Income Inequality Among the U.S. States: the Efficacy of Welfare Payment and Spending on Education

doi:10.21203/rs.3.rs-1828146/v1

The fifty American states plus the District of Columbia provide an excellent experimental setting to study income inequality and the efficacy of various programs that attempt to ameliorate income inequality. In this study I use state expenditures on some programs – specifically, education and welfare – to evaluate the extent to which such programs influence income inequality. The conclusion reached here is that while expenditures on education reduce income inequality, public welfare spending at best tends to not affect it and likely worsen it.

Other Economics

Education

Income Inequality

US States

Welfare

The issue of income and wealth inequality has attracted a great deal of attention and concern in recent years. Associated with this concern is how can we ameliorate inequality of income and wealth? The fifty American states plus the District of Columbia provide an excellent experimental setting to study income inequality and the efficacy of various programs to lessen it. In this study, I use state expenditures on some programs – specifically, education and welfare – to evaluate the extent to which such programs influence income inequality. The conclusion reached here is that while expenditures on education reduce income inequality, public welfare spending at best tends to not affect it and likely worsen it.

The fifty US states and the District of Columbia (collectively referred to as “states” allocate about a combined two-thirds of their total budget on education and public welfare; no other spending comes close to these two items. How does this allocation of resources affect income distribution? Numerous studies have found income inequality to be caused by, among other factors, education (Goldin and Katz, 2008; Galbraith, 2016), and a host of other things, such as innovation (Akcigit, et. al., 2019; Battisti, Dustmann and Schonberg, 2019; Aghion et. al. 2019; Tyson, 2021), geographical location of income earners and trade (Caliendo, Dvorkin and Parro 2019; Akerman, et. al., 2013), and migration (Card, 2009; Ottaviano and Peri, 2012; Piyapromdee, 2021). However, the role played by public welfare has not been emphasized, even though some 30% of the total general spending by states is allocated to public welfare. The American states, with their diversity of economic attributes, ideology and programs provide an excellent experimental setting to study income inequality and the efficacy of some major social programs that influence it. In this study, I use state expenditures on education and public welfare to evaluate the extent to which such programs influence income inequality. The conclusion reached here is that while the resources spent on education reduce income inequality, public welfare spending, at least at levels we observe, is not an efficacious use of resources to reduce it.

The annual data cover the period 2013-2019. The Gini Coefficient (also referred to as Gini Index) figures are obtained from the U.S. Census Bureau; these figures measure inequality before taxes and subsidies, using income quintiles. Population is measured by the number of state residents, obtained from the Census Bureau, which is also the source of median income in each state. Data on state expenditures on various programs are those reported by the US Census Bureau’s Annual Survey of State and Local Government Finances.

The mean value of the Gini Index for all states during the period 2013-2019 is 0.465 with a standard deviation of 0.021. The mean change of the Gini Index for 2013-2019 for the states is very close to zero at -6.275x10^-5 with standard deviation of 0.00782; this results in a test-statistic of -0.0080 (the p-value is essentially 0), which easily rejects the null hypothesis of change in the mean value of the Gini Index. So we can be quite confident that the average income distribution as measured by the Gini Index for all states has not changed during the seven years included in this study. That is surprising in view of the heightened interest shown in this issue.

Taking the average of the Gini Index during 2013-2019, the five top most unequal income states are, in decreasing order, DC at 0.528, NY at 0.513, CT at 0.498, LA at 0.494, and CA at 0.489[1]. The five least unequal states are AK at 0.423, UT at 0.426, WY at 0.435, NH at 0.440, and HI at 0.440. The inequality rankings remain remarkably stable from year to year. The range of the Gini Index for the top five the is 0.49 to 0.53, and the range for lowest five is 0.41 to 0.44. Of course, minor changes in ranking do take place among the states; e.g., NY replaces DC as the highest ranking Gini Index in 2019 and ID becomes the second least unequal in 2019, moving from the 7^th least unequal two years earlier. These observations strengthen the conclusion that income inequality among states has hardly changed in the years 2013 through 2019.

