Inequality Federalism and Economic Self-Interest in Subnational Progressive Tax Politics

Abstract

In a national context of unabated inequality growth and recurrent tax cuts benefiting the wealthy, we have witnessed a notable rise in “inequality federalism,” characterized by subnational initiatives to redress economic inequality through progressive taxation and social spending. While leading research documents the tendency for Americans to fall prey to “unenlightened self-interest” when evaluating complex national tax policy, recent research suggests that the simplicity and clarity of emerging subnational redistributive initiatives facilitates the enactment of economic self-interest, particularly among lower income citizens. This short article builds on prior work by offering the most extensive analysis to date of the role of economic self-interest in public support for subnational progressive tax policies. Drawing upon opinion polls covering eight separate state ballot measures or legislative enactments, we find that in each instance the greatest level of support for the respective progressive tax policy is observed among lower income citizens.

Keywords

inequality federalism public opinion redistribution taxation

In the realm of redistributive politics, what evidence is there of the exercise of economic self-interest by low-income citizens who stand to benefit the most from redistribution? Focusing on the United States, where economic inequality has steadily grown since the early 1980s (Stone et al. 2017) and inequality is higher and redistribution is lower relative to peer nations (Brandolini and Smeeding 2006), do we observe a relationship between an individual’s position on the income ladder and their support for redistribution? When evaluating the linkage between individual income and support for redistribution, we find an interesting disjuncture between studies providing evidence of self-interest in the realm of “abstract” policy support and work evaluating specific, and often complex, national policy proposals. To illustrate, existing research finds that lower income Americans express greater support than upper income Americans for federal welfare spending, guaranteed employment, taxation on inherited wealth, and the abstract principle of reducing economic inequality (Fong 2001; Kluegel and Smith 1986; Sears et al. 1980; Soroka and Wlezien 2008). This work suggests that, when it comes to views about general actions taken by the federal government in the economic realm, the preferences of lower income Americans appear to be informed by economic self-interest.

However, influential work by Bartels (2008) demonstrates that, when it comes to evaluating complex national tax policies, such as the “Bush tax cuts” in 2001 and 2003, three pieces of evidence for what is termed “unenlightened self-interest” emerge: (1) Americans who believed their own tax burden was too high were significantly more likely to support the cuts despite the fact that the cuts primarily benefited the wealthy, (2) lower income Americans expressed a greater level of support for the cuts than upper income Americans, and (3) support for a specific feature of the tax cuts—the repeal of the estate tax—was remarkably high among low-income Americans who likely did not experience personal material gain from its repeal. Bartels attributes these findings to deficits in political information and sophistication (cf. Lupia et al. 2007). Building on the work of Bartels, Piston (2018), using more recent survey data, found that low- and high-income Americans’ support for the federal estate tax did not significantly differ, which he similarly attributes to the fact that many Americas are not aware that only the wealthiest pay the cost of the tax. Moreover, Piston extends the work of Bartels by analyzing public support for a separate feature of the federal tax code, the home mortgage interest deduction (HMID). Despite the HMID largely benefiting higher income Americans, Piston failed to find significant differences in support for the deduction across by income lines. In sum, similar to Bartels, Piston (2018) concludes that lack of information about the distributional effects of these federal tax provisions underscores the seeming absence of economic self-interest as a factor shaping the tax policy preferences of lower- and middle-class Americans.

These findings are further complemented by experimental research by Ballard-Rosa et al. (2017), who find that lower income Americans do not express greater support than higher income Americans for national tax proposals levying heavier tax rates on the wealthy. In fact, when analyzing support for national tax plans specifying tax rates for those with incomes above $375,000 per year, Ballard-Rosa and colleagues find that the poorest Americans were less supportive of higher tax rates than the wealthiest Americans. Moreover, analyses of public opinion data over time suggest that increasing income inequality decreased low-income Americans’ support for redistribution (Kelly and Enns 2010), which runs contrary to the expectation, rooted in theories of self-interest, that support for redistribution among lower income citizens will increase as inequality grows (Meltzer and Richard 1981). Importantly, all of this work is undergirded by earlier experimental studies conducted by Roberts et al. (1994) demonstrating (1) higher support among subjects for progressive taxation as an abstract principle versus in concrete practice and (b) the disconnect between abstract and concrete support was most pronounced among low-knowledge subjects. Taken together, this work suggests that the average American—and particularly those at the bottom of the economic ladder—may lack the information or sophistication necessary to accurately determine the pocketbook effects of complex federal tax proposals and thus be limited in their ability to enact economic self-interest.

