How does inequality aversion affect inequality and redistribution?


We investigate the effects of inequality aversion on equilibrium labor supply, tax revenue, income inequality, and median voter outcomes in a society where agents have heterogeneous skill levels. These outcomes are compared to those which result from the behavior of selfish agents. A variant of Fehr-Schmidt preferences is employed that allows the externality from agents who are “ahead” to differ in magnitude from the externality from those who are “behind” in the income distribution. We find first, that inequality-averse preferences yield distributional outcomes that are analogous to tax-transfer schemes with selfish agents, and may either increase or decrease average consumption. Second, in a society of inequality-averse agents, a linear income tax can be welfare-enhancing. Third, inequality-averse preferences can lead to less redistribution at any given tax, with low-wage agents receiving smaller net subsidies and/or high-wage individuals paying less in net taxes. Finally, an inequality-averse median voter may prefer higher redistribution even if it means less utility from own consumption and leisure.

This is a preview of subscription content, log in to check access.


  1. Ackert, L.F., Martinez-Valdez, J., Rider, M.: Social preferences and tax policy: Some experimental evidence. Econ. Inq. 45, 487–501 (2007)

    Article  Google Scholar 

  2. Algood, S.: The marginal costs and benefits of redistributing income and the willingness to pay for status. J. Public. Econ. Theory. 8, 357–77 (2006)

    Article  Google Scholar 

  3. Alesina, A., Angeletos, G.: Fairness and redistribution. Am. Econ. Rev., pp. 960–980 (2005)

    Article  Google Scholar 

  4. Alm, J., McClelland, G.H., Schulze, W.: Changing the norm of tax compliance by voting. Kyklos 52, 141–71 (1999)

    Article  Google Scholar 

  5. Atkinson, A.B.: On the measurement of inequality. J. Econ. Theory. 2, 244–263 (1970)

    Article  Google Scholar 

  6. Beckman, S.R., Formby, J., Smith, J.S., Zheng, B.: Envy, malice and pareto efficiency: an experimental examination. Soc. Choice. Welfare. 19, 349–367 (2002)

    Article  Google Scholar 

  7. Bolton, G., Ockenfels, A.: ERC: A theory of equity, reciprocity, and competition. Am. Econ. Rev. 90, 166–193 (2000)

    Article  Google Scholar 

  8. Clark, A.E., Frijters, P., Shields, M.A.: Relative income, happiness, and utility: an explanation for the easterlin paradox and other puzzles. J. Econ. Lit. 46, 95–144 (2008)

    Article  Google Scholar 

  9. Dhami, S., Al-Nowaihi, A.: Existence of a condorcet winner when voters have other regarding preferences. J. Public. Econ. Theory. 12, 897–922 (2010)

    Article  Google Scholar 

  10. Dhami, S., Al-Nowaihi, A.: Redistributive policies with heterogeneous social preferences of voters. Eur. Econ. Rev. 54, 743–759 (2010)

    Article  Google Scholar 

  11. Dorfman, R.: A formula for the gini coefficient. Rev. Econ. Stat. 61, 146–149 (1979)

    Article  Google Scholar 

  12. Easterlin, R.A.: Will raising the incomes of all increase the happiness of all. J. Econ. Behav. Organ. 27, 35–48 (1995)

    Article  Google Scholar 

  13. Fehr, E., Schmidt, K.M.: A theory of fairness, competition and cooperation. Q. J. Econ. 114, 817–868 (1999)

    Article  Google Scholar 

  14. Fehr, E., Fischbacher, U.: Why social preferences matter—the impact of non-selfish motives on competition, cooperation and incentives. Econ. J. 112, C1–C33 (2002)

    Article  Google Scholar 

  15. Forsythe, R, Horowitz, J., Savin, N.E., Sefton, M.: Fairness in simple bargaining experiments. Game. Econ. Behav 6, 347–369 (1994)

    Article  Google Scholar 

  16. Frank, R.H.: Should public policy respond to positional externalities. J. Public. Econ. 92, 1777–86 (2008)

    Article  Google Scholar 

  17. Frohlich, N, Oppenheimer, J., Kurki, A.: Modeling other-regarding preferences and an experimental test. Publ. Choice 119, 91–117 (2004)

