SlideShare a Scribd company logo
The State-Level Effects of the Real
Minimum Wage on Income Distribution as
Measured by the Gini Index.
Author: Daniel Brinkerhoff, Spring 2016
1
Daniel Brinkerhoff, Spring 2016
Abstract
The goal of many proponents of increasing the minimum wage is to decrease income inequality.
This paper uses a multiple regression on the most commonly used determinants of income
inequality as well as the real minimum wage and the real minimum wage squared to attempt to
establish a link between an increasing minimum wage and a decreasing income inequality. This
paper finds that an increase in the real minimum wage does tend to decrease income inequality
for values below the turnaround point. In 2015 dollars the turnaround point was found to be
$15.10. However, due to lack of data at, or around this value, this is only a rough approximation.
If accurate, it would mean that increasing the wage to $15 an hour, like some organizations are
fighting for, would cause a decrease in income inequality.
Introduction
The minimum wage is an important enough topic to be a part of the 2016 presidential
election. Each candidate has been asked what their stances are, and movements like
fightfor15.org/ encourage people to vote on the basis of that answer. Fast-Food workers all over
the US have been going on strikes for a ‘living-wage.’ There are arguments claiming the low
income caused by the low minimum wage is putting more people into situations where they
require help from welfare programs; which acts as a pseudo subsidy for industries employing
workers for minimum wage. Yet there are many who think that an increase in minimum wage
would carry heavy costs in the form of decreasing the number of available jobs, especially for
low skill, low experience workers who need it most.
The decision to raise the minimum wage has continued to be a controversial topic. A
topic where empirical analysis seems to clash with basic economic theory. This topic resurfaces
periodically because the minimum wage is a nominal value. Every year it is not increased, is a
2
Daniel Brinkerhoff, Spring 2016
year where real minimum wage decreases by the inflation rate. This means that the answer to
the question ‘What are the costs and benefits of increasing the minimum wage?’ Remains
relevant year after year.
Unfortunately, the costs and the benefits remain uncertain. With basic theory predicting
a significant decrease to employment, but Card and Krueger1
failed to find an employment
decrease in their empirical study of New Jersey’s minimum wage increase. Most people think
that an increase to the minimum wage should increase the welfare of low income households,
but Neumark and Wascher2
found that it “more closely resemble income redistribution among
low-income families than income redistribution from high- to low-income families.” (Neumark
and Wascher, 2002. P333). In light of these contradictions, and the continued need to adjust the
minimum wage, more research on the subject is needed.
Literature Review
There is a lot of research going into minimum wages and their possible effects. As of
right now, there does not seem to be a definitive answer to the question “What are the effects
of Minimum wage increases?” Basic Supply and Demand theory would say that the obvious
outcome is an increase of income for some workers, a loss of income for firms, and the loss of
some jobs, with an overall total surplus decrease.
Draca et al3
looked at the introduction of the 1999 national minimum wage in the UK.
The authors used a difference in differences approach on private firms public accounting
information from the FAME database to study how the 1999 national minimum wage affected
wages and firm profitability. The authors found that “wages were significantly raised, and firm
1
CardandKrueger. 1994. “Minimum Wages and Employment: A Case Study of the Fast-FoodIndustryin NewJersey andPennsylvania.” The
American Economic Review, Vol. 84,No. 4.pp.772-793.
2
Neumark, David, andWascher,William. "DO MINIMUMWAGESFIGHT POVERTY."Economic Inquiry 40, no. 3 (July 2002):315-333.
Academic Search Premier, EBSCOhost (accessedSeptember 29, 2015)
3
Draca, Mirko, StephenMachin, andJohn Van Reenen. 2011."Minimum Wages andFirm Profitability." American Economic Journal:Applied
Economics,3(1): 129-51.
3
Daniel Brinkerhoff, Spring 2016
profitability was significantly reduced by the minimum wage” (Draca et al, 2011. P130). They
also found that firms with relatively high market power lost more profits but maintained their
employment and productivity over other firms. Their findings fit with the expectations of the
basic Supply and Demand theory that predicts a drop in producer surplus.
Neumark and Wascher4
calculated the change in the probability of a family coming out of
poverty vs the probability of a family falling into poverty due to an increase to the minimum
wage. The idea being that the families below the poverty line would get an income boost from
an increased minimum wage which could cause them to get above the poverty line while some
families suffer from the disemployment effect and could fall below the poverty line. The authors
found that there was an increased chance for families to get above the poverty line after a
minimum wage increase, but also found a similar increase to the number of families who would
fall below the poverty line. The net effect being an approximately equal change which causes a
minimum wage increase to “more closely resemble income redistribution among low-income
families than income redistribution from high- to low-income families.” (Neumark and Wascher,
2002. P333)
Card and Krueger5
did an empirical study and found something very different. They took
survey data of 410 fast-food restaurants in New Jersey and Pennsylvania before and after the
1992 increase to minimum wage in New Jersey. The authors found no evidence of lost
employment due to the increased minimum wage. This is not predicted by the simple Supply
and Demand model and implies that there is more happening than can be explained by the
simple model. So a different one will be needed.
Freeman’s6
Redistribution Theory tries to explain where the costs of a minimum wage
increase would come from. The first group who bear the costs of a minimum wage increase are
4
Neumark, David, and Wascher,William. "DO MINIMUMWAGESFIGHT POVERTY."Economic Inquiry 40, no. 3 (July 2002):315-333.
Academic Search Premier, EBSCOhost (accessedSeptember 29, 2015)
5
CardandKrueger. 1994. “Minimum Wages andEmployment: A Case Study of the Fast-FoodIndustryin NewJersey andPennsylvania.” The
American Economic Review, Vol. 84,No. 4.pp.772-793.
6
Freeman, Richard. 1996. "The MinimumWage as a Redistributive Tool."TheEconomic Journal 106(May): 639-649.
4
Daniel Brinkerhoff, Spring 2016
the consumers of products and services generated by minimum wage in the form of higher
prices for those goods and services. The simple assumption is that most people consume the
products of minimum wage workers regardless of personal income. Even if this assumption is
not strictly true, some of the costs of the minimum wage are borne by the consumers, which is
significant to income distribution. The second group that will pay for a minimum wage increase
will be the firms/stakeholders of companies that employ minimum wage workers. Because of the
labor is one of the costs, an increase in costs will be partially borne by the firm. This is
supported by Draca et al’s findings. The final group that would bear the costs of the minimum
wage increase are those who will lose their jobs due to the increase in pay. This is supported by
Neumark and Wascher but was not found to be an issue by Card and Krueger. A possible
solution to the contradictory findings on the minimum wage's effects on employment could be
that it is a non-linear relationship. Lee7
utilizes a quadratic minimum wage to account for the
nonlinear properties of an increasing minimum wage. The idea being that an increase to the
minimum wage is significantly more effective when it is binding (above natural equilibrium) in a
sector than if, after the increase, it remains non-binding (below the natural equilibrium). The
higher above the equilibrium price the minimum wage is, the more of a disemployment effect it
will have.
Trying to analyze and understand income inequality is far from a new idea. As such,
others have tried to describe it before. Hae-Young Lee et al8
attempts to describe income
inequality in Korea. Many of the variables they use will be applicable to the US, but they do not
add minimum wage into their considerations which would be important to the US explanation.
Bosch9
studies the effects of the real minimum wage on the increase in inequality observed in
7
Lee, D. S. (1999). Wage Inequality in the United States during the 1980s: Rising Dispersion or Falling Minimum Wage?. Quarterly
Journal Of Economics, 114(3), 977-1023.
8
Hae-YoungLee, JongsungKim, BeomCheol Cin. “Empirical Analysis on the Determinants ofIncome Inequalityin Korea.” International
Journal of AdvancedScience andTechnology, Vol. 53,April,2013.
9
Bosch, Mariano, andMarcoManacorda.2010. "Minimum Wages andEarnings Inequalityin UrbanMexico." AmericanEconomicJournal:
Applied Economics,2(4): 128-149.
5
Daniel Brinkerhoff, Spring 2016
urban Mexico, he finds that the real minimum wage decrease was the cause of much of the
large increase to income inequality. His findings should be very similar to mine because of the
similarities in the question. However, his findings are for a different place, at a different time. He
also does not attempt to understand the disemployment effect of the minimum wage, which will
be captured by the quadratic minimum wage.
Method
In order to test the relationship between the real minimum wage and income inequality,
this paper will use an Ordinary Least Square Regression on the following model:
𝐺𝑖𝑛𝑖 𝑠𝑡 = 𝛽0 + 𝛽1 𝑅𝑀𝑊𝑠𝑡 + 𝛽2 𝑅𝑀𝑊𝑠𝑡
2
+ 𝛽3 𝐼𝑛𝑓𝑡 + 𝛽4 𝛥𝐺𝐷𝑃𝑝𝐶 𝑠𝑡 + 𝛽5 𝑈𝑅 𝑠𝑡 + 𝛽6%𝑃𝑜𝑝65 𝑠𝑡
+ 𝛽7 𝑀𝑊𝑇𝑠𝑡 + 𝛽8 𝑀𝐿𝐺𝑇𝑠𝑡 + 𝛽9%𝐿𝐻𝑆 𝑠𝑡 + 𝛽10%𝐻𝑆𝐸 𝑠𝑡 + 𝛽11 %𝐶𝐴𝑆𝑠𝑡 + 𝛽12 %𝐵𝑜𝑟𝐻𝑠𝑡
+ 𝛽13 𝐼𝑇𝑠 + 𝛽14 𝑌𝑒𝑎𝑟 + 𝛽15 𝑆𝑡𝑎𝑡𝑒 + 𝜀
See Table 1 for variable names and descriptions
The hypothesis test is H0: 𝛽1= 0
Ha: 𝛽1< 0
And
H0: 𝛽2= 0
Ha: 𝛽2> 0
This model attempts to explain the effects the real minimum wage has on the gini index
while holding all other variables constant. The gini index is a measure of income inequality,
ranging from 0 to 1. A score of 0 indicates perfect equality of income; each person makes the
6
Daniel Brinkerhoff, Spring 2016
same income as every other person. A score of 1 indicates perfect inequality; only one person
makes all of the income (U.S. Census Bureau, Definitions). The definition of income used to
measure the gini index is defined as wages received on a regular basis before taxes. This figure
does not reflect the fact that some income can be received in non-cash benefits like food
stamps, health benefits, access to business transportation or facilities, and retirement plans.
(U.S. Census Bureau, About Income). This paper asserts that gini index is a product of 7
exogenous variables.
The first exogenous variable, and the focus variable of this paper, is the real minimum
wage. This paper predicts that as you increase the real minimum wage (to a point), the gini
index will decrease, yielding a decrease to inequality. This happens because the lower income
earners are more greatly boosted by an increase to real minimum wage than any other group.
Causing the overall distribution to become more centric. However, as the bindingness (amount
above the equilibrium price) increases, the disemployment effect increases more rapidly than
the benefits of the increase to wages. As the disemployment effect increases, fewer people gain
