How Corporate Sociopolitical Activism (CSA) impacts portfolio allocations: an experiment

Abstract

If a firm signals that they identify on one end of the conservative-liberal spectrum, will political affiliation help predict how an investor will allocate their investment dollars to that firm? Using an experimental survey design with 1,494 participants, evidence was found that political affiliation does materially predict the allocation decisions of investors to firms engaged in corporate sociopolitical activism (CSA). More specifically, Democrats were more likely than Republicans to allocate investment dollars towards firms signaling a liberal political identity through their support of LGBT social issues and Planned Parenthood. By comparison, Republicans were more likely than Democrats to allocate investment dollars towards firms signaling a conservative political identity through their support of the Religious Freedom Institute. In addition, evidence was found that both Democrats and Republicans will under-allocate (relative to a control group) to firms that engage in CSA. Corporations should consider the potential costs from investor under-allocation for taking public stands on controversial partisan issues that favor some constituents at the expense of others.

Introduction

Relative to any time in history, the United States is more ideologically polarized today than ever (Pew Research Center, 2020). There are numerous polarizing issues in society today from abortion rights to gun control to LGBT rights (Baldassari & Gelman, 2008; McCarty, 2019; Perry, 2022) to the more recent topic of teaching critical race theory in public schools (and many more issues). Firms that take a public stance on these issues can favor some constituents at the expense of others. For example, over 200 corporations have joined the Business Coalition for the Equality Act, which passed in the House 224 to 206 with only three Republicans in support, to promote equality and employment, housing, and credit protection to individuals based on sexual orientation and gender identity.

In the meantime, investors are becoming more active in their selection of investments in that align with their moral and ethical compass (Mohliver & Hawn, 2020). Some investors will react positively to information that supports their stance on a contested issue while other investors may react negatively to the same information. Individuals are deeply divided on polarizing issues, with some holding strong views for and some holding strong views against them (DiMaggio et al., 1996).

When a firm takes a public stance on a politically charged issue, this is known as corporate sociopolitical activism (CSA; Bhagwat et al,. 2020). This paper seeks to add to the relatively new, and growing, literature of CSA. The primary research question of this paper is: how do investor portfolio allocations change to signals that a firm supports a partisan position on a controversial social issue? This paper will use social identity theory of value congruence to predict that firms that engage in CSA issues should expect investors to align their investment allocations based on their political party’s affiliation to accrue the benefits of more value congruence. While prior studies have focused on trying to measure the stock market impact after a firm has signaled that it is engaged in CSA (Bhagwat et al., 2020; Gomes, 2021; Mohliver & Hawn, 2020), this is the first study to apply a survey experiment methodology to isolate the implications of CSA on investor behavior based on respondents’ self-assessed political affiliations.

As corporations consider wading into the water of sociopolitical issues, this paper explores the impact on investor’s portfolio allocation decisions. CSA could entail a cost to corporations if investors would allocate their portfolios away from corporations who have chosen a partisan issue that is contrary to their political affiliation. This study finds experimental evidence that an investor may potentially under-allocate to companies engaged in CSA up to 9%. Given that CSA may alienate a material portion of the investor base, a more inclusive course of action would be Corporate Social Responsibility (CSR), which has been shown to benefit multiple stakeholders (Mishra & Modi, 2016), including the firm (Baron, 2001; Nalick et al., 2016) and the firm’s shareholders (Hoepner & Schopohl 2020).

Literature review

Defining CSA

CSA refers to a firm’s public demonstration (statements and/or actions) of support for or opposition to one side of a partisan sociopolitical issue (Bhagwat et al., 2020). Sociopolitical issues may be defined as an issue where opinions are decisively split and will likely lead to contentious debate among the split groups (Nalick et al., 2016). CSA is a relatively new type of firm activity (Hambrick & Wowak, 2021). CSA is different from Corporate Social Responsibility (CSR) and Corporate Political Activity (CPA). Kang et al. (2016, pg. 59) defined CSR as “company actions that advance social good beyond that which is required by law.” These actions include “triple-bottom-line” reporting where a company will report certain stakeholder indicators that are not required by traditional U.S. GAAP accounting, such as environmental, social, and governance data (Slaper et al., 2011). The goal of CSR is centered around making a firm more sustainable through considering all important stakeholders and company resources over a longer period of time that which is considered in traditional profit and loss reporting (Slaper et al., 2011). By comparison, CPA involves corporate political activity at the federal and state level where economic benefits are accrued through trying to influence legislation (Bhagwat, et al., 2020). A primary goal of CPA is to further a specific goal that involves more direct financial payoffs versus gaining indirect payoffs from supporting social causes (Hillman et al., 2004).