Indeed, moving backwards in time, nearly a decade and a half before the Bush tax cuts, the American public was presented with a complex set of changes to the federal tax code by President Reagan in 1980s (i.e., “Reagan tax cuts”), which lowered top tax and capital gains tax rates. In an analysis of public opinion on the Reagan tax cuts, Lacy (1998) failed to find significant differences in support for the cuts between low- and high-income Americans, concluding that “popular explanations such as voter self-interest . . . perform poorly as predictors of voter opinion” (Lacy 1998, 299). One possible explanation for the lack of a self-interest effect with respect to the Reagan tax cuts is widespread voter ignorance about the cuts, with ABC News/Washington Post polls conducted in 1986 revealing that 63 percent of Americans did not know enough about the cuts to report an opinion (Kertscher 2017). Moving forward in time, nearly a decade and a half after the Bush tax cuts, the American public was presented with another complex tax plan involving a series of cuts, this time by President Trump in 2017, which would largely benefit wealthy corporations and individuals (Tax Policy Center 2017). Analysis of public opinion on these cuts indicates that lower income Americans did not express greater disapproval of the cuts than wealthy Americans (Doherty et al. 2018). For example, when asked about the expected effects of the Trump tax cuts, Doherty and colleagues found that 42 percent of those with incomes under $40,000 per year as well as 42 percent of those with incomes above $100,000 per year believed the cuts would have a negative effect over the coming years. Moreover, when asked about their approval of Trump’s overall tax plan, Doherty et al. again find that the same percentage (47%) of those earning below $40,000 or above $100,000 disapprove of the plan. In the end, Doherty and colleagues situate the lack of strong income-based differences in opinions within the fact that only 30 percent of those surveyed professed to have a strong understanding of the tax plan.

Inequality Federalism and Economic Self-Interest

One political development in the United States that may serve as antidote to the publics’ seemingly nonexistent, or merely “unenlightened,” exercise of self-interest in the realm of redistributive tax policy is the rise of inequality federalism. We define inequality federalism as the policies pursued at the state and local level to impact rising inequality and their strategic interaction with the prevailing federal policy environment. Leading scholarship documents how continued federal inaction to address growing inequality in conjunction with the devolution of welfare administration in the 1990s put state governments in a more important position to influence levels of inequality (Kelly and Witko 2012). This work highlights the various means available to state governments to influence levels of inequality, such as education and economic development policies, as well as minimum wage laws. Central to this development is the ballot initiative, which is argued to serve as a means for citizens to redress the perceived failure of the federal government to address policy problems by directing their efforts toward subnational arenas where they can have a direct role in enacting policy (Matsusaka 2004). In the context of inequality federalism, citizens and local politicians frustrated by the failure of the federal government to counter inequality growth can promulgate subnational redistributive policy initiatives (Franko and Witko 2018; Freeman and Rogers 2007).

Following the 2008 Financial Crisis, we have seen the emergence of state redistributive initiatives proposing tax increases on wealthy individuals to raise funds for social programs. One such policy was Proposition 1098 in Washington state in 2010, which was analyzed by Franko, Tolbert, and Witko (2013). The 2010 ballot in Washington state presented voters with a concise statement of the proposed policy:

This measure would tax “adjusted gross income” above $200,000 (individuals) and $400,000 (joint-filers), reduce state property tax levies, reduce certain business and occupation taxes, and direct any increased revenues to education and health.

This straightforward proposal stands in stark contrast to complex national tax proposals, such as the Economic Growth and Tax Relief Reconciliation Act of 2001 (i.e., “Bush tax cuts), which is a 117-page document with numerous sections and provisions affecting income tax rates, top tax rates, capital gains taxes, contribution limits for individual retirement accounts, and the estate tax. Drawing on the work of Chong, Citrin, and Conley (2001), who argue that self-interest is likely to obtain when the personal stakes of a policy are clear and substantial, Franko, Tolbert, and Witko (2013) argue that the simple language of the policy in conjunction with the straightforward costs and benefits for high- and low-income residents greatly facilitated the enactment of self-interest among the poor—and did so without requiring a substantial amount of political information.¹ Their main conclusion is that simple subnational redistributive tax initiatives (as compared to complex national tax plans) may facilitate the exercise of self-interest among lower income citizens. For example, Franko, Tolbert, and Witko (2013, 934) conclude their article by stating, “Our research shows that if tax policy is simply structured and redistribution is a central theme of the policy debate, at least some citizens can make better informed decisions that are in line with their attitudes and interests.” As noted by Franko, Tolbert, and Witko (2013), their findings are noteworthy in that they “challenge a general assumption in the literature that economic self-interest and attitudes about inequality do not influence preferences on redistributive taxation” (p. 934).

