the viability of voluntary funding mechanisms for generic advertising. Messer, Schmit, and Kaiser (2005) used experiments to investigate various aspects of the PPM and generic advertising, including optimal provision point levels--from both a producer surplus and a contribution perspective--and the role of producer referenda in increasing voluntary contributions. Norwood et al. (2006) surveyed beef producers to examine the feasibility of using VCMs or PPMs to fund the beef checkoff program and argued that the PPM may be inefficient because whenever the provisionpoint threshold is not reached, all funds are returned to producers and no advertising occurs. Messer, Kaiser, and Poe (2007) examined whether option assurance in the form of a two-tiered threshold (a lower one for administration and a higher one for advertising) improved contributions and producer welfare in funding generic advertising. This article contributes to this literature by testing for parallelism between the field and laboratory, examining the role of status quo bias on voluntary contributions, and exploring how PPMs can improve funding for voluntary generic advertising programs.

A Model of Generic Advertising