When comparing CSA, CSR, and CPA, Bhagwat et al. (2020) provided a useful 2 x 2 conceptual model where publicity is one dimension while partisanship is the other dimension. CSR is low in partisanship and high, or low, in publicity. CPA is high in partisanship and low in publicity. Finally, CSA is high in partisanship and high in publicity. CSR is considered low in partisanship given that Mishra and Modi (2016) found that it is intended to improve relationships with most stakeholders. Regarding the other dimension, publicity, CSR may be high, or low, since many CSR practices are done routinely, with little fanfare (e.g., producing quarterly sustainable report), while other CSR practices may be done with high promotion (e.g., introducing new eco-friendly products). CPA, by contrast, is high in partisanship since it must rely on political affiliations to influence public policy. CPA is considered low in publicity since it is most often effective when done quietly, behind closed doors (Lux et al., 2011). Finally, CSA is high in partisanship (like CPA), but does so in a way that has maximum publicity so that the firm’s values are clearly communicated to the public (Kotler & Sarkar, 2017).

CSA Studies

Studies are mixed about the impact to a firm’s economic value when engaging in CSA. Bhagwat et al. (2020) analyzed 293 CSA events initiated by 149 firms across 39 industries and found that on average, CSA elicits an adverse stock market reaction from investors. A few examples of the 293 CSA events that were studied include: Chipotle deciding to prohibit guns in their stores, Amazon removing Confederate flag merchandise from its website, and Target supporting national LGBT pride month. When compiling all 293 of these CSA events together, the authors found that there were abnormal negative changes in the stock market return of the sample that were not explained by traditional stock market factors. The authors predicted a negative reaction from investors using signaling and screening theories, which would argue that CSA is a signal from a firm that a firm is redirecting resources from profit-oriented objectives to more risk activities. By contrast, Gomes (2021) found marginal evidence of a positive stock market return for companies that made statements condemning racism after the death of George Floyd on May 25 of 2020. Wintoki and Xi (2020) found that fund managers are more likely to allocate assets to firms managed by executives and directors with whom they share a similar political partisan affiliation.

Mohliver and Hawn (2020) conducted an event study that predicted how investors would respond to the new information that a company scored highest or lowest on the publicly available Corporate Equality Index (CEI). The CEI ranks the largest U.S. firms on LGBTQ policies and practices, showing a list of the best performing companies and the worst performing companies. The authors found that, on the day the index is released, the best and worst performing companies experienced positive cumulative abnormal returns, whereas the companies ranked in the middle did not. The authors also found that firms headquartered in liberal states would have positive abnormal investor returns if newly listed with a high CEI score while firms headquartered in conservative states would have positive abnormal investor returns if newly with a low CEI score.

Bailey and Phillips (2020) conducted a survey experiment of 168 managers and MBA students seeking to understand consumer attitudes to CSA. Participants were all given baseline information about a mid-sized food service company and asked whether they would buy a product from the company or work for the company. Groups were then randomly selected to either a conservative or liberal experiment group. Most participants who were told the company had conservative values (conservative group) caused participants to lower their opinion of the company by 33% (completely driven by a change in opinion by the self-identified Democrats who made up 34% of survey participants). In addition, participants were 26% less likely to buy its products and 44% less likely to apply for a position there. However, when participants were told the company had liberal values (liberal group) there was no significant change in any opinion or intended behavior.

