Why is inequality so sticky? The political obstacles to a fairer economy

Andy Sumner Inequality, Research

Theory tells us that democracies should become more equal. So why are they still so unequal? Gideon Coolin, Emanuele Sapienza, and Andy Sumner on their new UNDP paper that unpicks the politics of inequality.

Panama City (Picture: UNDP Panama)

Joseph Stiglitz famously noted that inequality is “a choice”. In fact, a policy choice. So why don’t democracies choose to do more to address inequalities? 

That’s the topic of our new paper published by UNDP. We look at the politics of inequality and how and why governance systems do – or more often don’t – introduce new policies to address income and wealth distribution. We hope this work will be useful to anyone advocating for inequality reduction as they try to understand what is stopping policy spaces opening more for social and economic justice. 

Two facts about inequality: it’s high and it’s sticky

Let’s start here with two facts about national or ‘within-country’ inequality: first, inequality is generally high in most countries and, in much of the global South, it is at a level that is likely to hinder future income growth (so says the IMF research).

Second, inequality is kind of sticky – meaning higher inequality tends to persist over time even if it may rise or fall a bit (e.g. in Latin America in the 2000s, inequality came down somewhat from high levels).

Yet theory tells us that democracies should become more equal

Persistent and high inequality in democracies seems to contradict the theoretical, social and political expectation that democratization should lead to a fairer distribution of income and wealth in society.

Simon Kuznets argued that as countries urbanise and grow richer, the levels of inequality they experience will increase, before peaking and declining at higher levels of income. In democratic societies, he argued, taxation of capital and wealth would inevitably be introduced as the political power of lower-income groups increases. Kuznets was clear that politics have the potential to counterbalance any increase in inequality. He further contended that, in democracies, urban migrants would become politically organized, leading to redistribution. 

In a similar vein, Meltzer and Richards predicted that democracy tends to translate into lower levels of inequality. In a simple majority-rule system, there will be majority support for inequality-reducing measures when inequality is high. High inequality will incentivise political actors to advocate redistribution, and this will lead to redistributive policy – be it progressive taxation, social policy, or service provision.

If these theories were to hold, we would expect the distribution of income and wealth to be more equal in more ‘democratic’ societies (which of course can mean a wide range of things). However, the track record of democracies in reducing income and wealth inequality is mixed at best. Inequality remains high in many democracies and rather sticky. This important in a global context of democratic backsliding. There is evidence that people’s support for democracy increases when democracy is seen as delivering ‘distributional justice’. For this reason, democracies’ ability to bring about distributional justice should be regarded as central to support for democracy or the resilience of democracies.

How policy is really made

To understand why redistributive policy is not more widely adopted and what can be done, it is necessary to first get a better handle on how policy is made. We can understand this process as being driven by the interaction of three things:

  • the policy actors and networks, that contest and design or frame policy;
  • the policy institutions and context – the formal and informal structures, e.g. the legislative or party system, in which actors work; and
  • discourses and narratives that shape people’s understandings of policy.

How these interact determines the outcomes of a policy process and means the introduction of redistributive policies is more complex than implied by Kuznets or Meltzer and Richard.

Four factors that might block redistributive policies

In our survey, four factors emerge as particularly influential.

First, whether people want to see inequality-reducing policies adopted is not simply down to their ‘rational’ self-interest. For example, expectations of higher levels of income in the future, low trust in institutions to deliver policies, false perceptions of one’s position in the income distribution, cultural norms, and one’s ideological beliefs have been found to lead to lower support for redistribution among groups that would benefit from it. These can be shaped by the media environment or prevalent narratives (for instance, around ideas such as ‘meritocracy’). This means, in many cases, those that would stand to benefit in pure economic terms from redistribution may vote contrary to their material interests. 

Second, even when people do favour redistributive policies, there are often barriers to the collective action necessary to translate this wish into political demands. The existence of political parties, trade unions, or civic groups capable of speaking collectively for those with pro-redistribution preferences increases the bargaining power of lower-income groups. Where such organisations do not exist, or space for civic engagement is constrained, collective efforts to translate preferences into policy can falter. Systems where politics is organised on clientelism or identity will also tend to reduce the possibility of collective action for distributional justice.

Third, there is the obstacle posed by those with an interest in preserving high levels of inequality. Economic elites can use privileged access to capture policy processes, protect their economic position, or manipulate people’s preferences through control of media organisations and other sources of cultural influence. In extreme circumstances they may also use violence to block policies and prevent redistribution. 

Fourth, the institutions of a political system, along with the discourses and narratives in a society, can limit policy outcomes by influencing what is perceived as possible. Electoral and legislative systems can stymie policy change through building in a structural advantage for anti-redistribution groups. Constitutional provisions can be used to lock economic ideology into law. Further, when the prospect of progressive policy is raised, a common refrain is that is simply “is not possible” – how policy ideas fit into the current imagination of what is deemed acceptable matters hugely and is determined by dominant policy discourse.  

What does this mean for those advocating for tackling inequality?

So, if the goal is to understand and increase the policy space for social and economic justice, it’s important to consider:

  1. Non-economic factors determining support for redistribution across social groups such as ideology or people’s false perceptions of where they are in the income distribution.
  2. How to support collective action among lower-income groups, for instance by trade unions or other forms of civic organizing, as a crucial way through which a desire for redistribution is translated into policy.
  3. How political institutions might be structured to preserve inequality, as well as prevailing narratives – such as certain policies being labelled as not possible – that are key to enabling action for redistribution or blocking it.
  4. How elites leverage their higher levels of influence to capture the policy process and maintain their privileged position.

Andy Sumner

Andy Sumner is Professor of International Development at King’s College London


Emanuele Sapienza

Emanuele Sapienza is Global Lead - Open and Inclusive Public Sphere, UNDP


Gideon Coolin

Gideon Coolin is a Research Associate at Wageningen University & Research

You can read the full paper here: The politics of inequality: Why are governance systems not more responsive to the unequal distribution of income and wealth?

This blog first appeared on From Poverty to Power.