While the findings and conclusions of Franko, Tolbert, and Witko (2013) are engrossing, they are currently based on a single policy proposal at a single point in time. Many other states have held votes on initiatives comparable to Prop 1098, such as Oregon, Illinois, California, and Maine. For example, Measure 66 in Oregon in 2010 was nearly identical to that of Prop 1098 in Washington, as it proposed to state voters a plan that

Raises tax on household income at and above $250,000 (and $125,000 for individual filers). Reduces income taxes on unemployment benefits in 2009. Provides funds currently budgeted for education, health care, public safety, other services.

In addition, Proposition 55 in California in 2016, while slightly more complex, provided state voters with a relatively simple law:

Extends by twelve years the temporary personal income tax increases enacted in 2012 on earnings over $250,000, with revenues allocated to K-12 schools, California Community Colleges, and, in certain years, healthcare. Fiscal Impact: Increased state revenues—$4 billion to $9 billion annually from 2019-2030—depending on economy and stock market. Increased funding for schools, community colleges, health care for low-income people, budget reserves, and debt payments.

In contrast to the cumbersome nature of many federal tax plans, the subnational tax initiatives that have occurred within the past decade typically present state residents with a relatively concise redistributive proposal to tax high-income earners to raise funds for social spending. In addition to ballot initiatives, several states have pursued redistributive “tax-the-rich” policies via legislative enactment, such as in Maryland, New York, and New Jersey. While these types of proposals are increasing in prevalence, we do not know whether the findings by Franko, Tolbert, and Witko (2013) extend to other cases. Indeed, in their conclusion, Franko, Tolbert, and Witko (2013) state, “the results suggest that future attempts to raise taxes on the wealthy may succeed if they are openly debated in redistributive terms” (p. 934). We interpret this statement as a call for future research.

In the analysis that follows, we extend the work of Franko, Tolbert, and Witko (2013) by testing for economic self-interest in citizens’ support for progressive tax proposals in eight other states. The results from our analyses replicate the findings of Franko, Tolbert, and Witko (2013) and thus expand the base of empirical support for the claim that lower income Americans can enact economic self-interest in the face of straightforward redistributive tax policies. In addition to extending the results presented by Franko and colleagues to other redistributive state ballot measures, our analysis demonstrates that self-interest effects manifest for tax-the-rich policies directly enacted by state legislatures. This finding is noteworthy, as it suggests that the findings of Franko and colleagues are not induced by characteristics of the initiative process and may generalize to straightforward redistributive proposals regardless of the mode of adoption. We conclude by discussing the limitations of our findings, as well as directions for future research.

State Ballot Measures and Legislative Enactments

In addition to Prop 1098 in WA, four other U.S. states have held elections presenting voters with the opportunity to support what Franko, Tolbert, and Witko (2013) term a “robin hood” redistributive policy: Measure 66 in Oregon (Special Election, January 2010), Millionaires Tax Increase for Education Question in Illinois (General Election, November 2014), Proposition 55 in California (General Election, November 2016), and Question 2 in Maine (General Election, November 2016). In each case, the initiative proposed a straightforward tax on high-income individuals and households to raise revenues for social programs (e.g., education, health care, etc.). The language of each initiative is provided in Online Appendix A. In addition, we analyze the Massachusetts millionaire’s tax slated to appear on the ballot in the 2018 General Election that was ultimately struck from the ballot by the state’s high court and the millionaire’s tax proposed by New Jersey Governor Phil Murphy as part of the state budget in 2018. Finally, we analyze the millionaire’s taxes legislatively enacted in Maryland in 2008 and New York in 2009. Analysis of public opinion on these eight state-level redistributive tax policies provides an opportunity to assess the robustness of the results presented by Franko, Tolbert, and Witko (2013). While other states’ legislatures have adopted tax increases on the wealthy (Gramlich 2009; Henchman 2012), such as North Carolina and Wisconsin, we were unable to locate survey data for these other cases.² Thus, while our analysis involves the universe of state progressive tax increases involving ballot initiatives, our analysis of legislative enactment of progressive tax increases involves a subset of cases from states with polling organizations soliciting state residents’ opinions regarding such increases. In the end, the eight cases we analyze come from Democratic-leaning states, which is an issue we return to in latter sections of this article.

