Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Behavioural economics

Visible inequality breeds more inequality

Experiments suggest that when people can see wealth inequality in their social network, this propels further inequality through reduced cooperation and reduced social connectivity. See Letter p.426

Inequality is a growing concern in many societies1. Like most important social phenomena, it is a complex issue that has many interacting sources and consequences1,2,3. To understand inequality and its dynamics over time, multiple theoretical and empirical approaches are necessary. In this issue, Nishi et al.4 (page 426) use laboratory-style experiments (conducted online) to study how the visibility of wealth inequality in people's social environment shapes the behavioural dynamics of inequality. The attraction of an experimental approach is that it allows the control of factors that are inherently uncontrollable in naturally occurring data. Crucially, for example, the experimenter can control the initial level of inequality and see how inequality evolves as a function of people's behaviour alone5,6.

Nishi and colleagues' experimental model used an assessment of people's willingness to contribute to public goods to test how initial wealth inequality and the structure of the social network influence the evolution of inequality. The researchers were particularly interested in the role of visibility of wealth — can mere observation of your neighbour's wealth lead to more inequality over time, even if such information does not change economic incentives? Visible wealth might have a psychological effect by triggering social comparisons and thereby influencing economic choices that have repercussions for inequality3.

In their online laboratory, the researchers endowed all participants with tokens, worth real money. The endowment differed across individuals and treatments: in a treatment without inequality, all participants initially received the same number of tokens; in a low-inequality treatment, participants had similar but different initial endowments; and in the high-inequality treatment there was a substantial starting difference between participants.

The groups typically comprised 17 people arranged at random in a social network in which, on average, about 5 people were linked ('neighbours'). In each of the 10 rounds of the following game, participants had to decide whether to behave pro-socially ('cooperate') by reducing their own wealth by 50 tokens per connected neighbour to benefit each of them by 100 tokens, or to behave pro-selfishly ('defect') by keeping their tokens for themselves. These decisions had consequences for accumulated wealth levels and inequality. At the end of each round, the subjects learnt whether their neighbours had cooperated or defected and 30% of participants were given the opportunity to change their neighbour, that is, to either sever an existing link or to create a new one.

A crucial manipulation in this experiment was wealth visibility. Under invisible conditions, the participants could observe only their own accumulated wealth. Under visibility, they could see the accumulated wealth of their connected neighbours but not the whole network. Thus, there were six conditions in total: three levels of initial wealth inequality in each of the two visibility conditions.

The results are complex but illuminating. The authors find that, under high initial wealth inequality, visibility of neighbours' accumulated wealth increases inequality over time relative to the invisibility condition, although absolute inequality decreases over time under both visibility conditions. The reason for the relative increase under visibility is that inequality drops only moderately, whereas under invisibility the reduction in inequality is substantial. By contrast, in the case of initial wealth equality, inequality increases — similarly in both visibility conditions. Under moderate initial inequality, visibility leads to a small increase in inequality relative to invisibility.

Visibility of wealth also leads to lower social welfare, as measured by overall wealth (Fig. 1). By the end of the experiment, total accumulated wealth was substantially larger in the three conditions with invisible wealth than in the three conditions with visible wealth. The reason for this is that cooperativeness was lower under the condition of visible wealth compared to invisible wealth, and there were fewer links in the social network.

Figure 1: Wealth on display.

Florian Kopp/Robert Harding Picture Library

Nishi et al.4 use an experimental game to show that, when people can see the wealth of others whom they are linked with in a social network, inequality increases and the number of social connections decreases.

The most striking insight from these findings is the effect of wealth visibility on the dynamics of inequality: conspicuous inequality breeds more inequality. Although visibility of wealth does not change economic incentives in this experimental scenario, it invites social comparisons that, for various reasons3,7 worth exploring further, undermine cooperation and diminish social ties. This observation adds to existing8,9, but sparse, evidence that public information about individual pay-offs leads to more competition, which in a public-goods setting triggers more 'free-riding' by individuals (defecting when others cooperate), to improve their own pay-offs.

Nishi and colleagues' findings raise several intriguing methodological questions for future studies. For example, how much influence does the social network and its rewiring have on the main results of this experiment? Modelling interactions using a social network is certainly realistic, but is it crucial for the emergence of visibility effects in inequality? Another question concerns the result that visibility of wealth matters much less under initial equality of wealth. This is surprising, given that inequality of wealth increases over time and visibility effects should kick in, according to the results from the treatments with initial inequality. It is possible that these experiments, which used only ten iterations, might have been too short to allow for visibility effects arising as inequality grows.

