The Cost of Independence: An Argument on Not Leaving Home Just Yet

By: Gil Puyat

Abstract

This paper will examine the role of parental wealth in generating a strong positive effect on earnings. It compares the benefits of staying at home versus being independent by running a regression on collected population survey data from the IPUMS-CPS database. Parental aid can possibly reduce the burden on an individual, and allow them to be more productive in education; thus having a positive effect on future earnings. The results suggest that parental aid in the form of staying at home have a significant positive effect on an individuals future income.

Introduction

Most of us have faced the decision of leaving home and being independent, and in one way or another we have also all weighed the costs and benefits of our decisions. Here an argument is made and proven that staying at home has significant benefits to the quality of education received, be it through the aid of resources from parents, or through the relief from the burden of being on your own. The increase in quality of education is in turn affected by a direct positive relation to the amount of income an individual may earn. This paper tries to develop some evidence that may help argue against not leaving home just yet. Literature Review

There have been many studies on the relationship between education and earnings, and the role of parents in transferring wealth through generations. Maurer-Fazio & Dinh (2004) describe the role of education in determining a workers income in urban Chinas labor markets. They find that returns to education for workers who have found employment by means of a competitive method to be higher than those whose jobs were assigned. The authors decompose earnings differentials based on worker data from 1999-2000 by surveying 4873 individuals from 118 enterprises in six cities. They find that workers who

have labored for their positions tend to be more productive. Through random sampling of census data collected from a wide spread area, and applying regression analysis, they were able to accurately describe the population.

In, Blanden, Gregg & Macmillan (2007) they claim, those born to poor parents have restricted mobility and often do not achieve their economic potential. Their study proposes that education is the most obvious and efficient means of transmission explaining intergenerational mobility. A strong association between incomes across generations indicates weak intergenerational mobility. Cognitive ability offers a substantive contribution to mobility, but only if given ample opportunity to be nurtured. This study differs in such a way that it shall be looking at the simple relation of two variables to explain their causal relationships.

In Liu and Zeng (2007), they examine the role of genetic ability in generating strong positive role on intergenerational earnings in the U.S. The study finds the differences between adopted and non-adopted childrens ability and its importance in transmitting earnings ability from their parents. The authors measure IQ level and compare them to kinship correlations. Regression results from kinship correlations converge on the conclusion that differences in ability can explain a substantial fraction of the variation in IQ. They summarize that about half of the variation in IQ among individuals in the population can be explained by genetic ability.

In Behrman and Rosenzweig (2005), the authors find that parental resources are important in determining such factors as childrens human capital, returns to schooling, and future earnings. Newly available data on parent and in-laws indicate that parental resources of marital partners may affect resource distribution within marriage. Regression results from these data sets show the effects of parent and in-law characteristics such as transfers, bequests, and visits by parents and in-laws, have a large effect on childrens human capital, returns to schooling, and earnings. They find that parental wealth has economic advantages for children even as they become adults, as they tend to have more education, higher quality education, better connections for jobs, and transfers from gifts or bequests. Thus, in addition to direct

income advantages which are facilitated by greater human capital investment, those with wealthier parents enjoy additional consumption benefits.

Theoretical Argument

By the human capital theory, investment in human capital, such as education, and training, increases the productivity of an individual by cultivating skills and adding to knowledge. A well educated individual is more productive and capable, which makes them more profitable for employers. High profitability increases competition in the job market, thus possibly increasing wages as demand for jobs increase. An educated worker has acquired the basic skills in problem solving and analysis. Achieved through the academic process, training received during education makes workers more useful by increasing their cognitive ability. Another reason for profitability is that since employers have to spend less time and effort training workers for a position, workers can work independently with little or no guidance. Wages may also increase with the level of training due to higher competition in the job market. Employers prefer to hire, and may pay a premium for highly skilled and very productive individuals, who can increase their profits. All these factors contribute to possible explanations for the effects of education on earnings

More importantly, the effects of parental income on an individuals ability to attend school are considerable. By staying at home and receiving benefits from parents, an individual is less pressured and can freely pursue their educational goals with much more efficiency and less distraction. This leads to a higher overall quality experience that can translate into larger future financial gains by human capital accumulation. Staying at home alleviates the burden, which would otherwise be placed upon an individual. This can include a number of things like increased responsibility, or the incurring of expenses such as, but not limited to, rent, utilities, and meals. Free from excess load, physical and cognitive resources can be concentrated solely on the task of education.

Empirical testing

Methodology

Using Multiple Linear Regression techniques on IPUMS-CPS population survey data for the U.S. in 2008, the natural log of income wage was calculated as the dependent variable and factors such as age, sex, race, education level, region, hours worked, and a relation variable were used as independent variables, a regression line was estimated, and the relative probabilities were calculated for each variable. The main variable in question used to support my theory is a dummy variable named Independent, which is calculated from a set of relation variables in the IPUMS-CPS website that is equal to 0 if the individual lives at home and is equal to 1 if he is self-sufficient.