Data and Measures

To analyze the effect of economic self-interest on support for these policies, we conduct secondary analysis of state survey data for each case. For Measure 66 in Oregon, we rely on the Oregon January 2010 Special Tracking Survey #1 (N = 400, Nov 30–Dec 2, 2009) and #2 (N = 400, Jan 4–6, 2010) conducted by Lindholm Research, LLC. For the Millionaires Tax in Illinois, we rely on the Fall 2014 Simon Poll (N = 1,006, Sep 23–Oct 15, 2014) conducted by the Paul Simon Public Policy Institute at Southern Illinois University. For Proposition 55 in California, we rely on the September 2016 (N = 1,702, Sep 9–18) and October 2016 (N = 1,704, Oct 14–23) Statewide Surveys conducted by the Public Policy Institute of California (PPIC). For Question 2 in Maine, we rely on the September 2016 (N = 593, Sep 15–20) and October 2016 (N = 761, Oct 20–25) Portland Herald Press Polls conducted by the University of New Hampshire Survey Center. For the legislatively approved but judicially blocked millionaire’s tax in Massachusetts, we rely on the January 2017 (N = 508, Jan 15–17) and June 2017 (N = 504, Jun 19–22) WBUR Issues Surveys conducted by MassINC Polling Group. Finally, for the millionaire’s tax enacted by the Maryland, New York, and New Jersey legislatures, we rely on the October 2007 Maryland Poll (N = 1,103, Oct 18–22) conducted by TNS Intersearch, the October 2011 New York Poll (N = 800, Oct 10–12) conducted by the Siena College Research Institute, and the 2018 New Jersey Issues Poll (N = 728, Mar 22–29) conducted by the Stockton Polling Institute. For more information about these surveys, see Online Appendix A.

Each survey included questions tapping respondents’ support for the respective tax policy. The eight policies we analyze vary slightly in terms of rate increases, income brackets targeted for rate increases, and intended use of raised revenues (see Online Appendix A); however, the essence of each policy is the same—a tax increase on high-income state residents to raise funds for public goods and social spending. Moreover, despite minor variation in the complexity of these policies, within the context of each public opinion survey, the question asked to respondents about the respective tax increase on high-income earners was straightforward. For example, using Measure 66 in Oregon, while the ballot presented voters with greater detail regarding the finer points of the measure, the survey we utilize presented the measure more concisely by asking,

If the election were held today, would you vote yes or no on Measure 66: Raises tax on household income at and above $250,000 (and $125,000 for individual filers). Reduces income taxes on unemployment benefits in 2009. Provides funds currently budgeted for education, health care, public safety, other services.

Similar straightforward questions were asked to respondents to gauge support for the other measures we analyze (see Online Appendix A for question wording).

Each survey contained a self-reported measure of respondent income, which we use as the independent variable of interest in our analysis. We focus on respondent income as our indicator of economic self-interest because the policies under analysis define tax liability by income, such that income level clearly demarcates those who will and will not be subjected to the tax increase. As such, income is the measure used by Franko, Tolbert, and Witko (2013) and is the measure we use in our analysis. Each income measure presented respondents with ordered income categories and asked respondents to select the category containing their annual income. While the income categories range somewhat across surveys, nearly all presented respondents with options ranging from below $30,000 to above $150,000 per year. For example, the New York Poll provided only three ordered categories (<$50K, $50–100K, and >$100K) while the PPIC surveys provided seven ordered categories (<$20K, $20–40K, $40–60K, $60–80K, $80–100K, $100–200K, and >$200K). Online Appendix A provides the exact measure of income used in each survey. Our main analysis treats respondent income as a continuous variable; however, we demonstrate the robustness of our results to using dichotomous indicators of income quartile (Online Appendix Table C2).

The control variables available in each survey vary; however, each survey contained enough questions to enable each analysis to control for education, age, gender, and partisanship. The surveys conducted in OR and ME do not include questions about race, which according to the firms conducting each survey is due to the overwhelmingly Anglo composition of the state rendering negligible variation on race in probability samples. The surveys in the more diverse states of CA, IL, MA, MD, NJ, and NY each contain questions about respondents’ racial identification, enabling us to include separate controls for respondents identifying as Black, Latino, or Asian/Pacific Islander. In cases where other key variables are present, such as employment status, homeownership, church attendance, and gubernatorial approval, we control for these factors. Finally, in cases where two surveys were conducted and merged together (i.e., OR, CA, ME, and MA), the analysis includes a fixed effect for survey-month. We use multivariate regression analysis to test for a systematic relationship between a respondent’s income and their support for a respective ballot measure. While work in social science over the past decade has sought to employ more rigorous research designs for estimating causal effects, we are satisfied with a model-based approach in the context of our research question, as uncovering significant correlations between income and support for state redistributive policies represents a valuable contribution over our current knowledge. Furthermore, our analyses control for several feasible confounding factors (e.g., education, age, partisanship). While we were unable to control for additional factors such as core values (e.g., egalitarianism or individualism) or affect toward social groups (e.g., racial minorities, the poor, etc.), we do control for partisanship in each model and existing research indicates that party identification shapes and constrains core values (Goren 2005) and is substantially driven by feelings toward social groups (Kane et al. 2019; Miller et al. 1991; Robison and Moskowitz 2019).³

