Skip to main content

Advertisement

Log in

Subjective beliefs about the income distribution and preferences for redistribution

  • Original Paper
  • Published:
Social Choice and Welfare Aims and scope Submit manuscript

An Erratum to this article was published on 29 March 2016

Abstract

We investigate whether beliefs about the income distribution are associated with political positions for or against redistribution. Using a novel elicitation method, we assess individuals’ beliefs about the shape of the income distribution in the United States. We find that respondents’ beliefs approximate the actual distribution on average. However they tend to overestimate the median income and underestimate the level of inequality. Surprisingly we find that beliefs about overall inequality, measured in terms of income dispersion, play only a marginal role in political positions as well as prospects of future wealth. Political preferences, however, are predicted by first, beliefs about the level of income of the poorest members of society, and second, a belief in an open society with equal opportunities for all. Support for redistribution is lower for people who give higher estimates of the income level of the poorest members of society and for people who perceive that opportunities for upward mobility are available.

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

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3

Similar content being viewed by others

Notes

  1. Several models in the literature are equivalent to specific cases of (2) where u is additively separable in own and others income (Fehr and Schmidt 1999; Andreoni and Miller 2002). Such a separability has been axiomatised in the two person case by Karni and Safra (2002).

  2. We do not make claims of exhaustivity here. Our framework is not the only way to represent preferences for redistribution. More general models of preferences over distribution already exist in the one period case (Weymark 1981).

  3. This idea got a lot visibility during the 2008 US presidential campaign when a member of the public, afterward nick named “Joe the plumber”, criticized Obama for his proposed taxes which would hit Joe in the future as he was expecting to see his business grow.

  4. These questions about the average income of typical professions are taken from the International Social Survey Program and allow for the elicitation of beliefs with simple benchmarks (e.g., “What do you think is the average income in the USA of a chairman of a large national corporation?” and “What do you think is the average income in the USA of an unskilled worker in a factory?”).

  5. This selection is however not necessarily an issue when discussing the effect of resondents’ characteristics on their survey answers. Formally, Magee et al. (1998) show that this type of analysis is valid as long as there is no unobservable variable influencing the choice to self-select in the sample which is both correlated with the respondents’ characteristics and with the dependent variables studied (here the beliefs about inequalities and political positions). This is an issue which has been investigated in depth in experimental economics where most samples are self-selected and non representative. Studies such as Von Gaudecker et al. (2008), Harrison et al. (2009), and Cleave et al. (2013) have found that while samples of participants may differ from the population sampled the differences between groups of different demographics are not biased. These studies conclude that there is no selection bias based on social and risk preferences.

  6. A table comparing the answers of the retained and non retained sample is included in “Appendix 1”.

  7. The randomisation of the initial position of the markers on the left or on the right allows us to check whether the answers to the DB are very sensitive to the framing. Over the 839 participants retained, 1.61 % had the markers stacked on the left (proportion not significantly different from 50 %: p \(=\) 0.35). We did not find any significant differences in answers as a function of the markers initial location. A t test of means indicate that both groups gave very similar answers in average (p \(=\) 0.69). A plot of the two corresponding densities, similar to the one from Fig. 2 does not show any difference between the two distributions.

  8. This result is consistent with other studies. The DB estimates tend to be less variable and more accurate Goldstein and Rothschild (2014).

  9. Note that the cues provided by the DB would be limited: The upper end of the DB axis was labelled $205,000 and above, and participants were free to place any number of markers in this bin. Nonetheless, participants could try to read, through the choice of axis lablels, knowledge revealed by the experimenters. We tried to minimize this concern by choosing an upper label slightly above the true 95th percentile, not too far from the real value but unlikely to seem too low or too high. In practice, the tendency of respondents to draw skewed distribution with only few markers on the top category could suggest that they did not feel constrained by the x axis. A possible alternative to our choice of design for future studies would be to adopt as a 95th percentile the level from the answers to the questionnaire. The drawback of this alternative is that if participants do not understand the concept of 95th percentile, it can introduce noise in the elicitation procedure by creating an upper income category which is way off the mark.