Social Identity Theory (SIT)

In order to predict that a firm’s CSA behaviors would result in positive abnormal stock returns for both high and low CEI scores, Mohliver and Hawn (2020) argued that a firm’s social positioning and its financial valuation are connected through SIT of value congruence (Ashforth & Mael, 1989). Tajfel (1974) developed SIT and framed it as a component of one’s identity and concept of self by being part of a pertinent social group that is used to explain inter-group social behavior. According to SIT, an individual who belongs to a group receives value when firms signal something that is congruent with their ideology along the conservative-liberal spectrum (Sexton & Sexton, 2014). These ideological positions are aligned with political ideology (Baldassarri & Gelman, 2008) and portfolio choices are influenced by investors’ political ideology (Bolton et al., 2020).

Given these theoretical and empirical connections between CSA and investor behavior, this paper will conduct a survey experiment to test how investors react to signals that a firm has engaged in CSA. Firms that engage in CSA issues should expect investors whose political party is affiliated with that particular issue to over-allocate to those firms in order to receive value from becoming more socially congruent with the firm, as argued by SIT. In the same way, firms that engage in CSA issues should expect investors whose political party is not affiliated with that particular issue to under-allocate to those firms as a way to dissociate with firms that are not providing value congruence.

Three sociopolitical issues will be tested: LGBT, Planned Parenthood, and the Religious Freedom Institute. Regarding LGBT, while only 4.5% of American adults self-identify as a member of the LGTBQ population (according to 2017 Gallup Poll), most Americans have moderate to strongly held views on it that are highly polarized (Pew Research Center, 2020) with over 85% of Democrats in favor of teaching gay rights, sexual orientation, gender identity, and trans-rights in high school while less than 40% of Republicans were in favor (Polikoff, et al., 2022).

Planned Parenthood has become a polarizing institution because it is a material provider of the estimated 926,200 abortions in the United States (Jones & Jerman, 2017). One poll found that 63 percent of Republicans supported defunding the organization versus only 7 percent of Democrats (Washington Post, 2017). Although, in this case, Republicans show less uniformity in their views about what to do with Planned Parenthood (63% support defunding it), there is still a material difference by political affiliation, which qualifies the funding of this organization as a polarizing issue.

The Religious Freedom Institute is a nonprofit that sponsors legislation and activism related to protecting religious expression in the public domain (e.g., allowing a Muslim student to wear a headscarf in a classroom). While the issue of supporting religious freedom was largely bipartisan in the United States before the 2000s, Democratic sponsorship for religious-freedom legislation has diminished over the last 20 years (Lewis, 2021). For example, during 2014, only four votes of support were issued from Democrats in the five states that had religious freedom bills during that year (Lewis, 2021). Kazyak, et al., (2018) highlighted that Republicans were more than twice as likely as Democrats (73.75% vs 34.53%) to favor religious freedom laws.

Hypotheses

Using the SIT construct of value congruence, firms that engage in CSA issues should expect investors to allocate to them differently based on their political party’s affiliation to that CSA cause (in order to achieve more value congruence). Given this theoretical relationship, and support from prior studies, the following hypotheses will be tested:

• H1:Participant portfolio allocation to companies supporting LGTB social causes will be higher for Democrats relative to Republicans.

• H2:Participant portfolio allocation to companies supporting Planned Parenthood will be higher for Democrats relative to Republicans.

• H3:Participant portfolio allocation to companies supporting the Religious Freedom Institute will be higher for Republicans relative to Democrats.

• H4:Participants who are Democrats will allocate more to companies supporting LGTB and Planned Parenthood social causes than to the non-CSA control companies.

• H5:Participants who are Democrats will allocate less to companies supporting the Religious Freedom Institute than to the non-CSA control company.

• H6:Participants who are Republicans will allocate more to companies supporting the Religious Freedom Institute than the non-CSA control company.

• H7:Participants who are Republicans will allocate less to companies supporting LGTB and Planned Parenthood social causes than the non-CSA control companies.