Of a total of $2.023 trillion of their general expenditures in 2013, states allocated 66.3% of it to education and public welfare, with 30.7% spent on public welfare. By 2019, total expenditures had risen to $2.505 trillion with 68.5% of it allotted to education and public welfare, of which 34.5% went to public welfare; so welfare payments gained relative to education in those years. The relative decrease in spending on education may in part be due to somewhat lower number of people of all ages enrolled in educational institutions, from nursery and kindergarten through graduate university level. According to the Census Bureau (US Census, Current Population Survey, August 2021) 77.8 million people were enrolled in both public and private educational institutions in 2013; the number had declined to 76.8 million by 2018, and to 76.1 million by 2019. The corresponding numbers for only public school enrollment were 67.4 million in 2013, falling to 66.7 million in 2018 and further declining to 65.9 million in 2019.

These declines in school enrollment are probably due to the ageing of the U.S. population. Across all states, spending on education amounted from a low of 25.0% of the general expenditures to a high of 45.4% in 2013, with 2019 not much different, from a low of 22.9% to a high of 45.5%.

Although the fraction of general expenditures allotted to education and public welfare varies among the states over time, as percentage of total state expenditures ME, TN, NY and CA tend to spend the most on public welfare relative to other states and WY and ND the least. When it comes to education, UT tends to be the state that spends the most and MA the state that spends the least. There was a sharp reduction of spending on education in MA in 2014, from around 25% in the preceding and subsequent years; even excluding the 2014 figure, MA ranks among the states spending the least on education. However, these figures should not be interpreted to imply that, say, citizens of MA are less dedicated to educating their young than citizens of UT; it may well be the case that MA has fewer students per capita than UT, or that MA relies more heavily on not-state-supported schools for educating its children. By the same token, that CA and NY spend more on public welfare than most other states probably a reflects the higher cost of living in those states and the greater number of immigrants compared to other states[2]; a great proportion of new immigrants tend to earn less than the native workers (Card, 2009; Ottaviano and Perri, 2012).

For all the states, per capita spending on education rose by an annual average of 2.72% while per capita welfare payments increased at a rate of 5.07% per year during the period 2013-2019. Inflation amounted to slightly less than 1.6% per year in those years, so while both spending on education and welfare beat inflation, welfare payments rose by a much more generous amount.

[1] The list of states along with their abbreviations are presented after the References section.

[2] Admittedly, these explanations may not apply to states such as ME and TN that have lower immigrants and cost of living.

While the role of education, with education broadly defined to include various types of training, on income inequality has been studied extensively (Galbraith, 2016; Battisti, Dustmann and Schonberg, 2019; Ganon and Shaog, 2017), the efficacy of public welfare payment – in contrast to all social welfare programs – on income distribution has not attracted as much attention. The general finding by other researchers is that education, by narrowing the skill gap and increasing the mobility of workers, also narrows income inequality.

As long ago as 1955, Simon Kuznets (Kuznets, 1955) in his presidential address to the American Economic Association discusses the influence of national income, per capita income and population on income inequality. Unfortunately, Kuznets is not very clear which of the these variables he considers most important, although it seems that population is of primary concern. Furthermore, it is not clear whether it is the level or growth of these variables that he considers most important. In any case, given the very different sizes of the states, and in view of Kuznets observation, it is useful to consider the impact some measure of income may have on income inequality. Jason Furman argues that in addition to social safety net, increased “market incomes will likely be necessary as well” to narrow the income inequality gap (Furman, 2021, p. 242). Higher income increases labor mobility, both across jobs as well as geography, and makes workers better informed about the labor markets, thus helping them gain higher incomes.