The results also suggest substantive questions worthy of further research. As well as understanding the role of visibility of wealth (or pay-offs more generally) for cooperation, it would be interesting to gather evidence about how people's pro-social attitudes are affected by the ever-increasing amount of information about other people's consumption (as a signal of their wealth)10, which nowadays is spread on an almost global scale by social media. And how do visibility and social comparisons affect the dynamics of inequality when the relevant game is not one of cooperation but of competition? This is interesting because, in many interactions in our modern societies, not only initial endowments (wealth) matter but also resources that are allocated as people compete for scarce rewards — good jobs, for instance11.

These are just some questions that can be investigated with the experimental model put forward by Nishi and colleagues. Their most general contribution is to showcase the power of experiments to contribute to our understanding of the behavioural dynamics of inequality.Footnote 1


  1. 1.

    See all news & views


  1. 1

    Atkinson, A. B. Inequality: What Can Be Done? (Harvard Univ. Press, 2015).

    Book  Google Scholar 

  2. 2

    Chin, G. & Culotta, E. Science 344, 818–821 (2014).

    ADS  Article  Google Scholar 

  3. 3

    Frank, R. H. Falling Behind: How Rising Inequality Harms the Middle Class (Univ. California Press, 2013).

    Google Scholar 

  4. 4

    Nishi, A., Shirado, H., Rand, D. G. & Christakis, N. A. Nature 526, 426–429 (2015).

    ADS  CAS  Article  Google Scholar 

  5. 5

    Sadrieh, A. & Verbon, H. A. A. Eur. Econ. Rev. 50, 1197–1222 (2006).

    Article  Google Scholar 

  6. 6

    Gächter, S., Mengel, F., Tsakas, E. & Vostroknutov, A. (2014).

  7. 7

    Dohmen, T., Falk, A., Fliessbach, K., Sunde, U. & Weber, B. J. Publ. Econ. 95, 279–285 (2011).

    Article  Google Scholar 

  8. 8

    Huck, S., Normann, H.-T. & Oechssler, J. Int. J. Industr. Organiz. 18, 39–57 (2000).

    Article  Google Scholar 

  9. 9

    Nikiforakis, N. Games Econ. Behav. 68, 689–702 (2010).

    Article  Google Scholar 

  10. 10

    Veblen, T. The Theory of the Leisure Class (Macmillan, 1899).

    Google Scholar 

  11. 11

    Hopkins, E. & Kornienko, T. Am. Econ. J.: Microeconomics 2(3), 106–137 (2010).

    Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Simon Gächter.

Related links

Related links

Related links in Nature Research

Behavioural economics: A caring majority secures the future

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Gächter, S. Visible inequality breeds more inequality. Nature 526, 333–334 (2015).

Download citation

Further reading

  • Monetary Intelligence and Behavioral Economics: The Enron Effect—Love of Money, Corporate Ethical Values, Corruption Perceptions Index (CPI), and Dishonesty Across 31 Geopolitical Entities

    • Thomas Li-Ping Tang
    • , Toto Sutarso
    • , Mahfooz A. Ansari
    • , Vivien K. G. Lim
    • , Thompson S. H. Teo
    • , Fernando Arias-Galicia
    • , Ilya E. Garber
    • , Randy Ki-Kwan Chiu
    • , Brigitte Charles-Pauvers
    • , Roberto Luna-Arocas
    • , Peter Vlerick
    • , Adebowale Akande
    • , Michael W. Allen
    • , Abdulgawi Salim Al-Zubaidi
    • , Mark G. Borg
    • , Bor-Shiuan Cheng
    • , Rosario Correia
    • , Linzhi Du
    • , Consuelo Garcia de la Torre
    • , Abdul Hamid Safwat Ibrahim
    • , Chin-Kang Jen
    • , Ali Mahdi Kazem
    • , Kilsun Kim
    • , Jian Liang
    • , Eva Malovics
    • , Alice S. Moreira
    • , Richard T. Mpoyi
    • , Anthony Ugochukwu Obiajulu Nnedum
    • , Johnsto E. Osagie
    • , AAhad M. Osman-Gani
    • , Mehmet Ferhat Özbek
    • , Francisco José Costa Pereira
    • , Ruja Pholsward
    • , Horia D. Pitariu
    • , Marko Polic
    • , Elisaveta Gjorgji Sardžoska
    • , Petar Skobic
    • , Allen F. Stembridge
    • , Theresa Li-Na Tang
    • , Caroline Urbain
    • , Martina Trontelj
    • , Luigina Canova
    • , Anna Maria Manganelli
    • , Jingqiu Chen
    • , Ningyu Tang
    • , Bolanle E. Adetoun
    •  & Modupe F. Adewuyi

    Journal of Business Ethics (2018)


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing