close× Call Us
close×
Title Cooperation across income and racial categories: Pre-analysis plan
Post date 02/23/2017
C1 Background and Explanation of Rationale

Various research findings show that ethnoracially diverse communities have lower levels of solidarity and cooperation than homogeneous communities (Schaeffer 2014; Van der Meer and Tolsma 2014; Alesina, Baqir, and Easterly 1999; Alesina, Devleeschauwer, et al. 2003; Costa and Kahn 2003; Putnam 2007). A few reasons have been advanced: cooperation in diverse communities might be hindered by prejudice and outgroup discrimination, by cultural differences and disagreement concerning norms and preferred outcomes, or by weak social networks and difficulties in communication and social control across ethnic groups (Habyarimana et al. 2009; Koopmans and Veit 2014; Kimenyi 2006; Algan, H ́emet, and Laitin 2011; Miguel and Gugerty 2005; Lieberman and McClendon 2013; Enos and Gidron 2016).
However, ethnically diverse communities are often also poor communities, and thus diffused poverty, rather than ethnoracial diversity, might be at the basis of their lower cooperative capacity (Robert J. Sampson and Groves 1989; Robert J. Sampson, Raudenbush, and Earls 1997; Robert J Sampson 2012; Bruinsma et al. 2013; Abascal and Baldassarri 2015). Recent behavioral research suggests that scarcity produces a specific mind-set: in particular, people who are subjected to the stress of poverty 1) tend to discount the future more than people who do not live in conditions of chronic disadvantage, and 2) suffer from cognitive burden (Banerjee and Duflo 2011; Mani et al. 2013; Mullainathan and Shafir 2013; Haushofer and Fehr 2014). Here we theorize that these effects of poverty may influence cooperative behavior, as poor individuals might discard the future benefits of cooperation, and might be more erratic in their decision-making. In addition, poverty may also affect expectations: poor people might be considered less reliable cooperation partners, and might have fewer opportunities to participate in cooperative endeavors in the first place. In sum, poverty, as well as diversity, may be related to lower levels of cooperation. We turn to an experimental design to disentangle these two factors.
To test various hypotheses concerning the relative effect of ethnic diversity and poverty on cooperative behavior we developed a cooperative investment game and experimentally manipulate the identity of the interaction partner along racial and income lines. In addition, we compare behaviors in the cooperative investment game and a dictator game to distinguish between an underlying discriminatory bias against racial out-groups (and the poor) and discriminatory behavior based on expectations about alter’s behavior. Taken together, our results will contribute to the understanding of the mechanisms that bring about cooperation in diverse communities. Given that poor people often live in communities that are both poor and racially diverse, our research findings might have far-reaching consequences for overall cooperation levels in contemporary societies.

C2 What are the hypotheses to be tested?

We will test hypotheses concerning the causal effect of alter’s gender, income, and race in a bilateral cooperative investment task. Our major hypotheses concern the causal effect of alter’s identity – namely his/her ethnicity and economic status – on ego’s cooperative behavior. First of all, we expect that participants matched with a ‘poor’ alter will be less likely to invest. We speculate that this is because poor people tend to be more present-biased, while cooperation typically requires foresight and appreciation of future gains. Assuming that participants are (intuitively) aware of this, they will be more hesitant to invest when interacting with poor rather than rich subjects.

Based on previous observational research, we also expect to find a lower propensity to invest in the cooperative game when participants are matched to Black partners. Lower levels of cooperation with ethnic minorities can be due to a variety of reasons, from imbued prejudice towards them (e.g., dislike of Blacks) to specific expectations about minorities’ strategic behavior in cooperative contexts. If negative feelings toward Blacks were driving behavior, we would find lower contributions to Black partners in the dictator game and also find that contributions in the dictator game largely predict behavior in the cooperative investment game. In contrast, if lower cooperation is mainly due to expectations about Blacks’ behavior, then we need to figure out whether these negative expectations are related to ethnicity per se – i.e., the belief that Blacks are unreliable – or whether ethnicity is used as a proxy for other factors, such as poverty.

In particular, it is possible that part of the discriminatory behavior toward Blacks is driven by considerations regarding the economic conditions of most Blacks. In the latter case, varying information about the economic conditions may affect these expectations. Our research design, allowing us to compare cooperation with high- and low-income Blacks and Whites is mainly intended to address this aspect. Namely, if ethnicity is simply a proxy of economic conditions, we expect that there are no differences in levels of cooperation towards Blacks and Whites once we consider their income status. In contrast, if negative expectations about Blacks exist, beyond economic considerations, we would find that blacks elicit less cooperation than Whites who are in their same income category.