Results

To facilitate comparison of the marginal effects of income across surveys, we present the results from linear regression models. Our results do not meaningfully change when estimating models more technically suited to each dependent variable (Online Appendix Table C1), when accounting for possible nonlinear effects of income by estimating a model containing indicators of income quartile (Online Appendix Table C2), or when using multiple imputation to account for missing data on income (Online Appendix Table C3). For ease of interpretation, we recoded all non-dichotomous independent and dependent variables to range from 0 to 1. The results from our analysis are presented in Figure 1 (Online Appendix B). Figure 1 displays the marginal effect estimates of respondent income in each survey, along with associated sample sizes for each analysis. In each instance, income exerted a negative effect on policy support that was at minimum significant at the p < .10 level. In two of the eight cases (OR and NY) the effect of income was marginally significant (p < .10), in three cases (IL, MA, and NJ) the effect of income was significant at the p < .05 level, and in three cases (CA, MD, and ME) the effect of income was significant at the p < .001 level. Given our use of linear regression, the effect sizes are directly interpretable, ranging from roughly –.08 in the cases of OR to –.37 in the case of Maryland. The average effect is –.14, which is comparable to the effect observed for income by Franko, Tolbert, and Witko (2013) in their analysis of Prop 1098, the effect observed by Fong (2001) in her analysis of support for redistribution as a general principle, and the effect observed for lottery-induced affluence on tax policy preferences by Doherty et al. (2006).

Figure 1.

The effect of income on support for redistributive state tax policy.

The effect of income across these eight cases can also be evaluated in comparison to the effect of partisanship, which in each case exerted significant and substantively large effects (see Online Appendix B). For example, the smallest effect of income observed in Figure 1 is in Oregon, where the estimated effect of partisanship is β = −.35 (p < .001), which places the effect of income (β = −.07, p < .10) at roughly 80 percent smaller than the effect exerted by partisanship. The largest effect observed for income was in Maryland, where the estimated effect of partisanship is β = −.41 (p < .001), which places the effect of income (β = −.37, p < .001) at only 10 percent smaller than the effect exerted by partisanship. It is important to note that these relative effects are not insubstantial for a variable whose estimated influence is potentially reduced given that income is a known antecedent to partisanship (see Lewis-Beck et al. 2008). It is also notable that the estimated effect of income across these eight cases are remarkably similar, with Maryland standing out as having the largest effect of income. One explanation for the distinctly large effect of income in Maryland is wording of the question, where the tax increase was framed as a means to reduce the state budget deficit (see Online Appendix A). This may have made the measure more appealing to lower income Republicans than it otherwise would have been given that Republicans tend to disfavor tax-and-spend policies. The results in Figure 2 panel A suggest this may be the case, and we discuss this more below.

Figure 2.

Conditional effect of income on support for redistributive state policy by individual and contextual partisanship: (A) Respondent partisanship. (B) County Republican presidential vote share.

Moderation by Interest, Attention, and Partisanship

Bartels (2008) argues that political information was key in promoting the exercise of economic self-interest when evaluating the Bush tax cuts (cf. Lupia et al. 2007). Franko, Tolbert, and Witko (2013), however, suggest that in the context of straightforward redistributive proposals (i.e., Prop 1098), lower income citizens do not require a high level of policy information to enact self-interest. While none of the surveys measured respondent knowledge about the given redistributive policy under evaluation, in three cases—IL, CA, and ME—the surveys asked about general interest in, or attention to, public affairs. In the case of NY, the survey asks about attention to the Occupy Wall Street Movement; given that the NY millionaires tax was adopted in the wake of the 2008 Financial Crisis and extended in 2011 during the Occupy movement, and the movement greatly raised the salience of inequality (McCall 2013), we use attention to this movement as a proxy for attention to inequality. Across these four tests, we find no evidence that interest or attention moderates the effect of income (Online Appendix Table B2).

Another possibility is that partisanship moderates the effect of income. While Franko, Tolbert, and Witko (2013) do not uncover evidence of this, we find evidence of this in eight out of ten tests. Figure 2 panel A presents the marginal effect of income in each case among those identifying as Democrats or Republicans (full results in Online Appendix Table B3). Focusing on state-level policy, the trend is clearly that income matters more among Republicans, with low-income Republicans in each case being significantly more likely than high-income Republicans to support the redistributive policy. The most pronounced effect among Republicans was in MD, with one explanation being that the question asked to respondents framed the tax increase as a means of reducing the state budget deficit. This framing of the question may have made the measure more appealing to (low-income) Republicans by avoiding the depiction of the increase as typical liberal “tax-and-spend” fiscal politics. Among Democrats, we only observe an effect of income in ME. The income effects we observe among Republicans may reflect ceiling effects, as average Democratic support for these policies is high relative to Republicans, leaving less variation to explain; however, the effects may also be due to significant latent class conflict among Republicans across these states that is brought to the fore by these “tax-the-rich” policies. Indeed, a recent report by the Pew Research Center finds a similar cleavage among Republicans by income, with low-income Republicans significantly more likely than high-income Republicans to favor raising taxes on high-income households and wealthy corporations (Fingerhut 2017).