Perhaps the best measure for our purpose is median income; median income better represents the distribution of income than GDP per capita, especially if income is more heavily weighted toward lower income levels. Median income among states increased by an average of 23.2% in the seven-year period 2013–2019, with a range of growth of 4.5–36.5% during that period. The highest growth took place in DC where median income rose to more than $92,400 by 2019 and AK experienced the lowest growth where median income increased to just about $75,000; the 2019 median income in DC is slightly more than twice the level in MS at $45,800. The three Pacific coast states are ranked the highest median income states after DC.

There is substantial literature suggesting that minorities earn less than their majority counterparts (Daly et. al., 2017; Fortin and Lemieux, 1997). Research suggests that a larger proportion of non-whites may be recent immigrants, with lower ability to earn higher income for various reasons, such as inadequate language and work-related skills and deficient information about wages and job markets. Discrimination and simply taking advantage of some minority groups of workers are among these explanations. While it is better to distinguish among minorities since some groups may not be disadvantaged as much as others, that line is not pursued here since it is beyond the scope of this work. Of course, there is also the issue of the male-female earnings gap (Bertrand, 2018) which likely affects the Gini Coefficient.

The number of minority workers is proxied here by the ratio of nonwhites to total population. The white population used here excludes Hispanic whites as well as mixed races; this measure is obtained from the US Census Bureau. Of course, this results in a narrow definition of the white population. Another problem with the concept of minority used here is that the relevant population is the earning population, not the entire population used here.

In Table 1, below are presented a series of regressions, with the Gini Index as the regressand. Both OLS and TSLS are presented in the table.

Table 1. Determinants of the Gini Index

Regressor	Eq. 1: OLS	Eq. 2: TSLS	Eq. 3: OLS	Eq. 4: TSLS	E4. 5:TSLS
Constant	0.483 (0.007)	0.486 (0.008)	0.835 (0.055)	0.826 (0.050)	0.663 (0.060)
W/P	0.0115 (0.0018)	0.0056* (0.0049)
Ed/P	-0.0120 (0.0015)	-0.0096 (0.0021)
Med	-5.17 x10^-7 (9.63 x10^-8)	-4.62 x10^-7 (9.15 x10^-8)
Nw/P	0.0500 (0.0065)	0.0505 (0.0067)
ln(W/P)			0.0237 (0.0033)	0.0196 (0.0071)	0.0117 (0.0079)
ln(Ed/P)			-0.0256 (0.0033)	-0.0237 (0.0039)	-0.0254 (0.0046)
ln(Med)			-0.0314 (0.0051)	-0.0305 (0.0046)	-0.0170 (0.0170)
ln(Nw/P)			0.0161 (0.0013)	0.0160 (0.0014)
IV		Rev/P		Rev/P	Rev/P
R²	0.38		0.41
J-stat		-2.49 x10^-38		0.000	0.000

Standard errors are shown in parentheses.

*Excepting this coefficient, all coefficients are significant at better than 5% level.

W= Welfare Payment; P = Population; Ed = Spending on Education; Med =Median Income;

Nw = Non-White Population; Rev = Total State Revenue; IV= Instrumental Variable

Equations 2 an 4 are obtained using two-stage least squares (TSLS) because of the likelihood that Gini causes as well as is caused by welfare payment W. This can occur if in developing a budget, the state takes into account income inequality and allocates welfare funds in response to inequality. Accordingly, an instrument needs to be used that is itself independent of Gini. Total state revenue, Rev, is selected as instrumental variable because of its high correlation of 0.982 with welfare payment, W (the correlation between ln(W/P) and ln(Rev/P) is 0.61; Rev is relevant since W is a component of Rev, and as Rev rises W becomes more affordable. Considered too as instrumental variable was total state expenditures. However, Rev is preferable as an instrument since expenditures is likely caused by Gini since the state budget (in contrast with state revenue) may respond to income inequality and attempt to mitigate inequality by higher welfare payments. Regressions of both the variables and on their logarithms are used because of the possibility of nonlinear relationship. It turns out that when the log-linear model is used, the OLS and TSLS of Eq. 3 and Eq. 4 results differ very little.

