Ethnic identities, public spending, and political regimes

  • Blog Post Date 07 December, 2020
  • Articles
  • Print Page

Do democracies discriminate less against minorities as compared to non-democracies? How does the dominance of an ethnic group affect discrimination under various political regimes? Presenting a theoretical model to address these questions, this article shows that political regime changes can favour or disfavour minorities – in terms of allocation of spending on public goods – depending on the size of the ethnic majority.


Discrimination against minorities – ethnic, religious, linguistic, etc. – sometimes takes an overt form via directed violence, forcible segregation (residential and/or occupational), and sometimes a more covert form, working through discrimination in the labour market (manifesting in hiring decisions, glass ceilings, etc.), or even through the public offices by provision of inferior public goods in regions where large proportions of minorities reside or by excluding them from its use (roads, infrastructure, health facilities, educational institutions, etc.). The role of political institutions is potentially important in the context of such systematic exclusion of segments of the population, which is damaging not only from a normative perspective but also due to potential economic inefficiencies.

In our research, we seek to link political institutions with the issue of possible discrimination by asking the following questions: Do democracies discriminate less against minorities as compared to non-democracies? Recent events suggest otherwise (Ghosh and Mitra 2020). Consider the surge in violence against the Muslim Rohingya community after a democratically elected government assumed power in Myanmar (Lindblom et al. 2015).On the other hand, there was an increase in inter-ethnic cooperation in Rwanda under President Kagame's quasi-autocratic rule (Blouin and Mukand 2018). How does the dominance of an ethnic group affect discrimination under various political regimes? For Rwanda, Indonesia, and Myanmar, the Hutu, Javanese, and Bamar communities being numerically the largest ethnic groups, respectively, has made a difference to the outcomes as far as the treatment of minorities is concerned. This issue is in keeping with evidence of the treatment of minority communities in today’s India as well, without there being a formal transition from democracy to dictatorship in the country (Gettleman et al 2020, Gettleman and Raj 2017, Schultz 2019)

We put forward a tractable theory that attempts to address such questions. In our model, political leaders (democratically elected or not) decide on the allocation of spending on different types of public goods – a general public good and an ethnically-targetable public good that benefits the majority ethnic group while imposing a cost on the other minorities. On the allocation of public spending, among other things, democracies are perceived to be superior to non-democracies (for example, Acemoglu et al. 2019, Deacon 2009, Tavares and Wacziarg 2001), so the question we raise here is whether in this context the issue of discrimination against minorities could be mitigated by superior institutional structures like democracy.1

Broad theoretical framework

Our theory considers two alternative political regimes – democracy and dictatorship. Our work is related to Deacon (2009) where the differing incentives of political leaders from different regimes (democracy/dictatorship) are discussed in a theoretical framework. In our model, the society is composed of a dominant ethnic group and an amalgamation of many other (minority) groups. Irrespective of the political regime, one of the main tasks of the government is to allocate public spending, which has an important role to play in the economy, particularly in boosting output and economic growth. For political parties within a democracy, their terms in office would depend quite critically on this. For dictators, who are not elected through popular mandate, there is an alternative incentive to direct public spending in a certain way – they would typically embezzle a portion for themselves, while also ensuring that they minimise the chances of a popular uprising.

Notion of discrimination

We introduce the notion of discrimination in this setting in the following manner – two kinds of public spending are possible in this society. The first type is expenditure on a ‘general’ public good that benefits everyone irrespective of their ethnic background, while the other, an ‘ethnic’ public good, benefits only the dominant ethnic group. Moreover, this latter good apart from being exclusionary imposes a direct cost (psychological, material, etc.) on the members of the non-dominant ethnic groups. Instances of such targetting are not difficult to come by – for example, a formal proclamation of ethnic identities, in particular, delineating the national identity in terms of that of the dominant ethnic group (via the funding of ethnic or cultural-specific goods, festivals, etc.), which would reduce the stature of the minorities to ‘second-class citizens’. Hence, our stylised view of discrimination is in terms of a positive amount spent on the ethnic good – the greater the spending on the ethnic good, the higher the discrimination. Given that this ‘ethnic good’ is actually publicly provided, the theory we develop is pertinent to overt forms of discrimination – specifically, where the State has the potential to favour certain segments of society at the cost of others. There is, however, variation in the preferences for this ‘ethnic’ good within the dominant ethnic group – some value it more than others.2

Competing for citizens’ votes

We consider a democratic setting with two parties that compete for the citizens' votes by each proposing tax rates on incomes and thereby promising budgetary allocations on the two public goods. As in Acemoglu and Robinson (2006), we assume that citizens of any ethnic group can either be poor or rich. Here, we show that the ‘equilibrium’3 allocation of public spending depends heavily upon the extent of ethnic dominance. Below a certain threshold level of ethnic dominance, the entire budget is spent on the general public good. Above this threshold, the spending is entirely on the ethnic public good.4 This is intuitive, as in the absence of a ‘large’ dominant group, political parties would strive to compete for votes from all sections of the population (and hence invest in the general public good), while in the presence of such a dominant group, the parties would spend all of their energies in catering to that group (thereby investing in the ethnic good) even at the cost of antagonising the minorities. The fact that parties can adjust the tax rate suitably to garner support among the different income groups does not interfere with this core logic.