  10. We took the data from the Bureau of labor statistics (2010). While our survey questions asked for average income, the BLS only give data for median income. This is likely to give a lower bound for the average income in the profession, in particular for doctors and chairpeople where a negative skew in the distribution is likely to exist. For each category, we took a representative profession listed by the BLS (overall results are not sensitive to the choice of other specific professions. The BLS data gives: (1) unskilled factory worker (food processing workers) \(\$23{,}000\) per year, (2) skilled factory workers (industrial machinery mechanics and maintenance workers) \(\$44{,}000\) per year, (3) doctors (physicians and surgeons) \(\$166{,}400\) per year and (4) chairman of a large national corporation (based on 158 Standard & Poors 500 index companies) \(\$9,000,000\) per year.

  11. In addition, we also tested for possible correlations between more elaborate measures of inequality, such as the Gini coefficient, between the DB distribution and the local income distribution at the zip level or in the county (using both income standard deviation and local Gini coefficients). We did not find any correlation.

  12. Our sample does not contain enough African American or Asian American participants to estimate significant differences between these categories and White Americans in terms of beliefs.

  13. The results are robust to other specifications.

  14. One possible explanation could be that higher income respondents are both less informed about low household incomes (and as a consequence overestimated them) and more conservative. In that case, the correlation between beliefs and political position would just be a spurious link created by the correlation between the political position of the respondent and his/her degree of error made when asked to guess the level of income of the poorest households. We checked for such a possible explanation by running the same regression on the subsample of respondents with an income lower than $50,000 (median of the US distribution of household incomes) and on the subsample of those whose income is higher than $80,000 (75th percentile of the distribution) poorest respondents in our sample. In both samples, beliefs about the income of the poorest had a similarly positive marginal effect on political positions. This indicates that the observed correlation between beliefs about the lowest incomes in society and political positions is not reflecting different errors from respondents.

  15. A possible concern could be that low income respondents are more knowledgeable about the income of low income households than high income respondents who may overestimate the income from the poorest households in society. Such a situation would create the observed correlation if low income respondents tend to be in favor of redistribution and high income respondents tend to be against redistribution. The fact that the coefficient on the belief about unskilled workers’ income does not change between column (5) and column (6) when the income of the respondent is included as a covariate tends to suggest that it is not what is driving the results. The coefficient from the income variable should partially capture the correlation between income and political position in column (6). The link between participants income and political position could however be non linear and be imperfectly captured by the inclusion of the income variable in the regression. We therefore constructed a set of four dummies for the quartiles of the income distribution and we included them in the regression. The results show that the coefficients and their level of significance are almost unchanged. This suggest that for respondents of different income levels, estimates of income of unskilled workers is positively correlated with being against redistribution.

  16. We checked here again that this result could not simply reflect a better information from higher income respondents who tend to be more conservative. The magnitude of the coefficient does not decrease when the regressions are made within samples of richer and poorer respondents.

  17. We also investigated the effect of inequalities at the county level, with the same limited results. If another level of locality is appropriate, we suspect it may be a level closer to the respondent.

  18. The recent “Occupy Wall Street” political movement, rallying under the cry of “We are the 99 %”, used the high level of inequality at the top of the income distribution to mobilize support. Our results suggest that people may be more influenced towards redistribution by focusing on the lower end of the distribution instead.

References

  • Alesina A, Angeletos G (2005) Fairness and redistribution. The American Economic Review 95(4):960–980

    Article  Google Scholar 

  • Alesina A, Giuliano P (2009) Preferences for redistribution. National Bureau of Economic Research

  • Alesina A, La Ferrara E (2005) Preferences for redistribution in the land of opportunities. Journal of Public Economics 89(5–6):897–931

    Article  Google Scholar 

  • Alesina A, Rodrik D (1994) Distributive politics and economic growth. The Quarterly Journal of Economics pp 465–490

  • Andreoni J, Miller J (2002) Giving according to garp: An experimental test of the consistency of preferences for altruism. Econometrica 70(2):737–753

    Article  Google Scholar 

  • Ansolabehere S, Meredith M, Snowberg E (2014) Mecro-economic voting: Local information and micro-perceptions of the macro-economy. Economics & Politics 26(3):380–410

    Article  Google Scholar 

  • Atkinson A (1970) On the measurement of inequality. Journal of economic theory 2(3):244–263