The variable Nw/P does have a sizeable impact on the coefficient of ln(W/P); when dropped from the TSLS model, the coefficient of ln(W/P) falls by 40% from 0.0196 (t-ratio 2.76) in Eq. 4 to 0.0117 (t-ratio 1.48) in Eq. 5. The impact of the variable ratio of the non-white population to total population (Nw/P) is to reduce Gini, suggesting that reducing the earnings gap between whites and nonwhites is likely to reduce inequality.

The focus in this study is on the effects of welfare payment and education on income inequality. The results obtained here strongly suggest that welfare payments do not help lessen inequality, but education reduces income inequality. To decrease income inequality, the recent policies that have favored welfare payment over education need to be reconsidered. Research is proceeding to better understand the importance of frictions and inequalities in employment opportunities and how to eliminate them; results obtained here may then be modified. It is worth noting that in line with findings by others median income does help reduce income inequality, and policies that raise income are helpful. Finally, programs to improve the earnings ability of the minority groiups are also helpful in shrinking income inequality.

^aConflict of Interest Statement

This is to acknowledge that author did not received financial or any other type of support from anyone or any institution in conducting this research or the preparation of this paper. The author is solely responsible for the all the work and the findings of this research. The author has no interest of any kind in or obligations about this research or any its findings.

^bData Availability and Sharing

All of the data used to conduct this research are obtained from publicly available sources from the various databanks of the U.S. Census Bureau. None of the data are propriety. Furthermore, I would be willing to share my data and my worksheets, as well as the statistical analyses and results upon request.

Acemoglu, Daron, and David Autor. 2011. “Skills. Tasks and Technologies: Implications for Employment and Earnings.” In Orlie Ashenfelter and David Car (eds.) Handbook of Labor Economics, Vol 4 Part B:1043-1171

Aghion, Philippe, Ufuk Akcigit, Antoni Bergeaud, Richard Blundell, and David Hemous. 2015. “Innovation and Top Income Inequality.: NBER Working Paper 21247, Cambridge, MA: National Bureau of Economic Research.

Akcigit, Ufuk, Antonin Bergeaud, Richard Blundell, and David Hemous. 2019. “Innovation and Top Income Inequality.” Review of Economic Studies 86(1): 1-45.

Akerman, Anders, Elhanan Helpman, Oleg Itskhoki, Marc-Andreas Muendler and Stephen Redding. 2013. “Sources of Wage Inequality”, American Economic Review, Papers & Proceedings 103(3), 214-19.

Battisti, Michele, Christian Dustmann, and Uta Schonberg, 2019, “Technological and Organizational Change and the Careers of Workers”, University College, London.

Bertrand, Marianne.2018. Coase Lecture – the Glass Ceiling.” Economica 85(338);205-231.

Blanchard, Olivier, and Dani Rodrick (eds). 2021. Combating Inequality: Rethinking Government’s Role. Cambridge, MA: The MIT Press.

Caliendo, Lorenzo, Maximiliano Dvorkin, and Fernando Parro. 2019. “Trade and Labor Market Dynamics: General Equilibrium Analysis of the China Trade Shock.” Econometrica 87(3): 741-835.

Card, David. 2001. “The Effect of Unions on Wages Inequality in the US Labor Market.” Industrial and Labor Relations Review 54(2): 296-315.

________. 2009. “Immigration and Inequality.” American Economic Review: Papers & Proceedings 99 (2): 1–21. http://www.aeaweb.org/articles.php?doi=10.1257/aer.99.2.1

Daly, Mary C., Bart Hobjin, and Joseph H. Pedtke. 2017. “Disappointing Facts about Black-White Wage Gap.” FRBSF Economic Letters 2017-26, Federal Reserve of San Francisco, September 5, 2017.

