Inequality is the unequal distribution of income, wealth and opportunities between different groups in society. Over recent decades, inequality has risen substantially in many advanced economies. For instance, in the United Kingdom, the Institute for Fiscal Studies estimates that the share of income accruing to the top 1% rose from approximately 6% in 1980 to 15% in 2020, reflecting a significant increase in income concentration at the top (Blundell, 2024). Redistribution is often regarded as the primary pathway to remedy post-market inequalities, with policies such as progressive taxation or benefit payments reallocating wealth from higher to lower-income groups. As a government-led solution, redistribution depends not only on available resources but also on electoral support. Meltzer and Richard (1981) argue that rising inequality should increase public demand for redistribution, shift the median voter’s preferences, creating political incentives for governments to expand redistributive policies. However, despite widening inequality, public policy in many advanced democracies has become increasingly conservative since the 1960s (Hetherington, 2005). This essay argues that the gap between theoretical predictions and observed political outcomes is driven only partly by public distrust in government. Another important factor is the lack of information individuals have about the actual impact of redistributive policies. Lastly, this essay posits that the main driver of attitudes towards redistribution is the influence of personal and ideological beliefs on preferences.
While distrust in government is a key factor in weakening support for redistribution, restoring trust alone is unlikely to generate substantial demand for redistributive policies.
Redistribution requires citizens to entrust governments with the collection and reallocation of resources to reduce inequality while making material sacrifices. This creates a principal-agent problem in which citizens (the principals) delegate the implementation of redistributive policies to governments (the agents). However, perceptions that politicians prioritise re-election and personal interests over effective redistribution appear to have eroded public trust (Hetherington 2005). Kuziemko et al. (2015) find that over 89% of their sample agreed that “Politicians in Washington work to enrich themselves instead of working for the benefit of the majority of citizens”. Moreover, in the United States, 65% of adults believe political candidates run for office primarily to serve their own personal interests (Pew Research Center, 2022). Consistent with this, trust in the federal government fell from 77% in 1964 to just 18% in 2025 (Pew Research Center, 2025). Stantcheva (2024) contends that when voters doubt the government’s willingness or ability to redistribute resources effectively, rising inequality will not necessarily translate into greater support for redistribution.
Kuziemko et al. (2015) provide evidence for a causal link between political trust and support for redistributive preferences. Using an experimental treatment that reduced trust in government without significantly altering perceptions of inequality or poverty, they found a substantial decline in support for almost all poverty-alleviation policies. The sole exception was the minimum wage, a policy that does not rely on direct government transfers. This suggests that support for redistribution is influenced less by trust in government as an institution and more by confidence in the government’s ability to implement redistributive policies effectively.
Yet, while declining trust has contributed to falling support for redistribution, this does not imply that increasing trust in government will necessarily increase support for redistribution. Peyton (2020) finds that experimentally increasing political trust to a degree sufficient to eliminate the longstanding partisan gap in trust between Democrats and Republicans does not result in any significant increase in support for redistribution. This asymmetry suggests that although declining trust can weaken support for redistribution, restoring trust alone is insufficient to alter long-held distributive preferences. Moreover, longer-term evidence further weakens the case for political trust as a primary determinant of redistributive attitudes. Using longitudinal data, Devine (2025) finds that although individuals with higher levels of political trust are generally more supportive of redistribution, changes to levels of government trust do not produce lasting impacts on redistributive preferences. Hence, political trust appears more likely correlated with support for redistribution rather than causally responsible for it in the long-term.
Furthermore, political trust has not declined uniformly since the 1960s. Hetherington and Rudolph (2015) find that individuals’ trust in government fluctuates according to which party is in power, rising among the supporters of the governing party and falling among that of the opposition. This suggests that political trust is, at least in part, a consequence of partisan identity rather than an independent determinant of redistributive preferences. Therefore, distrust in government is unlikely to be the primary inhibitor of redistribution. Since, levels of trust themselves are shaped by partisanship, personal and ideological beliefs are likely to exert a more fundamental influence on redistributive preferences.
Voters’ lack of information is another important barrier to redistribution; however, increasing access to it alone is unlikely to increase support for redistributive policies.
Gimpelson and Treisman (2017) contend that low levels of social spending occur as a result of people’s misperceptions about how extreme the distribution is and where their income level lies within the wealth distribution. They argue that these perceptions regarding inequality are more influential than actual inequality levels in generating demand for redistributive policies. If the median voter is unaware of where they fall in the distribution, they will be unlikely to shift their preferences undermining Meltzer and Richard’s model. It therefore follows that correcting these misperceptions should result in higher demand for welfare spending. Kuziemko et al. (2015) find that concerns about inequality are highly responsive to information, with their informational treatment increasing the share of respondents who viewed inequality as a “very serious problem” by over 35%. However, this heightened concern rarely translated into greater support for redistributive policies. This suggests that voters’ lack of information regarding the operation and incidence of specific policies may be a greater barrier to redistribution than ignorance of inequality itself. This conclusion is supported by Slemrod (2006), who finds that correcting the misconception, held by 43.5% of respondents, that high-income earners would pay more tax under a flat income tax rate system, wrongly believing the current system to be regressive, would reduce support for the policy by around 10 percentage points. Correcting misconceptions about the distributional effects of policies therefore appears more effective at changing preferences than increasing awareness of inequality. However, Slemrod also notes that a majority would continue to support the flat tax even after this misconception was corrected, suggesting that voters’ information gaps alone cannot fully explain opposition to redistribution.