The Role of Intrastate Political Context

One issue that is important to address in our analysis is that the tax policies we analyze are from left-leaning states. This begs the question as to whether the self-interest effects we observe would materialize in right-leaning states? Republican-leaning states such as North Carolina and Wisconsin have implemented tax increases on high-income earners via legislative adoptions (Henchman 2012); however, we were unable to locate survey data for these states. While it is the case that most subnational “robin hood” tax policies have arisen in left-leaning states, one way to glean whether or not the self-interest effect we observe may hold in more conservative political environments is to reanalyze our data by county political context. States like CA, IL, ME, and NY, while Democratic-leaning in the aggregate, contain vast areas within the state voting heavily Republican. Take the liberal state CA for example: while Hilary Clinton received an estimated 61.7 percent of the vote statewide, in northern counties such as Modoc, Lassen, and Shasta, Donald Trump won anywhere from 64 to 71 percent of the vote. While residing in a predominately Republican county is not the same as residing in a predominately Republican state, we believe there is value in analyzing whether or not the self-interest effects we observe hold when analyzing state residents living in predominately Republican-voting counties within the liberal states in our analysis.

Figure 2 panel B (Online Appendix Table C5) presents the results from this analysis. Six of our cases (IL, CA, ME, MA, NY, and NJ) involved survey data including geocodes for respondents’ zip code or county of residence; however, the NY data contained insufficient sample size in Republican counties to enable a meaningful analysis, and there are no majority Republican-voting counties in MA in recent Presidential Elections. Thus, our analysis relies upon the four cases where we have geocoded survey data and enough counties and respondents within such counties to enable an analysis. We merged into these five datasets county-level election results for the 2012 (IL) or 2016 (CA, ME, NJ) Presidential Elections.⁴ Using this merged data, we re-estimate the models underlying Figure 1 on subsamples of the data defined by Republican Presidential vote share, estimating a model for respondents residing in counties where Republicans received less than or greater than 50 percent of votes cast.

The results in Figure 2 panel B reveal that the effect of income is more precisely estimated and consistently significant in left-leaning county contexts. That said, the effect of income in right-leaning counties is substantively larger in three out of four cases than the estimated effect in left-leaning counties. Importantly, the effect of income in Republican-voting counties is less precisely estimated, which could be due to greater variation in the preferences of residents in these areas and/or to the fact that fewer respondents in the state surveys we use reside in Republican-heavy counties. Indeed, a higher share of respondents in each survey reside in more heavily sampled urban areas which tend to vote Democrat. Thus, we cannot rule out the possibility that a larger number of respondents in Republican counties would have yielded more precise and thus significant effects for income. Finally, and perhaps more importantly, difference tests reveal that the estimated effects of income do not significantly differ between counties where the Republican vote share is less or greater than 50 percent (Online Appendix Table C6).

In sum, while we are unable to directly assess whether or not our results would hold in conservative states, the results from this ancillary analysis do not provide strong evidence indicating that the self-interest effects we observe are confined to liberal locales. Instead, the results from this ancillary analysis suggest that the self-interest effects we observe in our main analysis do not significantly differ between those residing in liberal versus conservative areas within a state. These results, however, should be interpreted as suggestive, with more research required to draw more definitive conclusions.

Conclusion

One of the central questions driving research on the politics of redistribution and inequality is whether or not average citizens—particularly the economically disadvantaged—are competent with respect to choosing economic policies that stand to improve their material self-interests and curtail growing economic inequality. The results presented in this short article corroborate previous findings (e.g., Franko, Tolbert, and Witko 2013) and suggest a robust economic self-interest effect enacted by low-income Americans: when investigating public support for five state ballot measures and three state legislative enactments that each tax the high-income earners to generate revenues for social programs, we find that lower income citizens are consistently more likely to support these policies than those residing at the upper end of the income distribution.