    Article  Google Scholar 

  • Atkinson A, Bourguignon F (1982) The comparison of multi-dimensioned distributions of economic status. The Review of Economic Studies 49(2):183–201

    Article  Google Scholar 

  • Bénabou R, Ok E (2001) Social mobility and the demand for redistribution: the poum hypothesis. The Quarterly Journal of Economics 116(2):447–487

    Article  Google Scholar 

  • Bénabou R, Tirole J (2006) Belief in a just world and redistributive politics. The Quarterly Journal of Economics 121(2):699–746

    Article  Google Scholar 

  • Bolton G, Ockenfels A (2000) Erc: A theory of equity, reciprocity, and competition. American economic review pp 166–193

  • Braas-Garza P (2006) Poverty in dictator games: Awakening solidarity. Journal of Economic Behavior & Organization 60(3):306–320

    Article  Google Scholar 

  • Brunner E, Ross S, Washington E (2011) Economics and policy preferences: Causal evidence of the impact of economic conditions on support for redistribution and other ballot proposals. The Review of Economics and Statistics 93(3):

  • Bureau of labor statistics (2010) Occupational outlook handbook. US Department of Labor, Spring

  • Chakravarty SR (2009) Inequality, polarization and poverty. Springer,

  • Charness G, Rabin M (2002) Understanding social preferences with simple tests. The Quarterly Journal of Economics 117(3):817

    Article  Google Scholar 

  • Cleave BL, Nikiforakis N, Slonim R (2013) Is there selection bias in laboratory experiments? the case of social and risk preferences. Experimental Economics 16(3):372–382

    Article  Google Scholar 

  • Corneo G, Grüner H (2002) Individual preferences for political redistribution. Journal of Public Economics 83(1):83–107

    Article  Google Scholar 

  • Cox J, Sadiraj V (2006) Direct tests of models of social preferences and a new model. Tech. Rep. 976027, SSRN

  • Cruces G, Perez Truglia R, Tetaz M (2013) Biased perceptions of income distribution and preferences for redistribution: Evidence from a survey experiment. Journal of Public Economics 98:100–112

    Article  Google Scholar 

  • Di Tella R, Galiani S, Schargrodsky E (2007) The formation of beliefs: evidence from the allocation of land titles to squatters. The Quarterly Journal of Economics 122(1):209

    Article  Google Scholar 

  • Engelmann D, Strobel M (2004) Inequality aversion, efficiency, and maximin preferences in simple distribution experiments. The American Economic Review 94(4):857–869

    Article  Google Scholar 

  • Engelmann D, Strobel M (2007) Preferences over income distribution, experimental evidence. Public Finance Review 35(2):285–310

    Article  Google Scholar 

  • Fehr E, Schmidt K (1999) A theory of fairness, competition, and cooperation. The Quarterly Journal of Economics 114(3):817–868

    Article  Google Scholar 

  • Fillippin A, Checchi D (2004) An experimental study of the poum hypothesis. Research on economic inequality 11

  • Fong C (2001) Social preferences, self-interest, and the demand for redistribution. Journal of Public Economics 82(2):225–246

    Article  Google Scholar 

  • Fong C, Bowles S, Gintis H (2006) Strong reciprocity and the welfare state. Handbook on the Economics of Giving, Reciprocity and Altruism 2:1439–1464

    Article  Google Scholar 

  • Gigerenzer G (1991) How to make cognitive illusions disappear: Beyond heuristics and biases. European review of social psychology 2(1):83–115

    Article  Google Scholar 

  • Giuliano P, Spilimbergo A (2009) Growing up in a recession: Beliefs and the macroeconomy. National Bureau of Economic Research

  • Goldstein DG, Rothschild D (2014) Lay understanding of probability distributions. Judgment and Decision Making 9(1):1–14

    Google Scholar 

  • Goldstein DG, Johnson E, Sharpe W (2008) Choosing outcomes versus choosing products: Consumer-focused retirement investment advice. Journal of Consumer Research 35(3):440–456

    Article  Google Scholar 

  • Gottschalk P, Spolaore E (2002) On the evaluation of economic mobility. The Review of Economic Studies 69(1):191–208