benefits from the increased minimum wage. Also, firms that had monopsony power may actually
increase employment for small increases to minimum wage, but still decrease employment at a
high enough minimum wage levels. Lee10
utilized a quadratic form with a negative minimum
wage, and a positive minimum wage squared, to capture this effect.
Second, Blinder11
found that inflation acted as a kind of ‘progressive tax’ where the lower
class and middle class tend to lose less than the upper class. This would have a negative effect
on the gini index and decrease inequality.
10
Lee, D. S. (1999). Wage Inequality in the United States during the 1980s: Rising Dispersion or Falling Minimum Wage?.
Quarterly Journal Of Economics, 114(3), 977-1023.
11
Blinder, A. S., & Esaki, H. Y. (1978). Macroeconomic Activity and Income Distribution in the Postw ar United States. Review Of
Economics And Statistics, 6(4), 604-609.
7
Daniel Brinkerhoff, Spring 2016
Third, the Kuznets curve12
indicates that as a developed country like the US grows,
income inequality would tend to decrease; this would have a negative effect on the gini index.
This paper uses the change in real income per capita to capture the state level growth.
Forth, Blinder11
also found that the unemployment rate disproportionately decreases the
income of the lowest 40% of income earners. This unequal effect of an increase to
unemployment causes the left hand tail of the gini index to grow, increasing inequality.
Unemployment is not an exogenous variable, however. This means it will likely have an
endogeneity problem. Endogeneity can cause the variable to have reverse causality and will
likely bias the coefficient and increase the associated p-value. This will not cause problems for
any of the other coefficients or disrupt the overall regressions, but that the findings on this
variable cannot be used as direct evidence of unemployment's’ effect on the gini index.
Fifth, Rubin13
has found that the elderly population between 1967 and 1997 have a
significantly higher gini index (more inequality) than the younger populations. The disparity was
shrinking during his research, indicating a change in the working habits of the elderly. This
means the value may have changed for the years between 2006-2014. The value has been
added in for completeness.
Sixth, taxes on the wealthiest income earners tends to act as a ‘braking system’
according to Alvaredo et al14
. The effect is that it lowered the return to effort on negotiating for
higher pay. Because the higher income earners tend to have more leverage in negotiations,
they tend to capture more the company's wealth at the cost of other employees. This means
that as taxes on top income earners increases, the gini index should decrease, indicating a
decrease to inequality. The measures being used to capture this effect are the State level max
wage tax. In addition, this paper adds the state level max long gains tax that would be applied to
12
S. Kuznets, “Economic Grow th and Income Inequality”, American Economic Review , vol. 45, (1955), pp. 1-28.
13
Rubin, Rose M, White-Means, Shelley I. And Daniel, Luojia Mao. “Income distribution of older Americans.” (2000). Bureau of
Labor Statistics.
14
Alvaredo, Facundo,Atkinson, Anthony B, Piketty, Thomas,andSaez, Emmanuel. (2013).”The Top 1 Percent in International andHistorical
Perspective.” Journal ofEconomic Perspectives, Volume 27, Number 3,Summer 2013, Pages 3–20.
8
Daniel Brinkerhoff, Spring 2016
capital investments. The State level max wage tax should directly measure the effect that
Alvaredo et al describes while the max long gains tax should be in opposition to Alvaredo et al’s
findings as it increases the importance of earned income by decreasing the wealth gained by
investing. An important reminder, the income measured by the gini index does NOT include the
income gained by long gains, and it applies before taxes. This means that it will not be directly
affected by those taxes, only indirectly in the ways described by Alvaredo et al.
Seventh, because wages are a representation of a worker's productivity, as productivity
increases, so should wage. If access to education is increased, so too will productivity and wage
on average. This paper uses the percent of the population who have less than a high school
education, high school education, some college or associate's degree, and Bachelor's or higher,
as a measure for the accessibility of schooling.
This paper also uses a fixed effect for the year to capture changes on a federal level that
would not be specifically captured by state level data. This paper also uses a fixed effects for
each state to capture policy and cultural differences that may not be represented by the
exogenous variables. One state level policy (whether or not a state has an income tax) has
been moved out of the state fixed effect and added as a dummy variable for better precision.
Analysis
This initial model utilizes 50 states over the course of 9 years. This generates 450 points
of data for each of the 12 continuous variables, including the endogenous variable. This paper
also uses a total of 58 dummy variables to represent the 50 states, the 9 years, and whether or
not a state has an income tax; note that Alabama and 2006 are used as the reference variables.
The initial model has 379 (450 - 58 - 12 -1) degrees of freedom.
The final model only removes one variable (Percent of population 65 years and older)
due to a low Test Statistic of -.44 and with an associated P-Value of .659. This increases the
9
Daniel Brinkerhoff, Spring 2016
degrees of freedom of the final model to 380. This final model has an adjusted R-Squared of
92.72% which is not uncommon in models that use a large number of fixed effects and dummy
variables.
See Regression Outputs in the appendix.
Problemed Variables
The proportion of the population above 65 has changed from the model prediction of
positively affecting the gini index, to negatively affecting the gini index. However, with a very
high P-Value, this number is not significantly different from zero. Rubin15
’s findings showed that
this difference was gradually decreasing between 1967-1997, and it is possible that there is no
longer a gap between the 65 year and over population, and those younger. For these reasons, it
has been dropped from the final model.
Inflation’s coefficient has a negative sign instead of a the positive sign that was
predicted. However, inflation was also the only variable that used year only level data and was
undifferentiated for the states. The measure used for inflation has a very low P-Value, indicating
that it is a good predictor, but due to the lack of specific data, the coefficient should be analyzed
critically.
The change in Real GDP per capita is strangely a positive figure when the Kuznets
curve would predict a negative coefficient. This value is also significant to the 5% level.
However, this wrong sign was also found in Hae-Young Lee et al16
’s research on the
determinants of income inequality in South Korea. A possible explanation could be that due to
the recession of 2007-2010 and the atypical monetary and fiscal response, the measures used
15
Rubin, Rose M, White-Means, Shelley I. And Daniel, Luojia Mao. “Income distribution of older Americans.” (2000). Bureau of
Labor Statistics.
16
Hae-YoungLee, JongsungKim, BeomCheol Cin. “Empirical Analysis on the Determinants ofIncome Inequalityin Korea.” International
Journal of AdvancedScience andTechnology, Vol. 53,April,2013.
10
Daniel Brinkerhoff, Spring 2016
were not typical and if the timeframe where to be increased, a different coefficient could possibly
be realized.
Unemployment is not statistically significant to the 10% level. This is likely due to the fact
that unemployment is an endogenous variable being used as an explanatory variable. This can
commonly give higher P-values than would be typically found if the determinants of
unemployment were used instead. For the purposes of this paper, the unemployment rate
controls for some of these variables, and the high P-value is not detrimental.
Only one of the measures used to define education was found to be statistically
significant to the 10% level. Because the values are linked (a decrease in one, necessitates the
increase of another) they are all kept in the final model. This does, however, mean that there is
likely a better measure for education than the percent of the population that has attained a
specific level of education.
Gini Index data set
The gini index between 2006-2014 increased in almost all states. Hawaii being the only
state that saw a decrease over that time period. While over the course of the entire range, each
state saw an increase, each year the gini index only increased approximately 65% of the time.
State Fixed Effect
The coefficient of the state fixed effect had a very low average with a comparatively high
standard deviation. This indicates that there was a very large variation, both positive and
negative, in the state coefficients. However, the sum of the coefficients was very close to zero.
The 5 states that had the largest positive coefficient, which indicates larger amounts of income
inequality, were Florida, Tennessee, Texas, new York, and Connecticut. These states tend to
be on or close to the East Coast, they have a relatively higher cost of living, and relatively higher
11
Daniel Brinkerhoff, Spring 2016
per capita GDP. While the 5 states with the largest negative coefficients, which indicates lower
amounts of income inequality, were Idaho, Iowa, Indiana, Nebraska, and Utah. These states
tend to be in the middle of the country without access to the ocean, they have a relatively lower
cost of living, and relatively lower GDP per capita.
The state fixed effect generally had a very low P-Value. This can mean a variety of
things. It is possible that there are determinants that are missing from the main model, that are
being captured by the state fixed effects. There could be state specific policies or culture that
reduce/add income inequality within its borders. Income inequality may also be strongly tied to
past income inequality, which would mean that states tend to cling to former values and only
adjust slowly to new values.
Minimum Wage’s Effect on Income Inequality
The Real Minimum Wage was found to be statistically significant at the 10%. This allows
us to reject the null hypotheses in favor of the alternative. The Real Minimum Wage decreases
income inequality. However, the Real Minimum Wage Squared was not found to be statistically
significant at the 10% level. We fail to reject the null hypothesis.
The real minimum wage and the real minimum wage squared did have the predicted
signs with the real minimum wage being negative and the real minimum wage squared being
positive. They have a turnaround point at a real wage of $13.20 (in 2007 dollars) where further
increases to the minimum wage would yield positive effects on income inequality. A potential
reason for the high P-Value on the minimum wage squared could be due to a lack of data at
higher levels of real minimum wage. Using a different data set where the real wage saw more
values at the higher levels may yield a lower P-Value.
12
Daniel Brinkerhoff, Spring 2016
Citations and References
Articles
Aaronson, Daniel, Sumit Agarwal,and Eric French. 2012. "The Spending and Debt Response to
Minimum Wage Hikes." American Economic Review,102(7): 3111-39.
Alvaredo, Facundo, Atkinson, Anthony B, Piketty, Thomas, and Saez, Emmanuel. (2013).”The Top 1
Percent in International and Historical Perspective.” Journalof Economic Perspectives,Volume
27, Number 3, Summer 2013, Pages 3–20.
Blinder, A. S., & Esaki, H. Y. (1978). Macroeconomic Activity and Income Distribution in the Postwar