Methods

Data collection

This study used survey data collected September, 2021. The dataset was funded by Inspire Investing, LLC. Participants of the survey were recruited through Amazon’s MTurk platform, which is known to provide access to a relatively low-cost, large pool of Americans that approximates the general U.S. population demographics for age and race but may not be representative when predicting religious affiliation or personality (Burnham et al., 2018). To qualify for the survey, participants from MTurk needed to have at least a 95% approval rating and at least 1,000 surveys (i.e., HITs) completed. A 95% approval rating means that 5% of the survey participant’s previously completed surveys have been rejected by the researcher. The lower the approval percentage, the higher the probability that the participant is a survey bot (Dreyfuss 2018; Miele 2018). The approval rating and HITs completed for this study were in-line with the recommended thresholds of 95-99% approval rating and 5,000 HITs (Matherly, 2019). Once survey participants agreed to participate in the survey through MTurk, they were sent to Qualtrics, which administered the survey and collected the data. The time to complete the survey was approximately 8 minutes and survey participants were compensated with $0.50 once the survey was submitted. For this study, the survey collected 1,494 participants on September 29, 2021.

Survey design

During survey collection, respondents were initially stratified by political affiliation (Democrat, Independent, Republican) to ensure at least 300 participants for each political affiliation. After that, participants were randomly assigned to one of three sociopolitical issues (LGBT, Planned Parenthood, Religious Freedom Institute) treatment groups or a non-controversial control group option. The study targeted a minimum of 1,200 total participants to ensure 100 participants in each of the categories (4 groups x 3 political affiliations). This is based on a sample size calculation required to have an effect size of 0.10, statistical power level of 0.90, and a two-tailed probability level of 0.025. This yields a required sample of 85 for each group (or 1,020 in total) which was rounded to a targeted respondent population of 100 for each cell of the analysis (or 1,200 in total). The survey ended up collecting slightly more than was needed according to these calculations (1,494 participants). Although participants were randomly assigned to one of four groups (LGBT, Planned Parenthood, Religious Freedom Institute, control), the ending distribution of the four groups was not even because the data was being stratified by political affiliation and many participants were dropped in a non-random way at the beginning of the survey in order to get a sufficient amount (greater than 300) of Independent/Other political participants.

Dependent variable

The dependent variable is the percentage allocated to a stock that has a sociopolitical community outreach program (or non-sociopolitical emphasis for the control group). The following question was the only question that was altered between the three treatment groups and control group.

“You begin work at a new company and transfer your $100,000 of retirement money to a retirement plan with your new company. Unfortunately, the new plan’s investment choices consist of only four stocks. The only information about these stocks is below. The ‘stock community outreach program’ refers to the main way the company uses its profits to give back to its community. Given this information, please choose what percentage of your $100,000 you would likely contribute to each investment choice.”

LGBT treatment group question

Investment Choice	Stock Sector	30 Yr. Avg. Annual Return	Stock Community Outreach Program
Stock 1	Technology	12%	Internet access for all program
Stock 2	Consumer Goods	12%	Food pantry program
Stock 3	Financial	12%	LGBT diversity program
Stock 4	Healthcare	12%	Mobile health clinic

Planned parenthood treatment group question

Investment Choice	Stock Sector	30 Yr. Avg. Annual Return	Stock Community Outreach Program
Stock 1	Technology	12%	Internet access for all program
Stock 2	Consumer Goods	12%	Food pantry program
Stock 3	Financial	12%	Financial literacy program
Stock 4	Healthcare	12%	Planned Parenthood

Religious freedom institute treatment group question

Investment Choice	Stock Sector	30 Yr. Avg. Annual Return	Stock Community Outreach Program
Stock 1	Technology	12%	Internet access for all program
Stock 2	Consumer Goods	12%	Religious Freedom Institute
Stock 3	Financial	12%	Financial literacy program
Stock 4	Healthcare	12%	Mobile health clinic

Control group question

Investment Choice	Stock Sector	30 Yr. Avg. Annual Return	Stock Community Outreach Program
Stock 1	Technology	12%	Internet access for all program
Stock 2	Consumer Goods	12%	Food pantry program
Stock 3	Financial	12%	Financial literacy program
Stock 4	Healthcare	12%	Mobile health clinic