    Article  Google Scholar 

  • Harrison GW, Lau MI, Elisabet Rutström E (2009) Risk attitudes, randomization to treatment, and self-selection into experiments. Journal of Economic Behavior & Organization 70(3):498–507

    Article  Google Scholar 

  • Harsanyi J (1955) Cardinal welfare, individualistic ethics, and interpersonal comparisons of utility. The Journal of Political Economy pp 309–321

  • Hirschman A, Rothschild M (1973) The changing tolerance for income inequality in the course of economic development. The Quarterly Journal of Economics 87(4):544

    Article  Google Scholar 

  • Isaksson AS, Lindskog A (2009) Preferences for redistributiona country comparison of fairness judgements. Journal of Economic Behavior & Organization 72(3):884–902

    Article  Google Scholar 

  • Karni E, Safra Z (2002) Individual sense of justice: a utility representation. Econometrica 70(1):263–284

    Article  Google Scholar 

  • Kitschelt H (1994) The transformation of European social democracy. Cambridge University Press,

  • Kolm SC (1969) The optimal production of social justice. In: Margolis J, Guitton S (eds) Public economics. Macmillan, London, pp 145–200

    Chapter  Google Scholar 

  • Kuziemko I, Norton MI, Saez E (2015) How elastic are preferences for redistribution? evidence from randomized survey experiments. American Economic Review 105(4):1478–1508

    Article  Google Scholar 

  • Li Y, Johnson EJ, Zaval (2011) Local warming: Daily temperature deviation affects beliefs and concern about climate change. Psychological Science 22(4):454–459

    Article  Google Scholar 

  • Magee L, Robb A, Burbidge J (1998) On the use of sampling weights when estimating regression models with survey data. Journal of Econometrics 84(2):251–271

    Article  Google Scholar 

  • Norton M, Ariely D (2011) Building a better ameerica one wealth quintile at a time. Perspectives on Psychological Science 6(1):9

    Article  Google Scholar 

  • Persson T, Tabellini G (2002) Political economics: explaining economic policy. MIT press,

  • Piketty T (1995) Social mobility and redistributive politics. The Quarterly Journal of Economics 110(3):551

    Article  Google Scholar 

  • Piketty T (1996) Mobilité économique et attitudes politiques face à la redistribution. Tech. Rep. 9603, CEPREMAP

  • Piketty T (2003) Attitudes vis-à-vis des inégalités de revenu en france: existerait-il un consensus. Comprendre 4:209–41

    Google Scholar 

  • Poole K, Rosenthal H (1991) Patterns of congressional voting. American Journal of Political Science pp 228–278

  • Przeworski A, Sprague J (1986) Paper stones: A history of electoral socialism. University of Chicago Press Chicago,

  • Ravallion M, Lokshin M (2000) Who wants to redistribute?: The tunnel effect in 1990s russia. Journal of Public Economics 76(1):87–104

    Article  Google Scholar 

  • Rawls J (1971) A theory of justice. Harvard University, Mass

  • Surowiecki J (2004) The wisdom of crowds: Why the many are smarter than the few and how collective wisdom shapes business. Economies, Societies and Nations

  • Von Gaudecker HM, Van Soest A, Wengström E (2008) Selection and mode effects in risk preference elicitation experiments. Tech. rep, IZA Discussion Paper

  • Weymark J (1981) Generalized gini inequality indices. Mathematical Social Sciences 1(4):409–430

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Lionel Page.

Appendices

Appendix 1: Variables description

See Table 7.

Table 7 Description of the explaining variables

Appendix 2: Description of the sampled counties

See Table 8 and Fig. 4.

Table 8 Detail of the selected counties
Fig. 4
figure 4

Census regions and divisions

Appendix 3: Survey sample

Tables 9 and 10 below compares the survey sampling firms panel demographics to our sample. The largest differences is observed for gender with more female respondents than in the overall panel. We have already experienced such a gender imbalance in previous uses of this firms panel as well as other commercial panels and panels we have curated ourselves. This suggests that it may be due to a general gender differences in the propensity to participate to a survey rather than a selection induced specifically by our topic. We also observe a smaller number of young participants, a larger number of participants with high income. We have now added this information to the description of our sample.