    Article  Google Scholar 

  18. Galasso, V: Redistribution and fairness: a note. Eur. J. Polit. Econ. 19, 885–892 (2003)

    Article  Google Scholar 

  19. Hochman, H.M., Rodgers, J.D.: Pareto optimal redistribution. Am. Econ. Rev. 59, 542–557 (2012)

    Google Scholar 

  20. Höchtl, W., Sausgruber, R., Tyran, J.: Inequality aversion and voting on redistribution. Eur. Econ. Rev. 56, 1406–1421 (2012)

    Article  Google Scholar 

  21. Hopkins, E.: Inequality, happiness and relative concerns: What actually is their relationship. J. Econ. Inequal. 6, 351–72 (2008)

    Article  Google Scholar 

  22. Hwang, S., Lee, J.: Conspicuous consumption and income inequality. Oxford. Econ. Pap. 69, 279–292 (2017)

    Google Scholar 

  23. Ireland, N.J.: Status seeking, income taxation and efficiency. J. Public. Econ. 70, 99–113 (1998)

    Article  Google Scholar 

  24. Ireland, N.J.: Optimal income tax in the presence of status effects. J. Public. Econ. 81, 193–212 (2001)

    Article  Google Scholar 

  25. Ledyard, J.O.: Public goods: A survey of experimental research. In: Kagel, J.H., Roth, A.E. (eds.) The Handbook of Experimental Economics, pp 111–194. Princeton University Press, Princeton (1995)

  26. Lind, JT: Fractionalization and the size of government. J. Public. Econ. 91, 51–76 (2007)

    Article  Google Scholar 

  27. Mankiw, N., Weinzierl, M., Yagan, D.: Optimal taxation in theory and practice. J. Econ. Perspect. 23, 147–74 (2009)

    Article  Google Scholar 

  28. Meltzer, A.H., Richard, S.F.: A rational theory of the size of government. J. Polit. Econ. 89, 914–927 (1981)

    Article  Google Scholar 

  29. Tyran, J., Sausgruber, R.: A little fairness may induce a lot of redistribution in democracy. Eur. Econ. Rev. 50, 469–485 (2006)

    Article  Google Scholar 

  30. Wendner, R., Goulder, L.H.: Status effects, public goods provision and excess burden. J. Public. Econ. 92, 1968–85 (2008)

    Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Langchuan Peng.

Electronic supplementary material

Below is the link to the electronic supplementary material.

(DOCX 26.7 KB)



Claim: Wage income, wL(w,τ), is increasing in w if and only if

σ(1 − αH(w) + β[1 − H(w)]) > w(α + β)h(w) for all \(w\in [\underline {w},\overline {w} ]\).


Using (2) we have

$${wL(w,\tau )=w^{\frac{\sigma} {\sigma -1}}[\frac{(1-\tau )\{1-\alpha H(w)+\beta [1-H(w)]\}}{\xi} ]}^{\frac{1}{\sigma -1}}. $$

Now differentiate the above with respect to w to get a necessary and sufficient condition for labor income to be increasing in the wage:

$$\begin{array}{@{}rcl@{}} \frac{\partial (wL(w,\tau ))}{\partial w}&=&{w^{\frac{\sigma} {\sigma -1}}[\frac{(1-\tau )\{1-\alpha H\left( w \right)+\beta [1-H(w)]\}}{\xi} ]}^{\frac{1}{\sigma -1}-1}(\frac{1}{\sigma -1})\\&&\times(\frac{-(1-\tau )(\alpha +\beta )h(w)}{\xi} ) \end{array} $$
$$+{w^{\frac{1}{\sigma -1}}[\frac{(1-\tau )\{1-\alpha H(w)+\beta [1-H(w)]\}}{\xi} ]}^{\frac{1}{\sigma -1}}(\frac{\sigma} {\sigma -1})>0. $$

The parametric condition for \(\frac {\partial (wL(w,\tau ))}{\partial w}>0\) is thus

$${w^{\frac{1}{\sigma -1}}[\frac{(1-\tau )\{1-\alpha H(w)+\beta [1-H(w)]\}}{\xi} ]}^{\frac{1}{\sigma -1}}(\frac{\sigma} {\sigma -1})>$$
$${w^{\frac{\sigma} {\sigma -1}}[\frac{(1-\tau )\{1-\alpha H(w)+\beta [1-H(w)]\}}{\xi} ]}^{\frac{1}{\sigma -1}-1}(\frac{1}{\sigma -1})(\frac{(1-\tau )(\alpha +\beta )h(w)}{\xi} ). $$