Regarding the overall design of the dependent variable question, the four stock sectors corresponded to the top four stock sectors by market capitalization as of October, 2021. The control group used for each of the treatment groups were meant to (1) correspond to the stock sector so that the stock would likely have some level of expertise in addressing the community need and (2) be a non-political issue so that it would not likely connect participants to their political social group. The non-sociopolitical control group question that corresponds to sociopolitical treatment question for LGBT is when a company is engaged in a ‘financial literacy program.’ The non-sociopolitical control group question that corresponds to sociopolitical treatment question for Planned Parenthood is when a company is engaged in a ‘roaming health clinic.’ The non-sociopolitical control group question that corresponds to sociopolitical treatment question for Religious Freedom Institute is when a company is engaged in a ‘food pantry program.’

When making asset allocation decisions among different investment options, in the absence of a controversial political issue (which may cause investors to allocate towards firms in order to achieve value congruence), it should be expected that participants will allocate using a naïve allocation strategy (1/n, where n is the number of investments from which to choose). Given that each question has four choices, a naïve allocation would mean participants would likely choose to allocate 25% to each investment option. It has been shown in prior studies that a naïve allocation (i.e., 1/n heuristic) is a typical strategy that investors use to help them make portfolio allocation decisions (Benartzi & Thaler, 2001). Every stock was given the same historical return to encourage investors to choose a 25% naïve allocation to take full advantage of principles of stock diversification since each stock was in a distinctly different market sector.

Independent variables

Socioeconomic information was collected from FINRA Educations’ National Financial Capability Study (NFCS) as tested and validated measures that help predict financial decision making. These variables include gender, age, race/ethnicity, marital status, level of education, income level, and employment status.

Given that the dependent variable is asking respondents to allocate to a retirement portfolio, tested and validated measures of financial perceptions and attitudes were used from the NFCS. These variables include a Likert-like scale of satisfaction with one’s personal financial condition (“Overall, thinking of your assets, debts, and savings, how satisfied are you with your current personal financial condition?), a Likert-like scale of willingness to take risks with their investments (“When thinking about your investments, how willing are you to take risks?”), and, finally, if respondents have figured out how much they need to save for retirement (yes/no/not sure; ‘not sure’ was coded as ‘no’).

Data analysis

Hypotheses 1-3 (participant’s portfolio allocation to companies supporting CSA will be different by political affiliation) will be tested, first, using an ANOVA test to ensure that there is a significant difference in portfolio allocation by party affiliation. Second, an OLS regression analysis will be utilized that includes socioeconomic and financial control variables to identify individual differences by political party. When testing hypotheses 4-7 (participants will over, or under, allocate to CSA issues based on political affiliation) a paired t-test calculation will be used between the CSA treatment question and the respective non-CSA control question. Democrats are expected to have a significant over-allocation (relative to the control question) to issues that provide them value congruence (LGBT and Planned Parenthood) and under-allocate (relative to the control question) to issues that do not provide them value congruence (Religious Freedom Institute). Likewise, Republicans are expected to have a significant over-allocation (relative to the control question) to issues that provide them value congruence (Religious Freedom Institute) and under-allocate (relative to the control question) to issues that do not provide them value congruence (LGBT and Planned Parenthood).

Results

Descriptive statistics

Table 1 shows the descriptive characteristics of the dependent variables. As expected, in the absence of a CSA issue, participants chose a naïve allocation (1/n heuristic) given that the median allocation for every non-CSA investment option was 25%. ANOVA results provide partial evidence in support of hypotheses 1-3 given that there are significant differences by political affiliation for the CSA options. Mean participant portfolio allocations were significantly different by political party for firms supporting LGBT social issues (H2; p=0.0051), Planned Parenthood (H3; p=0.0004), and the Religious Freedom Institute (H4; p=0.0019). In addition, when participants were given a similar portfolio allocation question without any sociopolitical options (control question, n=382), results show no significant mean differences by political party.