European Commission. 2021. Migrant Integration Statistics: 2020. Luxembourg: Publications Office of the European Union.

Fortin, Nicole M., and Thomas Lemieux. 1997. “Institutional Changes and Rising Wage Inequality: Is There a Linkage?” Journal of Economic Perspectives 11(2), 75-96.

Furman, Jason. 2021, “Enhancing Micro and Macro Resilience by Building on the Improvements in the Social Safety Net”. In O. Blanchard and D. Rodrik. Combating Inequality: Rethinking Government’s Role. Cambridge, MA: MIT Press.

Ganon, Peter and Daniel Shaog, 2017. “Why Has Regiona Income Convergence in the US Decline?”, Journal of Urban Economics, 102:76-90

Galbraith, James K. 2016. Inequality: What Everyone Needs to Know. Oxford, England: Oxford University Press

Goldin, Claudia, and Lawrence Katz. 2008. The Race between Education and Technology. Cambridge. MA: Belknap Press/Harvard University Press.

Kuznets, Simon. 1955. “Economic Growth and Income Inequality.” Presidential Address to the American Economic Association, American Economic Review 45(1): 1-28.

Molloy, Raven, Christopher L. Smith, Ricardo Trezzi, and Abigail Wozniak. 2016. “Understanding Declining Fluidity in the U.S. Labor Market.” Brookings Papers on Economic Activity 47(1): 183-259.

Ottaviano, Gianmarco, and Giovanni Peri. 2012. “Rethinking the Effects of Immigration and Wages.” Journal of the European Economic Association 10(1): 152-197.

Piketty, Thomas, and Emmanuel Saez. 2003. “Income Inequality in the United States, 1913-1998.” Quarterly Journal of Economics 118(1): 1-41; update data (1917-2017) available at https://eml.berkeley.edu?~saez/TabFig2018prel.xls

Piyapromdee, Suphanit. 2021. “The Impact of Immigration on Wages, Internal Migration, and Welfare.” The Review of Economic Studies 88(1), 406-453. https://doi.org/10.1093/restud/rdaa029

Tyson, Laura D’Andrea. 2021. “Technological Change, Income Inequality, and Good Jobs.” In O. Blanchard and D. Rodrik. Combating Inequality: Rethinking Government’s Role. Cambridge, MA: MIT Press.

U.S. Census Bureau.2021. Current Population Survey, August (www.census.gov/programs-surveys/cps.html, August 2021)

AL	Alabama
AK	Alaska
AZ	Arizona
AR	Arkansas
CA	California
CO	Colorado
AT	Connecticut
DE	Delaware
DC	District of Columbia
FL	Florida
GA	Georgia
HI	Hawaii
ID	Idaho
IL	Illinois
IN	Indiana
IA	Iowa
KS	Kansas
KY	Kentucky
LA	Louisiana
ME	Maine
MD	Maryland
MA	Massachusetts
MI	Michigan
MN	Minnesota
MS	Mississippi
MO	Missouri
MT	Montana
NE	Nebraska
NV	Nevada
NH	New Hampshire
NJ	New Jersey
NM	New Mexico
NY	New York
NC	North Carolina
ND	North Dakota

OH	Ohio
OK	Oklahoma
OR	Oregon
PA	Pennsylvania
RI	Rhode Island
SC	South Carolina
SD	South Dakota
TN	Tennessee
TX	Texas
UT	Utah
VT	Vermont
VA	Virginia
WA	Washington
WV	West Virginia
WI	Wisconsin
WY	Wyoming

Income Inequality Among the U.S. States: the Efficacy of Welfare Payment and Spending on Education

Status:

Version 1

Abstract

Figures

Introduction

Data And General Observations

Income Inequality: Education and Welfare Payments

Inequality And Income

Minorities And Income Inequality

Statistical Results

Conclusions

Declarations

References

List of states and their abbreviations

Status:

Version 1