United States. Review Of Economics And Statistics, 6(4), 604-609.
Bosch, Mariano, and Marco Manacorda. 2010. "Minimum Wages and Earnings Inequality in Urban
Mexico." American Economic Journal: Applied Economics,2(4): 128-149.
Card and Krueger. 1994. “Minimum Wages and Employment: A Case Study of the Fast-Food Industry in
New Jersey and Pennsylvania.” The American Economic Review, Vol. 84, No. 4. pp. 772-793.
Daly, M.c. & Valletta, R.G, (2004). Inequality and Poverty in the United States: The Effects of Rising
Male Wage Dispersion and Changing Family Behavior (Working Paper 2000-06). Federal
Reserve Bank of San Francisco: http://www.frbsf.org/economic-research/files/wp00-06bk.pdf
David H. Autor, Alan Manning, and Christopher L. Smith. 2016. “The Contribution of the Minimum
Wage to US Wage Inequality over Three Decades:A Reassessment.” The American Economic
Journal: Applied Economics, 8(1): 58-99.
Draca,Mirko, Stephen Machin, and John Van Reenen. 2011. "Minimum Wages and Firm
Profitability." American Economic Journal: Applied Economics,3(1): 129-51.
Engerman, S. L. & Gallman, R. E. (Eds.). (2000). The Cambridge Economic History of the United States
(Vol. 3). Cambridge, United Kingdom: Syndicate of the University of Cambridge. Retrieved
February 8, 2016.
Freeman, Richard. 1996. "The Minimum Wage as a Redistributive Tool." The Economic Journal 106
(May): 639-649.
13
Daniel Brinkerhoff, Spring 2016
Hae-Young Lee,Jongsung Kim, Beom Cheol Cin. “Empirical Analysis on the Determinants of Income
Inequality in Korea.” International Journal of Advanced Science and Technology, Vol. 53, April,
2013.
Kuznets S., “Economic Growth and Income Inequality”, American Economic Review, vol. 45, (1955),
pp. 1-28.
Lee,D. S. (1999). Wage Inequality in the United States during the 1980s: Rising Dispersion or Falling
Minimum Wage?. Quarterly Journal Of Economics, 114(3), 977-1023.
Litwin, Benjamin S., "Determining the Effect of the Minimum Wage on Income Inequality" (2015).
Student Publications. Paper 300. http://cupola.gettysburg.edu/student_scholarship/300
Neumark, David, and Wascher,William . "DO MINIMUM WAGES FIGHT POVERTY."Economic
Inquiry 40, no. 3 (July 2002): 315-333. Academic Search Premier, EBSCOhost (accessed
September 29, 2015).
Obstfeld, M. (1998). The Global Capital Market: Benefactor or Menace?. JournalOf Economic
Perspectives,12(4), 9-30. doi:http://dx.doi.org/10.1257/jep.12.4.9
Rittenberg, L. (2012). Macroeconomics Principles (Vol. 1.1). Creative Commons.
http://2012books.lardbucket.org/books/macroeconomics-principles-v1.1/s21-inequality-poverty-
and-discrim.html
Rubin, Rose M, White-Means, Shelley I. And Daniel, Luojia Mao. “Income distribution of older
Americans.” Bureau of Labor Statistics. http://www.bls.gov/opub/mlr/2000/11/art2full.pdf
Data Sources
Bankrate.com,Originally from Institute on Taxation & Economic:Policy. [List of States without Income
Tax]. Retreaved from http://www.bankrate.com/finance/taxes/state-with-no-income-tax-better-
or-worse-1.aspx
Bureau of Economic Analysis. Regional Economic Accounts: Download [Yearly Change in Real GDP
per Capita by State]. Retrieved from http://www.bea.gov/regional/downloadzip.cfm
Bureau of Economic Analysis. Regional Economic Accounts: Download [Yearly GDP by State].
Retrieved from http://www.bea.gov/regional/downloadzip.cfm
Bureau of Labor Statistics. Consumer Price Index Data from 1913 to 2016.[CPI and Inflation by of US by
year]. Retrieved from http://www.usinflationcalculator.com/inflation/consumer-price-index-and-
annual-percent-changes-from-1913-to-2008/
14
Daniel Brinkerhoff, Spring 2016
Iowa State University, Iowa Community Indicators Program [Unemployment by State]. Originally from
the Current Population Survey. Retrieved from
http://www.icip.iastate.edu/tables/employment/unemployment-states
Office of Communications Wage and Hour Division U.S. Department of Labor. last revised in December
2014. Changes in Basic Minimum Wage in Non-Farm Employment Under State law: Selected
years 1968 to 2016. United States Department of Labor.[Federaland state minimum wage by
state and year]. Retrieved From http://www.dol.gov/whd/state/stateMinWageHis.htm
The National Bureau of Economic Research,Maximum State Income Tax Rates 1977-2014 [Tax Rates
by State]. Retrieved from http://users.nber.org/~taxsim/state-rates/
U.S. Census Bureau, 2006-2014 American Community Survey. [Educational Attainment by State and
Share of Elderly]. Retreived from http://factfinder.census.gov/faces/nav/jsf/pages/index.xhtml
U.S. Census Bureau, 2006-2014 Gini Index of Income Inequality. [Gini Index by state and year].
Retrieved from http://factfinder.census.gov/faces/nav/jsf/pages/index.xhtml
Other Sources
Aaron, “Addressing ‘endogeneity’“ 2005-12-21.
http://pages.uoregon.edu/aarong/teaching/G4075_Outline/node9.html
U.S. Census Bureau, About Income. https://www.census.gov/hhes/www/income/about/
U.S. Census Bureau, Definitions.
https://www.census.gov/hhes/www/poverty/methods/definitions.html
15
Daniel Brinkerhoff, Spring 2016
Appendix
Table 1
Variable
Name
Description Expected
Sign on
Coefficient
Original
Source
of Data
Data
Set
Mean
Data
Set
STD
𝐺𝑖𝑛𝑖 𝑠𝑡 Gini Index by state and year. Measure
between 0-1, Where 0 is perfect equality,
and 1 is perfect inequality.
Endogenous
variable
U.S. Census
Bureau
0.4545 .02014
𝑅𝑀𝑊𝑠𝑡 Real minimum wage by state and year.
2007 as base year.
Negative U.S.
Department of
Labor
6.9824 .90260
𝑅𝑀𝑊𝑠𝑡
2
Real minimum wage squared, by state
and year. 2007 as base year
Positive U.S.
Department of
Labor
49.566 12.122
𝐼𝑛𝑓𝑡 Inflation by year. Negative Bureau of
Labor
Statistics
-0.0179 .12134
𝛥 𝐺𝐷𝑃𝑝𝐶 𝑠𝑡 Change in real GDP per capita by state
and year. 2007 as base year.
Negative Bureau of
Economic
Analysis
0.0029 .02612
𝑈𝑅𝑠𝑡 Unemployment rate by state and year. Positive Current
Population
Survey
6.5878 2.2343
%𝑃𝑜𝑝65 𝑠𝑡 Percent of population 65 and older, by
state and year.
Positive U.S. Census
Bureau
0.1769 .02107
𝑀𝑊𝑇𝑠𝑡 State Maximum tax rate on wages by
state and year.
Negative The National
Bureau of
Economic
Research
5.1386 2.9814
𝑀𝐿𝐺 𝑇𝑠𝑡 State Maximum tax rate on long gains by
state and year.
Positive The National
Bureau of
Economic
Research
4.7750 2.9949
%𝐿𝐻𝑆 𝑠𝑡 Percent of population with less than High
School education by state and year.
Positive U.S. Census
Bureau
0.1320 .03323
%𝐻𝑆𝐸𝑠𝑡 Percent of population with High School
education or equivalent by state and
year.
Negative U.S. Census
Bureau
0.3003 .03653
%𝐶𝐴𝑆 𝑠𝑡 Percent of population with some college
or Associate's degree by state or year.
Negative U.S. Census
Bureau
0.3158 .03579
16
Daniel Brinkerhoff, Spring 2016
%𝐵𝑜𝑟𝐻𝑠𝑡 Percent of population with Bachelor's
Degree or higher by state and year.
Negative U.S. Census
Bureau
0.2478 .04355
𝐼𝑇𝑠 Dummy variable (1) if state has income
tax. (0) if state does not have income
tax.
N/A Institute on
Taxation &
Economic:
Policy
N/A N/A
Year Fixed effect for years between 2009-
2014. (2009 as reference variable)
N/A N/A N/A N/A
State Fixed effect for each state. (Alabama as
reference variable)
N/A N/A N/A N/A
Data Conversion Methods
Not all data was found in the required format for use in this paper. Some variables had to be
converted or modified. This section explains all modifications done to change the data from its
original format, to the one used in this paper.
State Minimum Wage
● Data was originally in nominal dollars. It was converted to real dollars using the CPI
constructed by Bureau of Labor Statistics. The base year was set to 2007 to match the Real
GDP per Capita base year.
● Only years that saw a change in the federalor state minimum wage were recorded originally. The
values were extended to all years.
● All instances where the state minimum wage was lower than the federalminimum wage,the
federalvalue was used because that is what businesses would have been legally required to pay.
● Some states issued multiple minimum wages,where the smaller wage was to be used by
qualifying small businesses. These values were ignored in this paper.
17
Daniel Brinkerhoff, Spring 2016
Real GDP per Capita
● Original source gave total numbers by year and state. This paper converted them into
‘Change in’ values using the equation
○ (Current year total - Last year total)/(Current year Total)
Education
● Original data source had total number of people in each of the education categories. It
split people into 5 age categories and two gender categories. This paper added up all
people from each education category and divided it by the total population to get percent
of population.
● This data source was also used generate the percent of population over 65+.
Regression Outputs
Term Coef SE Coef T-Value P-Value VIF
Constant 0.5287 0.0489 10.81 0
Real Minimum
Wage -0.00865 0.00523 -1.65 0.099 331.9
Real Minimum
Wage Squared 0.000653 0.000412 1.58 0.114 371.85
Inflation -1.638 0.321 -5.11 0 22562.86
%Change GDP
per Capita 0.03 0.0141 2.13 0.034 2.01
Unemployment
Rate 0.000613 0.000385 1.59 0.112 10.98
State Max Wage
Tax -0.002474 0.000741 -3.34 0.001 72.62
State Max Long
Gains Tax 0.001703 0.000724 2.35 0.019 69.95
% Less Than
Highschool 0.0298 0.0682 0.44 0.662 77.22
% Highschool +
Equevelent -0.0438 0.0618 -0.71 0.479 76.34
18
Daniel Brinkerhoff, Spring 2016
% Some
College +
Associates -0.0131 0.0629 -0.21 0.836 75.45
% Bachelors or
Higher -0.1101 0.0621 -1.77 0.077 110.11
Has Income Tax 0.04356 0.00659 6.61 0 95.64
2007 0.00473 0.00121 3.9 0 2.17
2008 0.02275 0.00482 4.71 0 34.27
2009 -0.636 0.123 -5.16 0 22299.79
2010 -0.01169 0.00197 -5.92 0 5.73
2011 0.00927 0.00303 3.06 0.002 13.51
2012 -0.00494 0.00156 -3.16 0.002 3.59
2013 -0.01658 0.00373 -4.45 0 20.46
Alaska -0.01133 0.00329 -3.44 0.001 3.16
Arizona -0.00918 0.00446 -2.06 0.04 5.82
Arkansas -0.00073 0.00408 -0.18 0.857 4.86
California 0.01461 0.00757 1.93 0.054 16.75
Colorado 0.00032 0.00853 0.04 0.97 21.28
Connecticut 0.0338 0.00866 3.9 0 21.9
Delaware -0.01797 0.00521 -3.45 0.001 7.94
Florida 0.04994 0.00548 9.11 0 8.78
Georgia 0.00834 0.00407 2.05 0.041 4.85
Hawaii -0.01951 0.00727 -2.68 0.008 15.46
Idaho -0.03141 0.00572 -5.49 0 9.57
Illinois 0.00829 0.00588 1.41 0.159 10.1
Indiana -0.02595 0.00373 -6.97 0 4.06
Iowa -0.03181 0.0058 -5.49 0 9.82
Kansas -0.01225 0.00628 -1.95 0.052 11.5
Kentucky -0.00116 0.00333 -0.35 0.729 3.24
Louisiana 0.0103 0.00318 3.24 0.001 2.96
Maine -0.0134 0.00642 -2.09 0.037 12.03
Maryland -0.00882 0.00835 -1.06 0.291 20.35
Massachusetts 0.0225 0.01 2.25 0.025 29.24
Michigan -0.01258 0.00466 -2.7 0.007 6.35
Minnesota -0.01181 0.00806 -1.46 0.144 19
Mississippi 0.0034 0.00345 0.99 0.325 3.47
Missouri -0.00741 0.00414 -1.79 0.074 5.01
Montana -0.01168 0.00664 -1.76 0.079 12.88
Nebraska -0.02562 0.0063 -4.06 0 11.61
Nevada 0.01068 0.00539 1.98 0.048 8.49
19
Daniel Brinkerhoff, Spring 2016
New Hampshire 0.01413 0.00674 2.09 0.037 13.29
New Jersey 0.0154 0.00856 1.8 0.073 21.4
New Mexico 0.00341 0.00427 0.8 0.424 5.32
New York 0.0445 0.0068 6.54 0 13.53
North Carolina 0.00326 0.00452 0.72 0.471 5.96
North Dakota -0.01283 0.00706 -1.82 0.07 14.55
Ohio -0.00736 0.00466 -1.58 0.115 6.35
Oklahoma 0.00365 0.00548 0.67 0.506 8.77
Oregon -0.00987 0.00753 -1.31 0.191 16.57
Pennsylvania 0.0006 0.00603 0.1 0.92 10.64
Rhode Island 0.0038 0.00562 0.68 0.499 9.21
South Carolina 0.00272 0.00419 0.65 0.517 5.14
South Dakota 0.0176 0.00354 4.96 0 3.67
Tennessee 0.0462 0.0065 7.1 0 12.36
Texas 0.04596 0.00708 6.49 0 14.64
Utah -0.04365 0.00723 -6.04 0 15.27
Vermont -0.01537 0.00871 -1.76 0.078 22.18
Virginia 0.00721 0.00739 0.98 0.33 15.98
Washington 0.02181 0.00592 3.68 0 10.24
West Virginia -0.01269 0.00565 -2.24 0.025 9.34
Wisconsin -0.02232 0.00588 -3.79 0 10.11
Model Summary
S R-sq R-sq(adj) R-sq(pred)
0.0054943 93.81% 92.72% 91.27%