Albeit the lack of information plays a significant role in shaping preferences, even if this information gap were closed, myopic tendencies inhibit redistribution too. Kuziemko et al. (2015) find that when asked to identify the true share of decedents subject to the estate tax, 16% of liberals answered correctly compared with only 6% of conservatives. This suggests that policy misconceptions are not merely the product of limited information but are filtered through pre-existing ideological commitments. This is validated by Stantcheva (2021) whose surveys explored not just what people know about the tax system but how they reason it. She finds that partisan led differences dictate not only policy preferences but extend all the way down to basic perceptions of h, ow the current tax system actually works. Alesina et al. (2020) describe this as a “polarization of reality” where left and right-wing respondents do not perceive the current tax system in the same way, with the left tending to view it as both less progressive and lower than the right believes it to be. Consequently, while voter myopia may inhibit support for redistribution, personal beliefs and partisanship appear to be the more fundamental obstacles to redistributive reform.
Personal beliefs regarding the causes of inequality play an important role in shaping preferences for redistribution and, consequently, the type of welfare state that emerges.
Belief in a meritocratic society fosters greater tolerance of inequality. When income is perceived to be determined primarily by individual effort, high earners are viewed as deserving of their wealth and upward mobility is seen as attainable through hard work (Son Hing et al., 2019). By contrast, financial success is believed to be driven by luck, family background, or social connections, inequality is viewed as less legitimate and support for redistribution increases. This directly influences the welfare system that is established. Alesina and Angeletos (2005) argue that these beliefs can become self-reinforcing. Societies with strong faith in social mobility tend to favour lower taxes and less redistribution, creating an equilibrium in which effort is rewarded and the role of luck is limited. As market outcomes consequently appear fair, meritocratic beliefs become self-fulfilling. Conversely, societies that view opportunities for success as constrained by class differences tend to support greater redistribution, reinforcing the belief that economic outcomes are shaped by factors beyond individual effort. This framework helps explain why the United States, with its “American Dream”, has historically maintained a less redistributive welfare state than Europe despite experiencing higher levels of inequality. In the mid-1990s, the pre-tax Gini coefficient was 38.5 in the United States compared with an average of around 29 in Europe, yet European countries generally adopted more progressive tax systems and more extensive redistributive policies (Deininger and Squire, 1996).
Consistent with this explanation, the World Values Survey found that 71% of Americans, compared with only 40% of Europeans, believed that the poor could become rich if they worked hard enough (Alesina and Angeletos, 2002). These contrasting beliefs generate different levels of support for redistribution and, consequently, different welfare state regimes: universalistic, corporatist or liberal (Esping-Andersen, 1990). Furthermore, Kumlin et al. (2018) find that universalistic welfare institutions tend to build political trust, while means-tested programs erode it. This suggests a reverse causal relationship where institutional choices societies make based on their ideological beliefs and redistributive preferences shape levels of trust in governments.
However, these welfare arrangements are sustained because existing beliefs are reinforced and polarised by partisanship. Achen and Bartels (2016) argue that voter behaviour is driven more by partisan identities, an individual’s psychological attachment to a political party, than by rational policy reasoning. This directly challenges the central assumption of the Meltzer–Richard model that voters act rationally in pursuit of their economic self-interest. Instead, voters may continue to oppose redistribution despite rising inequality when redistributive policies conflict with their partisan loyalties. Partisanship also shapes how individuals process information. Gaines et al. (2007) find that partisans interpret the same facts differently to fit their existing beliefs, rather than rationally updating their preferences in response to new evidence. Consistent with this, Brooks and Harter (2021) show that partisanship exerts a greater influence on redistributive preferences than information about inequality across twelve different redistribution measures. They find that support for taxing individuals is at least over 1.6 times more strongly influenced by partisanship than by information about inequality. These findings help explain why rising inequality does not necessarily generate greater support for redistribution. If voters interpret economic outcomes through partisan identities rather than rationally responding to their material interests, increases in inequality alone are unlikely to shift redistributive preferences.
In conclusion, the extent of redistribution is fundamentally led by the electorate’s personal and ideological beliefs instead of distrust in government. While misconceptions regarding the extent of inequality and the nature of policies certainly prevent preferences from shifting, partisanship is crucial in understanding how individuals interpret the information they are presented with. Moreover, despite distrust in government causing levels of redistributive support to fall to such low levels, reversing this decline will not be sufficient to increase support for redistribution considering the effects of partisanship in distorting understandings and in alienating certain members of the electorate from different political parties. These dependencies undermine the Meltzer-Richard model as the median voter is unlikely to behave rationally as a result of the interactions between personal beliefs, information and trust in government. In order to increase support for redistribution to remedy inequality, all three of these factors need to be considered in synergy.