These findings provide an important replication and extension of past work and fortify our confidence in the capacity of simple and straightforward redistributive policy proposals to facilitate economically self-interested behavior. Indeed, when citizens are given a straightforward policy that clearly identifies high-income earners as the target of the tax increase and the allocation of the collected money to programs that benefit lower income earners, citizens across the economic spectrum appear consistently able to act on their economic self-interest. One intriguing caveat of our results is the finding that the effect of income is most pronounced among state citizens identifying with the Republican Party. These results suggest a stronger class-based cleavage within the Republican Party regarding redistributive policy than within the Democratic Party, where such income-based division is less notable and support for the analyzed state policies is comparably high among both low- and high-income individuals. While these results are intriguing, we leave a fuller investigation of them to future research. In addition, future research could explore whether the effect of income on support for subnational redistributive policies is shaped by local costs of living (see Ogorzalek, Piston, and Puig 2019), as economic self-interest effects may be more pronounced when accounting for respondent income relative to median income levels in one’s city of residence. Finally, one limitation of our analysis and findings is that we only analyze tax policies in left-leaning states. While we have endeavored to assess whether or not the self-interest effects we observe hold in Republican versus Democratic voting local political contexts, future research is needed to more directly assess the role of self-interest in public support for tax-the-rich policies in more politically conservative state contexts.

Supplemental Material

Online_Appendix – Supplemental material for Inequality Federalism and Economic Self-Interest in Subnational Progressive Tax Politics

Supplemental material, Online_Appendix for Inequality Federalism and Economic Self-Interest in Subnational Progressive Tax Politics by Benjamin J. Newman and Paul Teten in Political Research Quarterly

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Benjamin J. Newman

Supplemental Material

Supplemental materials and replication materials for this article are available with the manuscript on the Political Research Quarterly (PRQ) website.

Notes

References

Ballard-Rosa

Cameron

Martin

Lucy

Scheve

Kenneth

. 2017. “The Structure of American Income Tax Policy Preferences.” The Journal of Politics 79 (1): 1–16.

Bartels

Larry M.

2008. Unequal Democracy: The Political Economy of the New Gilded Age. Princeton: Princeton University Press.

Brandolini

Andrea

Smeeding

Timothy M.

2006. “Patterns of Economic Inequality in Western Democracies: Some Facts on Levels and Trends.” PS: Political Science & Politics 39 (1): 21–26.

Chong

Dennis

Citrin

Jack

Conley

Patricia

. 2001. “When Self-Interest Matters.” Political Psychology 22 (3): 541–70.

Condon

Meghan

Wichowsky

Amber

. 2019. “Inequality in the Social Mind: Social Comparison and Support for Redistribution.” The Journal of Politics 82 (1): 149–161.

Conover

Pamela Johnston

Feldman

Stanley

. 1981. “The Origins and Meaning of Liberal/Conservative Self-Identifications.” American Journal of Political Science 25 (4): 617–45.

Doherty

Kiley

Johnson

2018. “Public Has Mixed Expectations for New Tax Law.” Pew Research Center, January. https://www.people-press.org/2018/01/24/public-has-mixed-expectations-for-new-tax-law/

Doherty

Daniel

Gerber

Alan S.

Green

Donald P.

2006. “Personal Income and Attitudes toward Redistribution: A Study of Lottery Winners.” Political Psychology 27 (3): 441–58.

Fingerhut

Hannah

. 2017. “More Americans Favor Raising than Lowering Tax Rates on Corporations, High Household Incomes.” FacTank: Numbers in the News. Pew Research Center. https://www.pewresearch.org/fact-tank/.

10.

Fong

Christina

. 2001. “Social Preferences, Self-Interest, and the Demand for Redistribution.” Journal of Public Economics 82 (2): 225–46.

11.

Franko

William

Tolbert

Caroline J.

Witko

Christopher

. 2013. “Inequality, Self-Interest, and Public Support for “Robin Hood” Tax Policies.” Political Research Quarterly 66 (4): 923–37.

12.

Franko

William

Witko

Christopher

. 2018. The New Economic Populism: How States Respond to Economic Inequality. Oxford: Oxford University Press.

13.

Freeman

Rogers J.

Rogers

Joel

. 2007. “The Promise of Progressive Federalism.” In Remaking America: Democracy and Public Policy in an Age of Inequality, edited by Soss

Joe

Hacker

Jacob S.

Mettler

Suzanne

, 205–27. New York: Russell Sage.

14.

Goren

Paul

. 2005. “Party Identification and Core Political Values.” American Journal of Political Science 49 (4): 881–96.

15.

Gramlich

John

. 2009. “State Tax Hikes Take Aim at Top Earners.” Stateline Article, Pew Charitable Trusts, September 2. https://www.pewtrusts.org/en/research-and-analysis/blogs/stateline/2009/09/02/state-tax-hikes-take-aim-at-top-earners.

16.

Henchman

Joseph

. 2012. “Trend #1: ‘Millionaire’s Taxes.’” Fiscal Fact #313. The Tax Foundation, Washington, DC.

17.

Kane

John V.