More Related Content

Similar to Final Paper

UNEMPLOYMENT work ( slides).pptx
UNEMPLOYMENT work ( slides).pptxUNEMPLOYMENT work ( slides).pptx
UNEMPLOYMENT work ( slides).pptxSadam Jamaldin
 
Canadian Journal of Economics Revue canadienne d’économique.docx
Canadian Journal of Economics  Revue canadienne d’économique.docxCanadian Journal of Economics  Revue canadienne d’économique.docx
Canadian Journal of Economics Revue canadienne d’économique.docxdewhirstichabod
 
Thesis in full_5.6
Thesis in full_5.6Thesis in full_5.6
Thesis in full_5.6Max Berre
 
Insert your surname 3NameInstructorInstitutionDate.docx
Insert your surname 3NameInstructorInstitutionDate.docxInsert your surname 3NameInstructorInstitutionDate.docx
Insert your surname 3NameInstructorInstitutionDate.docxdirkrplav
 
Response one PADM-05  Mortgage interest rates are expected to ri.docx
Response one PADM-05  Mortgage interest rates are expected to ri.docxResponse one PADM-05  Mortgage interest rates are expected to ri.docx
Response one PADM-05  Mortgage interest rates are expected to ri.docxronak56
 
An Ampirical Assessment Of The Contribution Of Small Business Employment To U...
An Ampirical Assessment Of The Contribution Of Small Business Employment To U...An Ampirical Assessment Of The Contribution Of Small Business Employment To U...
An Ampirical Assessment Of The Contribution Of Small Business Employment To U...Martha Brown
 
Explaining Minimum Wage Law Variances Across States (2014)
Explaining Minimum Wage Law Variances Across States (2014)Explaining Minimum Wage Law Variances Across States (2014)
Explaining Minimum Wage Law Variances Across States (2014)Corey Kozak
 
O Behave! Issue 3 (June Edition)
O Behave! Issue 3 (June Edition)O Behave! Issue 3 (June Edition)
O Behave! Issue 3 (June Edition)#ogilvychange
 