‘Shadow of democracy’

In the case of a dictatorial regime, there is no explicit role for political parties. The dictator decides on the tax rate and the allocation of public spending with largely two considerations in mind – appropriation of the public funds, or ‘rents’, and surviving any potential uprising by the citizens.5 In the eventuality of a successful revolt, there is a return to the two-party democratic regime and the dictator is disallowed from appropriating any amount of the public budget. Thus, the dictator must factor in how the different income-earners within the ethnic groups will react – that is, support a rebellion or not – when they make their public spending allocation. Clearly, the decision by any citizen would depend upon what they think the alternative scenario (in this case, democracy) will deliver to them. Since, what a democracy delivers depends upon how large the dominant ethnic group of the society is, the dictator effectively operates in this ‘shadow of democracy’. Therefore, the dictator's actions also depend upon the level of ethnic dominance.

Level of ethnic dominance is key

We show that when ethnic dominance is lower than a certain threshold, the dictator tilts spending (if any) entirely towards the dominant ethnic group. This threshold corresponds to the one in democracy below which general public spending is the only equilibrium allocation. When ethnic dominance is sufficiently high, the dictator may invest only in the general public good; in fact, the spending on the ethnic good (if any) is strictly lower than that under democracy. In other words, in society with little ethnic dominance the dictator will actually only cater to the dominant ethnicity while neglecting the minorities. It is precisely a society with a large dominant ethnic group that will witness little or no discrimination. Observe that this is completely contrary to the equilibrium policy under democracy. The intuition for this result is the following: with low ethnic dominance, the minority group has a strong incentive to rebel since they know that they will benefit from the general public spending in case the dictator is ousted and elections take place. So, dissuading them is too costly for the dictator. In order to prevent members of the dominant group from joining the rebellion, targeted ethnic spending has to be offered to that group by the dictator. Alternatively, the dictator may simply cater to the rich citizens by lowering tax rates (this would benefit the rich since the model considers a proportional income tax rate) and not providing any public spending.

Conversely, with high ethnic dominance, the dominant group has an incentive to rebel since under democracy the entire spending will be directed towards them (maximum possible discrimination). In this situation, the minority groups will typically not rebel since democracy will not bring them any enjoyment of the public spending. Therefore, in order to dissuade some members of the majority from rebelling, a positive amount of only the general public good may be offered by the dictator to both majority and minority groups. Discrimination need not be optimal from the dictator's perspective since under democracy the entire spending would be in favour of the dominant ethnic group. Hence the dictator tries to dissuade rebellion by committing to little or no discrimination. As a result, the pattern of discrimination – particularly, how it varies with the size of dominant ethnic majority – is strikingly different in a democracy as opposed to a dictatorship.

Application to examples of changes in political regime or dominant ethnic groups

Our theory can be used to interpret certain historical events like the changing nature of Hutu-Tutsi relations in Rwanda, the treatment of Chinese Indonesians during and after the Suharto regime, and more recently, the issue of persecution of the Rohingya community in Myanmar. Each of these scenarios involve a change in political regime, and when viewed through the lens of our model, appear to be consistent with our predictions.

One may even apply the core logic of our theory to the Indian context, which does not involve a transition from democracy to dictatorship (or vice versa). In the case of India, one may plausibly argue that the notion of a dominant ethnic group6 – defined along religious lines – has really crystallised under the recent regime of the current ruling party, the Bharatiya Janata Party (BJP). The rallying cry of ‘Hindutva’ (ideology focussed on the Hindu religion) has undoubtedly provided a sense of identity for large sections of Hindus throughout the country. So, one may consider any of the pre-BJP regimes as a democracy characterised by a diffused dominant ethnic group, which stands in a marked contrast to the current BJP-ruled State where there is a clearly-defined sizeable dominant ethnic group – namely, the Hindus who endorse the idea of ‘Hindutva’. In this sense, under democracy, we now witness an increase in the size of the dominant ethnic group. Our model suggests that this would potentially lead to more ethnic spending at the cost of general public good provision. In other words, greater discrimination against minorities becomes more likely under the current political regime. There is a growing body of evidence which suggests likewise. The uptick in mob lynching in the recent years (Basu 2020) and the hostility against Muslims following the onset of the Covid-19 pandemic (Gettleman et al. 2020) consistently indicate such a pattern.

It is important to note that although our analysis demonstrates that minorities may face less discrimination under dictatorships, we do not claim that democracy necessarily imposes the ‘tyranny of the masses’: what is quite critical is the size of the dominant ethnic group, which is something that changes slowly over time. In conclusion, our findings suggest that extra safeguards (reservation of posts, quotas, etc.) for the most vulnerable need to be in place so as to rescue minorities from unfair treatment in electoral as opposed to liberal democracies.