Table 1 Descriptive characteristics of dependent variables

Table 2 shows the descriptive characteristics of the independent variables. Democrats and Republicans made up 40% and 35% of the sample, respectively. Independent/Other participants made up 25% of the sample. This compares with a Gallup poll on September 17, 2021, that found that 29% of Americans identified as Democrats, 29% identified as Republican, and 41% as Independent (Gallup, 2021). The sample of 1,494 participants, relative to the 2019 U.S. Census, had a higher representation of males (60% vs. 49%) and married (61% vs. 48%), and lower representation of white (67% vs. 76%). 60% of the group had calculated how much they needed for retirement and the average financial satisfaction and risk tolerance of the participants was 6 out of 10.

Table 2 Descriptive characteristics of independent variables (N=1,494)

OLS Results

OLS regression results (Table 3) provide further evidence for hypotheses 1-3, which more specifically predict the direction of portfolio allocation by specific political affiliation. After factoring in financial and socioeconomic control variables, political affiliation remains a significant predictive variable for sociopolitical portfolio allocation decisions. Results of the OLS show that, if a participant identified as Democrat, the portfolio allocation percentage to firms supporting LGBT issues, Planned Parenthood, and the Religious Freedom Institute, on average, were +4.3% higher (supports hypotheses 1), +7.5% higher (supports hypotheses 2), and -3.8% lower (supports hypothesis 3), respectively, relative to a participant who identified as Republican, holding all-else-equal. Identifying as an Independent/Other did not appear to significantly predict portfolio allocations to a sociopolitical issue, except for a lower allocation to a firm that supports the Religious Freedom Institute (-3.0% relative to a participant who identified as Republican, holding all-else-equal).

Table 3 OLS regression results

Paired T-Test Results

Table 4 shows paired t-test results to ascertain whether participants would over (or under) allocate to a CSA firm relative to a non-CSA control firm. The non-CSA control firm was supported a non-sociopolitical issue that was connected to the sector of the company (i.e., healthcare company would support a mobile health clinic instead of Planned Parenthood). Evidence was found in support of hypothesis 7 given that Republicans under-allocated to firms that supported LGBT and Planned Parenthood (-7.77% and -7.48% mean allocation difference, respectively, relative to the control allocation). Evidence was also found in support of hypothesis 5 since Democrats under-allocated to the firm supporting the Religious Freedom Institute (-8.88% mean allocation difference relative to the control allocation). Evidence was not found, however, in support of either hypotheses 4 or 6 since both Democrats and Republicans did not over-allocate (relative to the control group) to the sociopolitical issues that were aligned to their political affiliation. In all sociopolitical issues, both Democrats and Republicans under-allocated to the CSA firms relative to the non-CSA control firms.

Table 4 Paired t-test of mean % difference between CSA portfolio allocation vs. non-CSA control portfolio allocation by Republican or Democrat political affiliation

Discussion

According to SIT, individuals will naturally want to identify with social groups who are congruent with their existing concept of self (Tajfel, 1974). When a firm signals that they identify with one end of the conservative-liberal spectrum through corporate sociopolitical activism (CSA), this paper predicted that investors will invest differently, depending on their political affiliation, as well over-or-under allocate their investment dollars to that firm in order to gain the benefit of value congruence to a political social group. The results of a survey experiment found evidence that political affiliation did materially predict the allocation decision of investors to firms supporting CSA (hypotheses 1-3). These results are consistent with other prior literature that has shown that political affiliation does predict investment behavior (Bhagwat et al., 2020; Gomes, 2021; Mohliver & Hawn, 2020; Wintoki & Xi, 2020). These results uniquely contribute to existing literature through its use of an experimental design methodology to predict investor behavior.

More specifically, results provide evidence that participants who identified as Democrat were more likely than Republicans to allocate investment dollars towards firms signaling a liberal political identity through their support of LGBT and Planned Parenthood. By comparison, participants who identified as Republican were more likely than Democrats to allocate investment dollars towards firms signaling a conservative political identity through their support of the Religious Freedom Institute. Evidence was found that Democrats and Republicans will under-allocate (relative to a control group) to firms that signal support for sociopolitical causes that are not aligned to their political identity (hypotheses 5 and 7).