1Thesis Statement The minimum wage, the least possible pa.docx
1Thesis Statement The minimum wage, the least possible pa.docx1Thesis Statement The minimum wage, the least possible pa.docx
1Thesis Statement The minimum wage, the least possible pa.docxvickeryr87
 

Similar to Final Paper (12)

UNEMPLOYMENT work ( slides).pptx
UNEMPLOYMENT work ( slides).pptxUNEMPLOYMENT work ( slides).pptx
UNEMPLOYMENT work ( slides).pptx
 
Canadian Journal of Economics Revue canadienne d’économique.docx
Canadian Journal of Economics  Revue canadienne d’économique.docxCanadian Journal of Economics  Revue canadienne d’économique.docx
Canadian Journal of Economics Revue canadienne d’économique.docx
 
McNair Paper
McNair PaperMcNair Paper
McNair Paper
 
Thesis in full_5.6
Thesis in full_5.6Thesis in full_5.6
Thesis in full_5.6
 
Insert your surname 3NameInstructorInstitutionDate.docx
Insert your surname 3NameInstructorInstitutionDate.docxInsert your surname 3NameInstructorInstitutionDate.docx
Insert your surname 3NameInstructorInstitutionDate.docx
 
Response one PADM-05  Mortgage interest rates are expected to ri.docx
Response one PADM-05  Mortgage interest rates are expected to ri.docxResponse one PADM-05  Mortgage interest rates are expected to ri.docx
Response one PADM-05  Mortgage interest rates are expected to ri.docx
 
An Ampirical Assessment Of The Contribution Of Small Business Employment To U...
An Ampirical Assessment Of The Contribution Of Small Business Employment To U...An Ampirical Assessment Of The Contribution Of Small Business Employment To U...
An Ampirical Assessment Of The Contribution Of Small Business Employment To U...
 
Explaining Minimum Wage Law Variances Across States (2014)
Explaining Minimum Wage Law Variances Across States (2014)Explaining Minimum Wage Law Variances Across States (2014)
Explaining Minimum Wage Law Variances Across States (2014)
 
G.I.M. Final Paper
G.I.M. Final PaperG.I.M. Final Paper
G.I.M. Final Paper
 
O Behave! Issue 3 (June Edition)
O Behave! Issue 3 (June Edition)O Behave! Issue 3 (June Edition)
O Behave! Issue 3 (June Edition)
 
Wage Theories.pptx
Wage Theories.pptxWage Theories.pptx
Wage Theories.pptx
 
1Thesis Statement The minimum wage, the least possible pa.docx
1Thesis Statement The minimum wage, the least possible pa.docx1Thesis Statement The minimum wage, the least possible pa.docx
1Thesis Statement The minimum wage, the least possible pa.docx
 