Table 9 Comparison of the study sample with the sampling firm panel demographics

We retained 82 % of the initial sample and eliminated respondent who looked that they may not have taken the task seriously. Table 10 compares the answers of both samples (retained and not retained). On most answers, there are no significant differences. The standard deviation of the distribution elicited by the DB is significantly higher in the retained sample relative to the non retained sample. It is likely to be a mechanical effect of the rule we chose: we eliminated participants who clicked less than 5 times on the DB. This is likely to eliminate participants who created DB distributions with only limited dispersion. The retained sample also display a belief which is (marginally) significantly higher than for the non retained sample. Overall, our choice to retain the sample does not lead to a sample of respondents with very different answers. In particular, there are no significant differences in regard to the answers to the political questions.

Table 10 Comparison of the retained sample with the non retained sample

Appendix 4: Structural estimation

Estimating a parametric distribution from DB data requires taking into account that the DB allows participants to give an estimate of the percentage of households in a series of brackets 10,000 dollars wide each, up to the highest bracket “$205k and above”. To summarize beliefs about the income dispersion in one parameter, we model the DB observations as coming from a lognormal distribution with mean and standard deviation \(\mu \) and \(\sigma \). To estimate these parameters for each participant, we maximise the log-likelihood function for each individual i:

$$\begin{aligned} \ell (\mu _i,\sigma _i|M_i)=\sum _{k=1}^{100}\sum _{j=1}^{22} \ln \left( \varPhi \left( \frac{\ln (\overline{b}_j)-\mu _i}{\sigma _i}\right) -\varPhi \left( \frac{\ln (\underline{b}_j)-\mu _i}{\sigma _i}\right) \right) {\mathbbm {1}}_{\{m_{ki}\in [\underline{b}_j,\overline{b}_j]\}} \end{aligned}$$

where \(M_i\) is the vector of observed marker values \(m_{ki}\) for the individual i, with k being the id of the marker for each participant, j the id of the bracket and \(\underline{b}_j\), \(\overline{b}_j\) respectively the lower and upper bounds of bracket j (with \(\underline{b}_1=0\) and \(\overline{b}_{22}=+\infty \)).

This model allows us to estimate how different participant characteristics correlate with participants’ beliefs about the shape of the distribution. To do so, we sum the individual log likelihood over the whole sample:

$$\begin{aligned} \mathcal {L}(\mu ,\sigma |M)=\sum _{i=1}^N \ell (\mu _i,\sigma _i|M_i) \end{aligned}$$
(5)

where M is the vector of all observed marker values \(m_{ki}\) in the sample. In order to estimate links between individual and local characteristics of participants and their beliefs, we parametrize the coefficients as linear functions of vectors of observed variables \(X_1\) and \(X_2\) respectively:

$$\begin{aligned} \mu&=\beta _1X_1\nonumber \\ \sigma&=\beta _2X_2 \end{aligned}$$
(6)
Table 11 Beliefs about the moments of the distribution from the distribution builder

Table 11 shows the results of the estimations of parameters \(\mu \) and \(\sigma \) from the lognormal distribution by maximizing the likelihood (5). The parameters are written as linear functions of variables characterizing the individual and local characteristics following Eq. (6). To take the non-independence of observations within participants into account, we use a robust matrix of variance clustered by participants. We find that education and age are significant, while no local variable is significant.

Appendix 5: Median comparison

An interesting feature of our design is the elicitation of the subjective beliefs about the median household income using two different methods: a direct question and the Distribution Builder. Figure 5 shows the scatterplot of the individual answers to these two methods. There is clearly substantial variation across the two elicitation methods reflected in the overall correlation of 0.6 between the two types of answers. We think that these difference can be due to the abstract nature of the direct question which requires participants to think about the notion of percentiles. This may lead to more noisy answer from guesses. Another possibility is that people may take more or less care in answering each question. Overall 60 % of participants estimates about the median via these two methods are within $10,000 from each other and 78

Fig. 5
figure 5

Individual answers about the median in the questionnaire and in the DB. p values indicated in brackets

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Page, L., Goldstein, D.G. Subjective beliefs about the income distribution and preferences for redistribution. Soc Choice Welf 47, 25–61 (2016). https://doi.org/10.1007/s00355-015-0945-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00355-015-0945-9

Keywords

Navigation