Finally, rewriting the above yields the necessary and sufficient condition (given in (3)) that must be satisfied at all w in order for labor income to be monotonically increasing in the wage: σ(1 − αH(w) + β[1 − H(w)]) > w(α + β)h(w). □

Proof of Proposition 1

An inequality-averse individual i with an hourly wage wi provides weakly greater labor than a self-interested individual with the same hourly wage if and only if:

$$ {[\frac{(1-\tau )w_{i}\{1-\alpha H(w_{i})+\beta [1-H(w_{i})]\}}{\xi}]}^{\frac{1}{\sigma -1}}\ge {\mathrm{[[}\frac{(1-\tau )w_{i}}{\xi}]}^{\frac{1}{\sigma -1}}. $$

Rearranging the above yields the condition

$$ H(w_{i})\le \frac{\beta} {\alpha +\beta}. $$

Recall that \(\tilde {w}\) is defined as the wi such that (9) holds with equality. Since H is a strictly increasing function, (9) is satisfied if and only if \(w_{i}\le \tilde {w}\). □

Before stating the proofs of Proposition 2 and 3, define average pre-tax income as

$$\text{I}(\alpha, \beta )={[\frac{(1-\tau )}{\xi} ]}^{\frac{1}{\sigma -1}}{\int}_{\underline{w}}^{\overline{w}} {z^{\frac{\sigma} {\sigma -1}}\{\left[ 1-\alpha H\left( z \right)+\beta \left[ 1-H\left( z \right) \right] \right]^{\frac{1}{\sigma -1}}\}h(z)dz} , $$

with I(0,0) corresponding to selfish agents. It is clear that I(α,β) is increasing in β and decreasing in α. For a given τ, both average tax revenue, S = τ I(α,β), and average consumption, (1 − τ)I(α,β), and are increasing in I(α,β), so the following proofs focus on I(α,β).

Proof of Proposition 2


We know from above that I(0,0) < I(0,β). Suppose β ≤ 1. If I(β,β) ≥I(0,0), the result holds for α = β. If I(β,β) < I(0,0) < I(0,β), the result follows from the observation that I(α,β) is a continuous function. Next, suppose β > 1. If I(α,β) ≥I(0,0), the result holds for α = 1. If I(1,β) < I(0,0) < I(0,β), the result follows from the observation that I(α,β) is a continuous function.


Follows from 2(i) and Proposition 1.

Proof of Proposition 3


We first establish that I(β,β) < I(0,0) for a symmetric distribution with σ > 2. The result then follows from the fact that I(α,β) is a continuous decreasing function of α. Take the derivative of ρ(β) ≡I(β) to yield

$$ \frac{d\rho (\beta )}{d\beta} =\frac{1}{\sigma -1}{[\frac{(1-\tau)}{\xi} ]}^{\frac{1}{\sigma -1}}{\int}_{\underline{w}}^{\overline{w}} {z^{\frac{\sigma} {\sigma -1}}\{{[1+\beta [1-2H(z)]]}^{\frac{1}{\sigma -1}-1}\}[1-2H(z)]h(z)dz}. $$

Let the median wage be wM. The above is negative if and only if

$$\begin{array}{@{}rcl@{}} &&{\int}_{\underline{w}}^{w_{M}} {z^{\frac{\sigma} {\sigma -1}}\{\left[ 1+\beta [1-2H(z)] \right]^{\frac{1}{\sigma -1}-1}\}[1-2H(z)]h(z)dz}\\ &&{\kern58pt} <{\int}_{w_{M}}^{\overline{w}} {z^{\frac{\sigma} {\sigma -1}}\{{[1+\beta [1-2H(z)]]}^{\frac{1}{\sigma -1}-1}\}[2H(z)-1]h(z)dz} . \end{array} $$

Using symmetry, which implies g(x) ≡ h(wMx) = h(wM + x), and reversing the limits of integration we arrive at (12) which is equivalent to (11):

$$\begin{array}{@{}rcl@{}} {\int}_{0}^{\frac{\overline{w} -\underline{w}}{2}} {{(w_{M}-x)}^{\frac{\sigma} {\sigma -1}}\{{[1+\beta \left[ 1-2H(w_{M}-x) \right]]}^{\frac{1}{\sigma -1}-1}\}[1-2H(w_{M}-x)]g(x)dx}\\ <{\int}_{0}^{\frac{\overline{w} -\underline{w}}{2}} {{(w_{M}+x)}^{\frac{\sigma} {\sigma -1}}\{{[1+\beta [2H(w_{M}-x)-1]]}^{\frac{1}{\sigma -1}-1}\}[1-2H(w_{M}-x)]g(x)dx}.\\ \end{array} $$