Final Paper

  • 1. The State-Level Effects of the Real Minimum Wage on Income Distribution as Measured by the Gini Index. Author: Daniel Brinkerhoff, Spring 2016
  • 2. 1 Daniel Brinkerhoff, Spring 2016 Abstract The goal of many proponents of increasing the minimum wage is to decrease income inequality. This paper uses a multiple regression on the most commonly used determinants of income inequality as well as the real minimum wage and the real minimum wage squared to attempt to establish a link between an increasing minimum wage and a decreasing income inequality. This paper finds that an increase in the real minimum wage does tend to decrease income inequality for values below the turnaround point. In 2015 dollars the turnaround point was found to be $15.10. However, due to lack of data at, or around this value, this is only a rough approximation. If accurate, it would mean that increasing the wage to $15 an hour, like some organizations are fighting for, would cause a decrease in income inequality. Introduction The minimum wage is an important enough topic to be a part of the 2016 presidential election. Each candidate has been asked what their stances are, and movements like fightfor15.org/ encourage people to vote on the basis of that answer. Fast-Food workers all over the US have been going on strikes for a ‘living-wage.’ There are arguments claiming the low income caused by the low minimum wage is putting more people into situations where they require help from welfare programs; which acts as a pseudo subsidy for industries employing workers for minimum wage. Yet there are many who think that an increase in minimum wage would carry heavy costs in the form of decreasing the number of available jobs, especially for low skill, low experience workers who need it most. The decision to raise the minimum wage has continued to be a controversial topic. A topic where empirical analysis seems to clash with basic economic theory. This topic resurfaces periodically because the minimum wage is a nominal value. Every year it is not increased, is a
  • 3. 2 Daniel Brinkerhoff, Spring 2016 year where real minimum wage decreases by the inflation rate. This means that the answer to the question ‘What are the costs and benefits of increasing the minimum wage?’ Remains relevant year after year. Unfortunately, the costs and the benefits remain uncertain. With basic theory predicting a significant decrease to employment, but Card and Krueger1 failed to find an employment decrease in their empirical study of New Jersey’s minimum wage increase. Most people think that an increase to the minimum wage should increase the welfare of low income households, but Neumark and Wascher2 found that it “more closely resemble income redistribution among low-income families than income redistribution from high- to low-income families.” (Neumark and Wascher, 2002. P333). In light of these contradictions, and the continued need to adjust the minimum wage, more research on the subject is needed. Literature Review There is a lot of research going into minimum wages and their possible effects. As of right now, there does not seem to be a definitive answer to the question “What are the effects of Minimum wage increases?” Basic Supply and Demand theory would say that the obvious outcome is an increase of income for some workers, a loss of income for firms, and the loss of some jobs, with an overall total surplus decrease. Draca et al3 looked at the introduction of the 1999 national minimum wage in the UK. The authors used a difference in differences approach on private firms public accounting information from the FAME database to study how the 1999 national minimum wage affected wages and firm profitability. The authors found that “wages were significantly raised, and firm 1 CardandKrueger. 1994. “Minimum Wages and Employment: A Case Study of the Fast-FoodIndustryin NewJersey andPennsylvania.” The American Economic Review, Vol. 84,No. 4.pp.772-793. 2 Neumark, David, andWascher,William. "DO MINIMUMWAGESFIGHT POVERTY."Economic Inquiry 40, no. 3 (July 2002):315-333. Academic Search Premier, EBSCOhost (accessedSeptember 29, 2015) 3 Draca, Mirko, StephenMachin, andJohn Van Reenen. 2011."Minimum Wages andFirm Profitability." American Economic Journal:Applied Economics,3(1): 129-51.
  • 4. 3 Daniel Brinkerhoff, Spring 2016 profitability was significantly reduced by the minimum wage” (Draca et al, 2011. P130). They also found that firms with relatively high market power lost more profits but maintained their employment and productivity over other firms. Their findings fit with the expectations of the basic Supply and Demand theory that predicts a drop in producer surplus. Neumark and Wascher4 calculated the change in the probability of a family coming out of poverty vs the probability of a family falling into poverty due to an increase to the minimum wage. The idea being that the families below the poverty line would get an income boost from an increased minimum wage which could cause them to get above the poverty line while some families suffer from the disemployment effect and could fall below the poverty line. The authors found that there was an increased chance for families to get above the poverty line after a minimum wage increase, but also found a similar increase to the number of families who would fall below the poverty line. The net effect being an approximately equal change which causes a minimum wage increase to “more closely resemble income redistribution among low-income families than income redistribution from high- to low-income families.” (Neumark and Wascher, 2002. P333) Card and Krueger5 did an empirical study and found something very different. They took survey data of 410 fast-food restaurants in New Jersey and Pennsylvania before and after the 1992 increase to minimum wage in New Jersey. The authors found no evidence of lost employment due to the increased minimum wage. This is not predicted by the simple Supply and Demand model and implies that there is more happening than can be explained by the simple model. So a different one will be needed. Freeman’s6 Redistribution Theory tries to explain where the costs of a minimum wage increase would come from. The first group who bear the costs of a minimum wage increase are 4 Neumark, David, and Wascher,William. "DO MINIMUMWAGESFIGHT POVERTY."Economic Inquiry 40, no. 3 (July 2002):315-333. Academic Search Premier, EBSCOhost (accessedSeptember 29, 2015) 5 CardandKrueger. 1994. “Minimum Wages andEmployment: A Case Study of the Fast-FoodIndustryin NewJersey andPennsylvania.” The American Economic Review, Vol. 84,No. 4.pp.772-793. 6 Freeman, Richard. 1996. "The MinimumWage as a Redistributive Tool."TheEconomic Journal 106(May): 639-649.
  • 5. 4 Daniel Brinkerhoff, Spring 2016 the consumers of products and services generated by minimum wage in the form of higher prices for those goods and services. The simple assumption is that most people consume the products of minimum wage workers regardless of personal income. Even if this assumption is not strictly true, some of the costs of the minimum wage are borne by the consumers, which is significant to income distribution. The second group that will pay for a minimum wage increase will be the firms/stakeholders of companies that employ minimum wage workers. Because of the labor is one of the costs, an increase in costs will be partially borne by the firm. This is supported by Draca et al’s findings. The final group that would bear the costs of the minimum wage increase are those who will lose their jobs due to the increase in pay. This is supported by Neumark and Wascher but was not found to be an issue by Card and Krueger. A possible solution to the contradictory findings on the minimum wage's effects on employment could be that it is a non-linear relationship. Lee7 utilizes a quadratic minimum wage to account for the nonlinear properties of an increasing minimum wage. The idea being that an increase to the minimum wage is significantly more effective when it is binding (above natural equilibrium) in a sector than if, after the increase, it remains non-binding (below the natural equilibrium). The higher above the equilibrium price the minimum wage is, the more of a disemployment effect it will have. Trying to analyze and understand income inequality is far from a new idea. As such, others have tried to describe it before. Hae-Young Lee et al8 attempts to describe income inequality in Korea. Many of the variables they use will be applicable to the US, but they do not add minimum wage into their considerations which would be important to the US explanation. Bosch9 studies the effects of the real minimum wage on the increase in inequality observed in 7 Lee, D. S. (1999). Wage Inequality in the United States during the 1980s: Rising Dispersion or Falling Minimum Wage?. Quarterly Journal Of Economics, 114(3), 977-1023. 8 Hae-YoungLee, JongsungKim, BeomCheol Cin. “Empirical Analysis on the Determinants ofIncome Inequalityin Korea.” International Journal of AdvancedScience andTechnology, Vol. 53,April,2013. 9 Bosch, Mariano, andMarcoManacorda.2010. "Minimum Wages andEarnings Inequalityin UrbanMexico." AmericanEconomicJournal: Applied Economics,2(4): 128-149.
  • 6. 5 Daniel Brinkerhoff, Spring 2016 urban Mexico, he finds that the real minimum wage decrease was the cause of much of the large increase to income inequality. His findings should be very similar to mine because of the similarities in the question. However, his findings are for a different place, at a different time. He also does not attempt to understand the disemployment effect of the minimum wage, which will be captured by the quadratic minimum wage. Method In order to test the relationship between the real minimum wage and income inequality, this paper will use an Ordinary Least Square Regression on the following model: 𝐺𝑖𝑛𝑖 𝑠𝑡 = 𝛽0 + 𝛽1 𝑅𝑀𝑊𝑠𝑡 + 𝛽2 𝑅𝑀𝑊𝑠𝑡 2 + 𝛽3 𝐼𝑛𝑓𝑡 + 𝛽4 𝛥𝐺𝐷𝑃𝑝𝐶 𝑠𝑡 + 𝛽5 𝑈𝑅 𝑠𝑡 + 𝛽6%𝑃𝑜𝑝65 𝑠𝑡 + 𝛽7 𝑀𝑊𝑇𝑠𝑡 + 𝛽8 𝑀𝐿𝐺𝑇𝑠𝑡 + 𝛽9%𝐿𝐻𝑆 𝑠𝑡 + 𝛽10%𝐻𝑆𝐸 𝑠𝑡 + 𝛽11 %𝐶𝐴𝑆𝑠𝑡 + 𝛽12 %𝐵𝑜𝑟𝐻𝑠𝑡 + 𝛽13 𝐼𝑇𝑠 + 𝛽14 𝑌𝑒𝑎𝑟 + 𝛽15 𝑆𝑡𝑎𝑡𝑒 + 𝜀 See Table 1 for variable names and descriptions The hypothesis test is H0: 𝛽1= 0 Ha: 𝛽1< 0 And H0: 𝛽2= 0 Ha: 𝛽2> 0 This model attempts to explain the effects the real minimum wage has on the gini index while holding all other variables constant. The gini index is a measure of income inequality, ranging from 0 to 1. A score of 0 indicates perfect equality of income; each person makes the
  • 7. 6 Daniel Brinkerhoff, Spring 2016 same income as every other person. A score of 1 indicates perfect inequality; only one person makes all of the income (U.S. Census Bureau, Definitions). The definition of income used to measure the gini index is defined as wages received on a regular basis before taxes. This figure does not reflect the fact that some income can be received in non-cash benefits like food stamps, health benefits, access to business transportation or facilities, and retirement plans. (U.S. Census Bureau, About Income). This paper asserts that gini index is a product of 7 exogenous variables. The first exogenous variable, and the focus variable of this paper, is the real minimum wage. This paper predicts that as you increase the real minimum wage (to a point), the gini index will decrease, yielding a decrease to inequality. This happens because the lower income earners are more greatly boosted by an increase to real minimum wage than any other group. Causing the overall distribution to become more centric. However, as the bindingness (amount above the equilibrium price) increases, the disemployment effect increases more rapidly than the benefits of the increase to wages. As the disemployment effect increases, fewer people gain benefits from the increased minimum wage. Also, firms that had monopsony power may actually increase employment for small increases to minimum wage, but still decrease employment at a high enough minimum wage levels. Lee10 utilized a quadratic form with a negative minimum wage, and a positive minimum wage squared, to capture this effect. Second, Blinder11 found that inflation acted as a kind of ‘progressive tax’ where the lower class and middle class tend to lose less than the upper class. This would have a negative effect on the gini index and decrease inequality. 10 Lee, D. S. (1999). Wage Inequality in the United States during the 1980s: Rising Dispersion or Falling Minimum Wage?. Quarterly Journal Of Economics, 114(3), 977-1023. 11 Blinder, A. S., & Esaki, H. Y. (1978). Macroeconomic Activity and Income Distribution in the Postw ar United States. Review Of Economics And Statistics, 6(4), 604-609.