C3 How will these hypotheses be tested? *

In order to measure cooperation, we developed a cooperative investment game. In the cooperative investment game, participants are given an endowment, and will have to decide whether they want to keep the endowment, or whether they want to invest their endowment. Keeping the endowment results in participants having the amount paid out immediately. Investing promises a 33% return on their investment after a two weeks wait. This return is only realized if alter chooses to invest, too, however. If the partner chooses not to invest, participants lose 20% of their endowment and still have to wait two weeks to be paid.

However, our cooperative investment game adds a time dimension to what is basically a stag-hunt game (Skyrms 2004). This time dimension captures the extent to which individuals tend to discount future gains in favor of an immediate reward. According to recent psychological research (Mullainathan and Shafir 2013; Mani et al. 2013; Haushofer and Fehr 2014), this ‘present bias’ is quite common among the poor, and, we argue, could be at the origin of their reduced propensity to cooperate. The game was intentionally constructed this way to give us a measure of cooperation that sensitively depends on expectations about alter’s willingness to cooperate, but also to reflect the fact that many real-life cooperation situations involve a time dimension.

In terms of data analysis, all hypotheses will be tested using OLS. For the cooperative investment game, we will additionally provide probit estimates. We use heteroskedasticity-robust (White) standard errors to account for the fact that our treatment groups will most likely be slightly unbalanced (as we use a simple randomisation procedure, where participants are allocated to conditions as they arrived) (Samii and Aronow 2012). We will control for pre-treatment characteristics of our subjects (sex, age, race, education, income, household size, occupation) to improve the precision of our estimates.

We will test hypotheses concerning the causal effect of our treatments, as well as gather empirical evidence from observational data, and estimate heterogenous treatment effects for subgroups of our sample. Our major hypotheses concern the causal effect of alter’s identity – namely his/her ethnicity and economic status – on ego’s cooperative behavior. First of all, we expect that participants matched with a ‘poor’ alter will be less likely to invest. We speculate that this is because poor people tend to be more present-biased, while cooperation typically requires foresight and appreciation of future gains. Assuming that participants are (intuitively) aware of this, they will be more hesitant to invest when interacting with poor rather than rich subjects.

Based on previous observational research, we also expect to find a lower propensity to invest in the cooperative game when participants are matched to Black partners. Lower levels of cooperation with ethnic minorities can be due to a variety of reasons, from imbued prejudice towards them (e.g., dislike of Blacks) to specific expectations about minorities’ strategic behavior in cooperative contexts. If negative feelings toward Blacks were driving behavior, we would find lower contributions to Black partners in the dictator game and also find that contributions in the dictator game largely predict behavior in the cooperative investment game. In contrast, if lower cooperation is mainly due to expectations about Blacks’ behavior, then we need to figure out whether these negative expectations are related to ethnicity per se – i.e., the belief that Blacks are unreliable – or whether ethnicity is used as a proxy for other factors, such as poverty.

In particular, it is possible that part of the discriminatory behavior toward Blacks is driven by considerations regarding the economic conditions of most Blacks. In the latter case, varying information about the economic conditions may affect these expectations. Our research design, allowing us to compare cooperation with high- and low-income Blacks and Whites is mainly intended to address this aspect. Namely, if ethnicity is simply a proxy of economic conditions, we expect that there are no differences in levels of cooperation towards Blacks and Whites once we consider their income status. In contrast, if negative expectations about Blacks exist, beyond economic considerations, we would find that blacks elicit less cooperation than Whites who are in their same income category.

C4 Country USA
C5 Scale (# of Units) ~1,200
C6 Was a power analysis conducted prior to data collection? Yes
C7 Has this research received Insitutional Review Board (IRB) or ethics committee approval? Yes
C8 IRB Number 11/07/2016
C9 Date of IRB Approval 7 Nov 2016
C10 Will the intervention be implemented by the researcher or a third party? Researchers
C11 Did any of the research team receive remuneration from the implementing agency for taking part in this research? not provided by authors
C12 If relevant, is there an advance agreement with the implementation group that all results can be published? not provided by authors
C13 JEL Classification(s) C90, D71, I32, J71, Z13