I4I is now on Telegram. Please click here (@Ideas4India) to subscribe to our channel for quick updates on our content.


  1. The paper by Mukand and Rodrik (2020), in distinguishing between different regimes (electoral and liberal democracies) in terms of the discrimination against minorities resonates with the main theme in our work. They, however, focus on the conditions as to when liberal democracy may arise. Liberal democracy is one where individual rights and freedoms are recognised, and the rule of law prevails, while an electoral democracy simply refers to a representative form of government based on elections, where the other attributes of a liberal democracy may not be present.
  2. The literature that links ethnic diversity with public goods provision is important in many respects in the context of our research: see, for example, Alesina et al. (1999), Alesina and La Ferrara (2005), Miguel and Gugerty (2005), but in most of this literature the focus has been on coordination issues arising from taste diversity.
  3. Equilibrium is a state where aggregate supply and aggregate demand are in balance.
  4. Fernandez and Levy (2008) show how diversity in preferences affects the basic conflict between rich and poor in a framework where there are variations across people both in preferences and in incomes, and in which political parties and party platforms are endogenous (in other words the parties are formed from coalitions of individuals, and parties thus formed propose policy platforms, that is, both parties and platforms are endogenous). However, they do not deal with an autocratic setup and hence do not engage with a comparison across regimes for minorities.
  5. In an ethnically divided society with weak institutions, Padro-i-Miquel (2007) argues how survival is possible for grossly inefficient rulers who often extract enormous rents. But the issue of one group being numerically/politically dominant is absent there. Also, our comparison across different regimes (democracy and dictatorship) is not the focus of that paper. By contrast, Burgess et al. (2015) do perform a comparison across regimes, but they are also concerned with the ethnic identity of the political leader, unlike us.
  6. Hindus constitute over 80% of the total Indian population, while Muslims – numbering around 200 million – are the largest religious minority.

Further Reading

  • Acemoglu, Daron, Suresh Naidu, Pascual Restrepo and James A Robinson (2019), “Democracy Does Cause Growth”, Journal of Political Economy, 127(1): 47-100.
  • Acemoglu, D and JA Robinson (2006), Economic Origins of Dictatorship and Democracy, Cambridge University Press, Cambridge UK.
  • Alesina, Alberto and Eliana La Ferrara (2005), “Ethnic Diversity and Economic Performance”, Journal of Economic Literature, 43(3): 762-800.
  • Alesina, Alesina, Reza Baqir and William Easterly (1999), “Public Goods and Ethnic Divisions”, Quarterly Journal of Economics, 114(4):1243-1284.
  • Basu, D (2020), ‘Majoritarian Politics and Hate Crimes Against Religious Minorities: Evidence from India, 2009—2018’, Political Economy Research Insititute, University of Massachusetts, Amherst, Working Paper No. 493.
  • Blouin, Arthur and Sharun W Mukand (2019), “Erasing Ethnicity? Propaganda, Nation Building and Identity in Rwanda”, Journal of Political Economy, 127(3):1008-1062.
  • Burgess, Robin, Remi Jedwab, Edward Miguel, Amit Morjaria and Gerard Padro-i-Miquel (2015), “The value of democracy: Evidence from road building in Kenya”, American Economic Review, 105(6):1817-1851.
  • Deacon, Robert T (2009), “Public good provision under dictatorship and democracy”. Public Choice, 139: 241-262.
  • Fernandez, Raquel and Gilat Levy (2008), “Diversity and Redistribution”, Journal of Public Economics, 92(5-6):925- 943.
  • Gettleman J, K Schultz and S Raj (2020), “In India, Coronavirus Fans Religious Hatred”, The New York Times, April 12.
  • Gettleman J and S Raj (2017), ‘His Defense of Hindus Was to Kill a Muslim and Post the Video’, The New York Times, 8 December.
  • Ghosh, Sugata and Anirban Mitra (2020), “Ethnic identities, public spending, and political regimes”, mimeo.
  • Lindblom, A, E Marsh, T Motala and K Munyan (2015). “Persecution of the Rohingya Muslims: Is Genocide Occurring in Myanmar's Rakhine State?”, Yale Law School Report.
  • Miguel, Edward and Mary Kay Gugerty (2005), “Ethnic diversity, social sanctions, and public goods in Kenya”, Journal of Public Economics, 89(11):2325-2368.
  • Mukand, Sharan W and Dani Rodrik (2020), "The Political Economy of Liberal Democracy", The Economic Journal, 130 (627): 765–79.
  • Padro-i-Miquel, Gerard (2007), “The Control of Politicians in Divided Societies: The Politics of Fear”, Review of Economic Studies, 74(4):1259-1274.
  • Schultz, Kai (2019), ‘Murders of Religious Minorities in India Go Unpunished, Report Finds’, The New York Times, 18 February.
  • Tavares, José and Romain Wacziarg (2001), “How Democracy Affects Growth”, European Economic Review, 45(8):1341-1378
No comments yet
Join the conversation
Captcha Captcha Reload

Comments will be held for moderation. Your contact information will not be made public.

Related content

Sign up to our newsletter