  • 8. 7 Daniel Brinkerhoff, Spring 2016 Third, the Kuznets curve12 indicates that as a developed country like the US grows, income inequality would tend to decrease; this would have a negative effect on the gini index. This paper uses the change in real income per capita to capture the state level growth. Forth, Blinder11 also found that the unemployment rate disproportionately decreases the income of the lowest 40% of income earners. This unequal effect of an increase to unemployment causes the left hand tail of the gini index to grow, increasing inequality. Unemployment is not an exogenous variable, however. This means it will likely have an endogeneity problem. Endogeneity can cause the variable to have reverse causality and will likely bias the coefficient and increase the associated p-value. This will not cause problems for any of the other coefficients or disrupt the overall regressions, but that the findings on this variable cannot be used as direct evidence of unemployment's’ effect on the gini index. Fifth, Rubin13 has found that the elderly population between 1967 and 1997 have a significantly higher gini index (more inequality) than the younger populations. The disparity was shrinking during his research, indicating a change in the working habits of the elderly. This means the value may have changed for the years between 2006-2014. The value has been added in for completeness. Sixth, taxes on the wealthiest income earners tends to act as a ‘braking system’ according to Alvaredo et al14 . The effect is that it lowered the return to effort on negotiating for higher pay. Because the higher income earners tend to have more leverage in negotiations, they tend to capture more the company's wealth at the cost of other employees. This means that as taxes on top income earners increases, the gini index should decrease, indicating a decrease to inequality. The measures being used to capture this effect are the State level max wage tax. In addition, this paper adds the state level max long gains tax that would be applied to 12 S. Kuznets, “Economic Grow th and Income Inequality”, American Economic Review , vol. 45, (1955), pp. 1-28. 13 Rubin, Rose M, White-Means, Shelley I. And Daniel, Luojia Mao. “Income distribution of older Americans.” (2000). Bureau of Labor Statistics. 14 Alvaredo, Facundo,Atkinson, Anthony B, Piketty, Thomas,andSaez, Emmanuel. (2013).”The Top 1 Percent in International andHistorical Perspective.” Journal ofEconomic Perspectives, Volume 27, Number 3,Summer 2013, Pages 3–20.
  • 9. 8 Daniel Brinkerhoff, Spring 2016 capital investments. The State level max wage tax should directly measure the effect that Alvaredo et al describes while the max long gains tax should be in opposition to Alvaredo et al’s findings as it increases the importance of earned income by decreasing the wealth gained by investing. An important reminder, the income measured by the gini index does NOT include the income gained by long gains, and it applies before taxes. This means that it will not be directly affected by those taxes, only indirectly in the ways described by Alvaredo et al. Seventh, because wages are a representation of a worker's productivity, as productivity increases, so should wage. If access to education is increased, so too will productivity and wage on average. This paper uses the percent of the population who have less than a high school education, high school education, some college or associate's degree, and Bachelor's or higher, as a measure for the accessibility of schooling. This paper also uses a fixed effect for the year to capture changes on a federal level that would not be specifically captured by state level data. This paper also uses a fixed effects for each state to capture policy and cultural differences that may not be represented by the exogenous variables. One state level policy (whether or not a state has an income tax) has been moved out of the state fixed effect and added as a dummy variable for better precision. Analysis This initial model utilizes 50 states over the course of 9 years. This generates 450 points of data for each of the 12 continuous variables, including the endogenous variable. This paper also uses a total of 58 dummy variables to represent the 50 states, the 9 years, and whether or not a state has an income tax; note that Alabama and 2006 are used as the reference variables. The initial model has 379 (450 - 58 - 12 -1) degrees of freedom. The final model only removes one variable (Percent of population 65 years and older) due to a low Test Statistic of -.44 and with an associated P-Value of .659. This increases the
  • 10. 9 Daniel Brinkerhoff, Spring 2016 degrees of freedom of the final model to 380. This final model has an adjusted R-Squared of 92.72% which is not uncommon in models that use a large number of fixed effects and dummy variables. See Regression Outputs in the appendix. Problemed Variables The proportion of the population above 65 has changed from the model prediction of positively affecting the gini index, to negatively affecting the gini index. However, with a very high P-Value, this number is not significantly different from zero. Rubin15 ’s findings showed that this difference was gradually decreasing between 1967-1997, and it is possible that there is no longer a gap between the 65 year and over population, and those younger. For these reasons, it has been dropped from the final model. Inflation’s coefficient has a negative sign instead of a the positive sign that was predicted. However, inflation was also the only variable that used year only level data and was undifferentiated for the states. The measure used for inflation has a very low P-Value, indicating that it is a good predictor, but due to the lack of specific data, the coefficient should be analyzed critically. The change in Real GDP per capita is strangely a positive figure when the Kuznets curve would predict a negative coefficient. This value is also significant to the 5% level. However, this wrong sign was also found in Hae-Young Lee et al16 ’s research on the determinants of income inequality in South Korea. A possible explanation could be that due to the recession of 2007-2010 and the atypical monetary and fiscal response, the measures used 15 Rubin, Rose M, White-Means, Shelley I. And Daniel, Luojia Mao. “Income distribution of older Americans.” (2000). Bureau of Labor Statistics. 16 Hae-YoungLee, JongsungKim, BeomCheol Cin. “Empirical Analysis on the Determinants ofIncome Inequalityin Korea.” International Journal of AdvancedScience andTechnology, Vol. 53,April,2013.
  • 11. 10 Daniel Brinkerhoff, Spring 2016 were not typical and if the timeframe where to be increased, a different coefficient could possibly be realized. Unemployment is not statistically significant to the 10% level. This is likely due to the fact that unemployment is an endogenous variable being used as an explanatory variable. This can commonly give higher P-values than would be typically found if the determinants of unemployment were used instead. For the purposes of this paper, the unemployment rate controls for some of these variables, and the high P-value is not detrimental. Only one of the measures used to define education was found to be statistically significant to the 10% level. Because the values are linked (a decrease in one, necessitates the increase of another) they are all kept in the final model. This does, however, mean that there is likely a better measure for education than the percent of the population that has attained a specific level of education. Gini Index data set The gini index between 2006-2014 increased in almost all states. Hawaii being the only state that saw a decrease over that time period. While over the course of the entire range, each state saw an increase, each year the gini index only increased approximately 65% of the time. State Fixed Effect The coefficient of the state fixed effect had a very low average with a comparatively high standard deviation. This indicates that there was a very large variation, both positive and negative, in the state coefficients. However, the sum of the coefficients was very close to zero. The 5 states that had the largest positive coefficient, which indicates larger amounts of income inequality, were Florida, Tennessee, Texas, new York, and Connecticut. These states tend to be on or close to the East Coast, they have a relatively higher cost of living, and relatively higher
  • 12. 11 Daniel Brinkerhoff, Spring 2016 per capita GDP. While the 5 states with the largest negative coefficients, which indicates lower amounts of income inequality, were Idaho, Iowa, Indiana, Nebraska, and Utah. These states tend to be in the middle of the country without access to the ocean, they have a relatively lower cost of living, and relatively lower GDP per capita. The state fixed effect generally had a very low P-Value. This can mean a variety of things. It is possible that there are determinants that are missing from the main model, that are being captured by the state fixed effects. There could be state specific policies or culture that reduce/add income inequality within its borders. Income inequality may also be strongly tied to past income inequality, which would mean that states tend to cling to former values and only adjust slowly to new values. Minimum Wage’s Effect on Income Inequality The Real Minimum Wage was found to be statistically significant at the 10%. This allows us to reject the null hypotheses in favor of the alternative. The Real Minimum Wage decreases income inequality. However, the Real Minimum Wage Squared was not found to be statistically significant at the 10% level. We fail to reject the null hypothesis. The real minimum wage and the real minimum wage squared did have the predicted signs with the real minimum wage being negative and the real minimum wage squared being positive. They have a turnaround point at a real wage of $13.20 (in 2007 dollars) where further increases to the minimum wage would yield positive effects on income inequality. A potential reason for the high P-Value on the minimum wage squared could be due to a lack of data at higher levels of real minimum wage. Using a different data set where the real wage saw more values at the higher levels may yield a lower P-Value.