In contradiction to hypothesis 4 and 6, both Democrats and Republicans under-allocated (relative to the control group) to firms that engaged in sociopolitical issues aligned to their political identity. For example, Democrats allocated 1.75% less (versus control option) to a firm supporting LGBT social issues, even though this sociopolitical issue is largely congruent with a Democrat viewpoint. Also, Republicans allocated 1.89% less (versus control option) to the firm supporting the Religious Freedom Institute even though this sociopolitical issue is largely congruent with a Republican viewpoint. This under-allocation was also relative to a naïve allocation strategy (1/n). A one sample t-test of each CSA treatment question confirms that the mean allocation to each CSA firm was significantly less than a naïve allocation of 25% for both Democrats and Republicans for all three sociopolitical issues. This under-allocation to all CSA issues, regardless of political affiliation, may be due to investors receiving a negative signal from a CSA firm that they are redirecting resources from profit-oriented objectives to riskier activities, which is consistent with the findings of Bhagwat et al. (2020). This financial threat perceived by investors could potentially overpower the perceived benefit of value congruence from over-allocation.

Limitations to this study include the use of Amazon MTurk survey data, which may not fully represent the average American adult. As mentioned in the descriptive analysis, the survey used in this study had a lower representation of those who identified as white (67% vs. 76%) and a higher representation of those who identified as male (60% vs. 49%) and married (61% vs. 48%) versus the U.S. Census (U.S. Census, 2019). In addition, the thought experiment survey design is only measuring intention versus actual behavior. As a result, there is the potential for Hypothetical Bias (Bohm, 1972) where stated preference does not match actual behavior (Bosworth & Taylor, 2012). If participants were managing their actual money, results may be different (Cumming, et al., 2022) Future research that measures what investors actually do with their investment allocations relative to CSA issues is warranted. Other future research should seek qualitatively to understand why a company’s leadership engages in CSA.

Implications

This paper provided experimental evidence that an investors’ political affiliation does help predict asset allocation decisions. This paper provides evidence that CSA could potentially cause an investor to under-allocate to a firm on both ends of the political spectrum. SIT of value congruence was used to help explain why, for example, a firm that supported Planned Parenthood could experience an under-allocation by Republicans due to a lack of value congruence. However, for this example, value congruence does not appear to be enough to prevent Democrats to under-allocate, as well. This under-allocation by Democrats to an issue that is widely supported by their party (Norman, 2022) may signal that they view CSA as a sub-optimal use of the firm’s resources (Bhagwat et al., 2020).

Given these results, it would be appropriate for firms to fully reckon with the potential costs of engaging in CSA, as it relates to investor portfolio allocation. This study finds experimental evidence that an investor may potentially under-allocate to companies engaged in CSA by up to 9% of their portfolio allocation. This under-allocation from investors should significantly increase the cost of capital (Ghoul et al., 2011). The threat of investors responding to CSA is rising given the rise of corporate social reporting (CSR) where investors are becoming more accustomed to using non-GAAP, non-financial information in their investment process. Certain companies are even compiling CSA data in order to help portfolio managers and individual investors use this data as a part of a exclusionary screening process. For example, Inspire Investing, LLC recently developed a screening tool, Insight Pro, that investment managers are using to gather CSA data as a part of a Christian exclusionary screening process. On the other side of the liberal-conservative spectrum, allies of the LGBT community have used the CSA information from the Corporate Equality Index (CEI) in order to exclude companies that are ‘worst performers.’ As more CSA data becomes fully available to investors, it is likely that more investors will under-allocate to firms who engage in CSA, which increases their cost of capital. If the motivation for CSA is for a firm to benefit their community, only engaging in CSR may be a more appropriate way for public corporations to spend resources since prior literature has shown CSR to benefit both firms (Baron, 2001; Nalick et al., 2016) and investors (Hoepner & Schopohl, 2020). Additionally, CSR has been shown to lower a firm’s cost of capital as well (Ghoul et al., 2011).

Human subjects

Given the use of human subjects, this study was reviewed and approved by the Biola University Protection of Human Rights in Research Committee (F21-013_SE). Given the use of a survey, informed consent was required of participants before beginning the survey.