To show that (12) holds it is sufficient to show that (13) holds weakly for x = 0 and strictly for all \(x\in \left (0 \right .\frac {\overline {w} -\underline {w}}{2}]\).

$$\begin{array}{@{}rcl@{}} &&{(w_{M}-x)}^{\frac{\sigma} {\sigma -1}}{[1+\beta \left[ 1-2H(w_{M}-x) \right]]}^{\frac{1}{\sigma -1}-1}\\ &&{\kern95pt} \le {(w_{M}+x)}^{\frac{\sigma} {\sigma -1}}{[1+\beta [2H(w_{M}-x)-1]]}^{\frac{1}{\sigma -1}-1}. \end{array} $$

Clearly, (13) holds weakly for x = 0 since H(wM) = 1/2. We now show that (13) holds strictly for \(x\in \left (0 \right .\frac {\overline {w} -\underline {w}}{2}]\), which implies \(\frac {\overline {w} -\underline {w}}{2}\le w_{M}\). So we have 2H(wMx) < 1 and, given that \(\frac {1}{\sigma -1}-1<0\) for σ > 2, it follows that

$$ {[1+\beta [1-2H(w_{M}-x)]]}^{\frac{1}{\sigma -1}-1}<{[1+\beta [2H(w_{M}-x)-1]]}^{\frac{1}{\sigma -1}-1}. $$

Finally, (14) and (wMx) < (wM + x) implies that (12) holds strictly for

$$x\in \left( 0 \right.\frac{\overline{w} -\underline{w}}{2}]. $$

(ii). Follows from part (i) and Proposition 1. □

Proof of the Lemma

Observe that γ(wM,α,β) can be written as

$$\begin{array}{@{}rcl@{}} \gamma (w_{M},\alpha, \beta )&\equiv& \frac{1}{\sigma -1}(\frac{1}{2})[\beta \int\limits_{w_{M}}^{\overline{w}} {zL(z,\tau )} 2h(z)dz-\alpha \int\limits_{\underline{w}}^{w_{M}} {zL(z,\tau )} 2h(z)dz]\\ &+&(\frac{\beta} {2})[{\int}_{w_{M}}^{\overline{w}} {zL(z,\tau )} 2h(z)dz-w_{M}L(w_{M},\tau )]\\ &&+(\frac{\alpha} {2} )[w_{M}L(w_{M},\tau )-{\int}_{\underline{w}}^{w_{M}} {zL(z,\tau )} 2h(z)dz]. \end{array} $$

Average income of those above the median is given by \({\int }_{w_{M}}^{\overline {w}} {zL\left (z,\tau \right )} 2h(z)dz\) and average income of those below the median is given by \({\int }_{\underline {w}}^{w_{M}} {zL(z,\tau )} 2h(z)dz\). For all β > 0 and α ∈ [0,β] the sum in the first bracketed term is positive since average income of those above the median is greater than average income of those below the median. The sum in the second bracketed term is positive because average income of those above the median is greater than the median income. The sum in the third bracketed term is positive because median income is greater than the average income of those below the median. □

Proof of Proposition 5

Let τM denote the most preferred tax for an other-regarding median agent and let τS denote the most preferred tax for a selfish median agent. The condition α = β implies that labor income for the selfish median voter equals that of the inequality averse median voter; namely, wML(wM,τ). Thus, τS must satisfy

$$ -w_{M}L(w_{M},\tau^{S})+{\int}_{\underline{w}}^{\overline{w}} {[zL(z,\tau^{S})+\tau^{S}z\frac{\partial L(z,\tau^{S})}{\partial \tau} ]} h(z)dz= 0. $$

Now if we evaluate the expression in (7) at τ = τS we get

$$-w_{M}L(w_{M},\tau^{S})+{\int}_{\underline{w}}^{\overline{w}} {[zL(z,\tau^{S})+\tau^{S}z\frac{\partial L(z,\tau^{S})}{\partial \tau} ]} h(z)dz+ \gamma (w_{M},\alpha, \beta )>0. $$

Given that τM must satisfy (6) it follows that τM > τS. □

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Murray, M.N., Peng, L. & Santore, R. How does inequality aversion affect inequality and redistribution?. J Econ Inequal 16, 507–525 (2018).

Download citation


  • Income distribution
  • Inequality aversion
  • Redistribution