Mason

Liliana

Wronski

Julie

. 2019. “Who’s at the Party? Group Sentiments, Knowledge, and the Social Bases of Political Partisanship.” Paper presented to the American Political Science Association Annual Meeting, Washington DC, August.

18.

Kelly

Nathan J.

Enns

Peter K.

2010. “Inequality and the Dynamics of Public Opinion: The Self-reinforcing Link between Economic Inequality and Mass Preferences.” American Journal of Political Science 54 (4): 855–70.

19.

Kelly

Nathan J.

Witko

Christopher

. 2012. “Federalism and American Inequality.” The Journal of Politics 74 (2): 414–26.

20.

Kertscher

Tom

. 2017. “Paul Ryan Claims 1986 Tax Reform, Like the Current One, Had Low Public Support Just before Passage.” @Politifact, December 18. https://www.politifact.com/wisconsin/statements/2017/dec/18/paul-ryan/paul-ryan-claims-1986-tax-reform-current-one-had-l/.

21.

Kluegel

James R.

Smith

Eliot R.

1986. Beliefs about Inequality: Americans’ Views of What Is and What Ought to Be. Aldine De Gruyter.

22.

Lacy

Dean

. 1998. “Electoral Support for Tax Cuts: A Case Study of the 1980 American Presidential Election.” American Politics Quarterly 26 (3): 288–307.

23.

Lewis-Beck

Michael S.

Norpoth

Helmut

Jacoby

William G.

Weisberg

Herbert F.

2008. The American Voter Revisited. University of Michigan Press.

24.

Lupia

Arthur

Levine

Adam Seth

Menning

Jesse O.

Sin

Gisela

. 2007. “Were Bush Tax Cut Supporters ‘Simply Ignorant?’ A Second Look at Conservatives and Liberals in ‘Homer Gets a Tax Cut.’” Perspectives on Politics 5 (4): 773–84.

25.

Matsusaka

John G.

2004. “Direct Democracy Works.” Journal of Economic Perspectives 19 (2): 185–206.

26.

McCall

Leslie

. 2013. The Undeserving Rich: American Beliefs about Inequality, Opportunity, and Redistribution. Cambridge University Press.

27.

Meltzer

Allan H.

Richard

Scott F.

1981. “A Rational Theory of the Size of Government.” Journal of Political Economy 89 (4): 914–27.

28.

Miller

Arthur H.

Wlezien

Christopher

Hildreth

Anne

. 1991. “A Reference Group Theory of Partisan Coalitions.” The Journal of Politics 53 (4): 1134–49.

29.

Ogorzalek

Thomas

Piston

Spencer

Puig

Luisa Godinez

. 2019. “Nationally Poor, Locally Rich: Income and Local Context in the 2016 Presidential Election.” Electoral Studies. Published electronically September 23. https://doi.org/10.1016/j.electstud.2019.102068.

30.

Piston

Spencer

. 2018. Class Attitudes in America: Sympathy for the Poor, Resentment of the Rich, and Political Implications. Cambridge University Press.

31.

Roberts

Michael L.

Hite

Peggy A.

Bradley

Cassie F.

1994. “Understanding Attitudes toward Progressive Taxation.” Public Opinion Quarterly 58 (2): 165–90.

32.

Robison

Joshua

Moskowitz

Rachel L.

2019. “The Group Basis of Partisan Affective Polarization.” The Journal of Politics 81 (3): 1075–79.

33.

Sears

David O.

Henry

Patrick J.

2003. “The Origins of Symbolic Racism.” Journal of Personality and Social Psychology 85 (2): 259–275.

34.

Sears

David O.

Lau

Richard R.

Tyler

Tom R.

Allen

Harris M.

1980. “Self-Interest vs. Symbolic Politics in Policy Attitudes and Presidential Voting.” The American Political Science Review 74 (3): 670–84.

35.

Soroka

Stuart N.

Wlezien C

Christopher

. 2008. “On the Limits to Inequality in Representation.” PS: Political Science & Politics 41 (2): 319–27.

36.

Stone

Chad

Trisi

Danilo

Sherman

Arloc

Beltrán

Jennifer

. 2017. “A Guide to Statistics on Historical Trends in Income Inequality.” Center on Budget and Policy Priorities. https://www.cbpp.org/research/poverty-and-inequality/a-guide-to-statistics-on-historical-trends-in-income-inequality.

37.

Tax Policy Center. 2017. “Distributional Analysis of the Conference Agreement for the Tax Cuts and Jobs Act.” December 18. https://www.taxpolicycenter.org/sites/default/files/publication/150816/2001641_distributional_analysis_of_the_conference_agreement_for_the_tax_cuts_and_jobs_act.pdf.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.68 MB