  • 13. 12 Daniel Brinkerhoff, Spring 2016 Citations and References Articles Aaronson, Daniel, Sumit Agarwal,and Eric French. 2012. "The Spending and Debt Response to Minimum Wage Hikes." American Economic Review,102(7): 3111-39. Alvaredo, Facundo, Atkinson, Anthony B, Piketty, Thomas, and Saez, Emmanuel. (2013).”The Top 1 Percent in International and Historical Perspective.” Journalof Economic Perspectives,Volume 27, Number 3, Summer 2013, Pages 3–20. Blinder, A. S., & Esaki, H. Y. (1978). Macroeconomic Activity and Income Distribution in the Postwar United States. Review Of Economics And Statistics, 6(4), 604-609. Bosch, Mariano, and Marco Manacorda. 2010. "Minimum Wages and Earnings Inequality in Urban Mexico." American Economic Journal: Applied Economics,2(4): 128-149. Card and Krueger. 1994. “Minimum Wages and Employment: A Case Study of the Fast-Food Industry in New Jersey and Pennsylvania.” The American Economic Review, Vol. 84, No. 4. pp. 772-793. Daly, M.c. & Valletta, R.G, (2004). Inequality and Poverty in the United States: The Effects of Rising Male Wage Dispersion and Changing Family Behavior (Working Paper 2000-06). Federal Reserve Bank of San Francisco: http://www.frbsf.org/economic-research/files/wp00-06bk.pdf David H. Autor, Alan Manning, and Christopher L. Smith. 2016. “The Contribution of the Minimum Wage to US Wage Inequality over Three Decades:A Reassessment.” The American Economic Journal: Applied Economics, 8(1): 58-99. Draca,Mirko, Stephen Machin, and John Van Reenen. 2011. "Minimum Wages and Firm Profitability." American Economic Journal: Applied Economics,3(1): 129-51. Engerman, S. L. & Gallman, R. E. (Eds.). (2000). The Cambridge Economic History of the United States (Vol. 3). Cambridge, United Kingdom: Syndicate of the University of Cambridge. Retrieved February 8, 2016. Freeman, Richard. 1996. "The Minimum Wage as a Redistributive Tool." The Economic Journal 106 (May): 639-649.
  • 14. 13 Daniel Brinkerhoff, Spring 2016 Hae-Young Lee,Jongsung Kim, Beom Cheol Cin. “Empirical Analysis on the Determinants of Income Inequality in Korea.” International Journal of Advanced Science and Technology, Vol. 53, April, 2013. Kuznets S., “Economic Growth and Income Inequality”, American Economic Review, vol. 45, (1955), pp. 1-28. Lee,D. S. (1999). Wage Inequality in the United States during the 1980s: Rising Dispersion or Falling Minimum Wage?. Quarterly Journal Of Economics, 114(3), 977-1023. Litwin, Benjamin S., "Determining the Effect of the Minimum Wage on Income Inequality" (2015). Student Publications. Paper 300. http://cupola.gettysburg.edu/student_scholarship/300 Neumark, David, and Wascher,William . "DO MINIMUM WAGES FIGHT POVERTY."Economic Inquiry 40, no. 3 (July 2002): 315-333. Academic Search Premier, EBSCOhost (accessed September 29, 2015). Obstfeld, M. (1998). The Global Capital Market: Benefactor or Menace?. JournalOf Economic Perspectives,12(4), 9-30. doi:http://dx.doi.org/10.1257/jep.12.4.9 Rittenberg, L. (2012). Macroeconomics Principles (Vol. 1.1). Creative Commons. http://2012books.lardbucket.org/books/macroeconomics-principles-v1.1/s21-inequality-poverty- and-discrim.html Rubin, Rose M, White-Means, Shelley I. And Daniel, Luojia Mao. “Income distribution of older Americans.” Bureau of Labor Statistics. http://www.bls.gov/opub/mlr/2000/11/art2full.pdf Data Sources Bankrate.com,Originally from Institute on Taxation & Economic:Policy. [List of States without Income Tax]. Retreaved from http://www.bankrate.com/finance/taxes/state-with-no-income-tax-better- or-worse-1.aspx Bureau of Economic Analysis. Regional Economic Accounts: Download [Yearly Change in Real GDP per Capita by State]. Retrieved from http://www.bea.gov/regional/downloadzip.cfm Bureau of Economic Analysis. Regional Economic Accounts: Download [Yearly GDP by State]. Retrieved from http://www.bea.gov/regional/downloadzip.cfm Bureau of Labor Statistics. Consumer Price Index Data from 1913 to 2016.[CPI and Inflation by of US by year]. Retrieved from http://www.usinflationcalculator.com/inflation/consumer-price-index-and- annual-percent-changes-from-1913-to-2008/
  • 15. 14 Daniel Brinkerhoff, Spring 2016 Iowa State University, Iowa Community Indicators Program [Unemployment by State]. Originally from the Current Population Survey. Retrieved from http://www.icip.iastate.edu/tables/employment/unemployment-states Office of Communications Wage and Hour Division U.S. Department of Labor. last revised in December 2014. Changes in Basic Minimum Wage in Non-Farm Employment Under State law: Selected years 1968 to 2016. United States Department of Labor.[Federaland state minimum wage by state and year]. Retrieved From http://www.dol.gov/whd/state/stateMinWageHis.htm The National Bureau of Economic Research,Maximum State Income Tax Rates 1977-2014 [Tax Rates by State]. Retrieved from http://users.nber.org/~taxsim/state-rates/ U.S. Census Bureau, 2006-2014 American Community Survey. [Educational Attainment by State and Share of Elderly]. Retreived from http://factfinder.census.gov/faces/nav/jsf/pages/index.xhtml U.S. Census Bureau, 2006-2014 Gini Index of Income Inequality. [Gini Index by state and year]. Retrieved from http://factfinder.census.gov/faces/nav/jsf/pages/index.xhtml Other Sources Aaron, “Addressing ‘endogeneity’“ 2005-12-21. http://pages.uoregon.edu/aarong/teaching/G4075_Outline/node9.html U.S. Census Bureau, About Income. https://www.census.gov/hhes/www/income/about/ U.S. Census Bureau, Definitions. https://www.census.gov/hhes/www/poverty/methods/definitions.html
  • 16. 15 Daniel Brinkerhoff, Spring 2016 Appendix Table 1 Variable Name Description Expected Sign on Coefficient Original Source of Data Data Set Mean Data Set STD 𝐺𝑖𝑛𝑖 𝑠𝑡 Gini Index by state and year. Measure between 0-1, Where 0 is perfect equality, and 1 is perfect inequality. Endogenous variable U.S. Census Bureau 0.4545 .02014 𝑅𝑀𝑊𝑠𝑡 Real minimum wage by state and year. 2007 as base year. Negative U.S. Department of Labor 6.9824 .90260 𝑅𝑀𝑊𝑠𝑡 2 Real minimum wage squared, by state and year. 2007 as base year Positive U.S. Department of Labor 49.566 12.122 𝐼𝑛𝑓𝑡 Inflation by year. Negative Bureau of Labor Statistics -0.0179 .12134 𝛥 𝐺𝐷𝑃𝑝𝐶 𝑠𝑡 Change in real GDP per capita by state and year. 2007 as base year. Negative Bureau of Economic Analysis 0.0029 .02612 𝑈𝑅𝑠𝑡 Unemployment rate by state and year. Positive Current Population Survey 6.5878 2.2343 %𝑃𝑜𝑝65 𝑠𝑡 Percent of population 65 and older, by state and year. Positive U.S. Census Bureau 0.1769 .02107 𝑀𝑊𝑇𝑠𝑡 State Maximum tax rate on wages by state and year. Negative The National Bureau of Economic Research 5.1386 2.9814 𝑀𝐿𝐺 𝑇𝑠𝑡 State Maximum tax rate on long gains by state and year. Positive The National Bureau of Economic Research 4.7750 2.9949 %𝐿𝐻𝑆 𝑠𝑡 Percent of population with less than High School education by state and year. Positive U.S. Census Bureau 0.1320 .03323 %𝐻𝑆𝐸𝑠𝑡 Percent of population with High School education or equivalent by state and year. Negative U.S. Census Bureau 0.3003 .03653 %𝐶𝐴𝑆 𝑠𝑡 Percent of population with some college or Associate's degree by state or year. Negative U.S. Census Bureau 0.3158 .03579
  • 17. 16 Daniel Brinkerhoff, Spring 2016 %𝐵𝑜𝑟𝐻𝑠𝑡 Percent of population with Bachelor's Degree or higher by state and year. Negative U.S. Census Bureau 0.2478 .04355 𝐼𝑇𝑠 Dummy variable (1) if state has income tax. (0) if state does not have income tax. N/A Institute on Taxation & Economic: Policy N/A N/A Year Fixed effect for years between 2009- 2014. (2009 as reference variable) N/A N/A N/A N/A State Fixed effect for each state. (Alabama as reference variable) N/A N/A N/A N/A Data Conversion Methods Not all data was found in the required format for use in this paper. Some variables had to be converted or modified. This section explains all modifications done to change the data from its original format, to the one used in this paper. State Minimum Wage ● Data was originally in nominal dollars. It was converted to real dollars using the CPI constructed by Bureau of Labor Statistics. The base year was set to 2007 to match the Real GDP per Capita base year. ● Only years that saw a change in the federalor state minimum wage were recorded originally. The values were extended to all years. ● All instances where the state minimum wage was lower than the federalminimum wage,the federalvalue was used because that is what businesses would have been legally required to pay. ● Some states issued multiple minimum wages,where the smaller wage was to be used by qualifying small businesses. These values were ignored in this paper.
  • 18. 17 Daniel Brinkerhoff, Spring 2016 Real GDP per Capita ● Original source gave total numbers by year and state. This paper converted them into ‘Change in’ values using the equation ○ (Current year total - Last year total)/(Current year Total) Education ● Original data source had total number of people in each of the education categories. It split people into 5 age categories and two gender categories. This paper added up all people from each education category and divided it by the total population to get percent of population. ● This data source was also used generate the percent of population over 65+. Regression Outputs Term Coef SE Coef T-Value P-Value VIF Constant 0.5287 0.0489 10.81 0 Real Minimum Wage -0.00865 0.00523 -1.65 0.099 331.9 Real Minimum Wage Squared 0.000653 0.000412 1.58 0.114 371.85 Inflation -1.638 0.321 -5.11 0 22562.86 %Change GDP per Capita 0.03 0.0141 2.13 0.034 2.01 Unemployment Rate 0.000613 0.000385 1.59 0.112 10.98 State Max Wage Tax -0.002474 0.000741 -3.34 0.001 72.62 State Max Long Gains Tax 0.001703 0.000724 2.35 0.019 69.95 % Less Than Highschool 0.0298 0.0682 0.44 0.662 77.22 % Highschool + Equevelent -0.0438 0.0618 -0.71 0.479 76.34
  • 19. 18 Daniel Brinkerhoff, Spring 2016 % Some College + Associates -0.0131 0.0629 -0.21 0.836 75.45 % Bachelors or Higher -0.1101 0.0621 -1.77 0.077 110.11 Has Income Tax 0.04356 0.00659 6.61 0 95.64 2007 0.00473 0.00121 3.9 0 2.17 2008 0.02275 0.00482 4.71 0 34.27 2009 -0.636 0.123 -5.16 0 22299.79 2010 -0.01169 0.00197 -5.92 0 5.73 2011 0.00927 0.00303 3.06 0.002 13.51 2012 -0.00494 0.00156 -3.16 0.002 3.59 2013 -0.01658 0.00373 -4.45 0 20.46 Alaska -0.01133 0.00329 -3.44 0.001 3.16 Arizona -0.00918 0.00446 -2.06 0.04 5.82 Arkansas -0.00073 0.00408 -0.18 0.857 4.86 California 0.01461 0.00757 1.93 0.054 16.75 Colorado 0.00032 0.00853 0.04 0.97 21.28 Connecticut 0.0338 0.00866 3.9 0 21.9 Delaware -0.01797 0.00521 -3.45 0.001 7.94 Florida 0.04994 0.00548 9.11 0 8.78 Georgia 0.00834 0.00407 2.05 0.041 4.85 Hawaii -0.01951 0.00727 -2.68 0.008 15.46 Idaho -0.03141 0.00572 -5.49 0 9.57 Illinois 0.00829 0.00588 1.41 0.159 10.1 Indiana -0.02595 0.00373 -6.97 0 4.06 Iowa -0.03181 0.0058 -5.49 0 9.82 Kansas -0.01225 0.00628 -1.95 0.052 11.5 Kentucky -0.00116 0.00333 -0.35 0.729 3.24 Louisiana 0.0103 0.00318 3.24 0.001 2.96 Maine -0.0134 0.00642 -2.09 0.037 12.03 Maryland -0.00882 0.00835 -1.06 0.291 20.35 Massachusetts 0.0225 0.01 2.25 0.025 29.24 Michigan -0.01258 0.00466 -2.7 0.007 6.35 Minnesota -0.01181 0.00806 -1.46 0.144 19 Mississippi 0.0034 0.00345 0.99 0.325 3.47 Missouri -0.00741 0.00414 -1.79 0.074 5.01 Montana -0.01168 0.00664 -1.76 0.079 12.88 Nebraska -0.02562 0.0063 -4.06 0 11.61 Nevada 0.01068 0.00539 1.98 0.048 8.49
  • 20. 19 Daniel Brinkerhoff, Spring 2016 New Hampshire 0.01413 0.00674 2.09 0.037 13.29 New Jersey 0.0154 0.00856 1.8 0.073 21.4 New Mexico 0.00341 0.00427 0.8 0.424 5.32 New York 0.0445 0.0068 6.54 0 13.53 North Carolina 0.00326 0.00452 0.72 0.471 5.96 North Dakota -0.01283 0.00706 -1.82 0.07 14.55 Ohio -0.00736 0.00466 -1.58 0.115 6.35 Oklahoma 0.00365 0.00548 0.67 0.506 8.77 Oregon -0.00987 0.00753 -1.31 0.191 16.57 Pennsylvania 0.0006 0.00603 0.1 0.92 10.64 Rhode Island 0.0038 0.00562 0.68 0.499 9.21 South Carolina 0.00272 0.00419 0.65 0.517 5.14 South Dakota 0.0176 0.00354 4.96 0 3.67 Tennessee 0.0462 0.0065 7.1 0 12.36 Texas 0.04596 0.00708 6.49 0 14.64 Utah -0.04365 0.00723 -6.04 0 15.27 Vermont -0.01537 0.00871 -1.76 0.078 22.18 Virginia 0.00721 0.00739 0.98 0.33 15.98 Washington 0.02181 0.00592 3.68 0 10.24 West Virginia -0.01269 0.00565 -2.24 0.025 9.34 Wisconsin -0.02232 0.00588 -3.79 0 10.11 Model Summary S R-sq R-sq(adj) R-sq(pred) 0.0054943 93.81% 92.72% 91.27%