Hypothesized Results

For the main variable in question, which is the dummy variable named Independent; I hypothesize a pretty large significant positive number for its coefficient. I predict that the coefficients value will be considerably large enough to warrant a conclusion that parental income has a major effect on and individuals future income. Empirical Testing

Data Description The data collected is based on a 2008 nationwide population survey from the IPUMS-CPS data website. Below is a description of the data:

Descriptive Statistics

A. Continuous Variables Min Wage and Salary Income Age Weeks worked last year Usual hours worked per week (last yr) $1 15 1 1 Max $706,117 85 52 99 Mean $40,818 40.27 46.56 38.89 Standard Deviation $47,739 13.68 12.06 11.89

B. Dummy Variables Percent Educational Attainment Dropout Highsch Some College College Postgrad Region West Northeast Midwest South Race/Ethnicity White Latino Afamer Asianam Otherace Sex Female Male Parental Aid Independent Percent

12.44 28.37 19.55 29.30 10.34 22.94 18.49 22.84 35.73

65.48 15.34 10.75 5.31 3.12 48.74 51.26 12.39

The average salary in the data set is $40,818, the average working age is 40 years old, and the average work week was about 52 hours. Under the variables for education the highest percentage of people falls under the category of having a college degree with 29.30% of the people falling under that category. The regional variables had the most people coming from the south with 35.73%, and the least from the northeast with 18.49 percent. The largest ethnic group in the race category is the whites with 65.48% and other races being 3.12 %. The sex variables are almost evenly split with the males just a bit over at 51.26% of the sample.

Regression Results
Regression Results Table Model 1 Coefficient (t-stat) Constant Age Weeks Worked Hours worked Race Ethnicity African American Hispanic Asian American Other Racial/Ethnic Gender Female Region Northeast Midwest South Educational Attainment Less Than Highs School High School Four Year College Degree Post Graduate Degree Parental Aid Independent Adjusted R Squared F Statistic Sample Size 0.585 18,863,536.706 101,536 0.636 17,152,886.348 101,536 -0.032 -165.212 -0.134 -730.199 -0.120 -725.414 -0.045 -251.194 -0.103 -598.366 -0.103 -668.327 -0.745 -3291.686 -0.461 -2801.361 -0.354 -2153.462 0.203 941.145 -0.039 -221.271 -0.107 -626.382 -0.106 -693.405 -0.699 -3092.194 -0.446 -2735.953 -0.335 -2049.656 0.204 956.388 -0.339 -1672.199 0.643 16,561,095.283 101,536 -0.173 -1385.170 -0.214 -1825.769 -0.227 -1949.036 6.169 17372.213 0.011 2434.840 0.048 8610.547 0.038 6556.100 -0.142 -734.686 -0.277 -1528.596 0.053 184.613 -0.126 -294.725 Model 2 Coefficient (t-stat) 6.741 18415.795 0.009 2154.586 0.046 8740.324 0.035 6326.909 -0.074 -405.479 -0.092 -515.870 -0.039 -145.539 -0.066 -164.524 Model 3 Coefficient (t-stat) 7.005 17708.854 0.006 1373.125 0.045 8536.137 0.033 6029.476 -0.078 -433.786 -0.118 -668.805 -0.047 -174.677 -0.056 -141.503

Interpretation

To ascertain the effects of each of the added variables three separate regression were calculated. In the first model we estimated a model considered to be the baseline in order to measure the effects of our other hypothesized variables. In the first model age, races, gender, region, hours and weeks worked were used as control variables to estimate a regression line with an adjusted R-squared value of 0.585 on a sample of 101,536 individuals. A fairly high value of R-squared means that about 58.5% of the variability can be explained by the model, giving us a baseline as we add in the hypothesized variables. The extremely large value of the F-statistic of 18,863,536.706 indicates that the factors in this model are highly significant in determining the dependent variable.

In the second model the education variable has been added in to explain the effect of education on earnings. We have come up with a more statistically significant result as the value of R-square his increased to a 0.636, meaning that education creates a much better explanation for the increase in earning, as 63.6% of the variability in the data can be explained by the model. For those who have less than a high school degree, they make around -74% less than then their cohorts with at least some college. Having a high school degree decreases that negative effect on earnings, with the members of this group earning -46% less than their counterparts. Having a college degree is better by roughly 11 percentage points than having a high school diploma. Post Graduate degree holders make on average 20.3% more than their counterparts with some college.

In the third model with the included hypothesized variable estimating the amount of parental aid through a relation variable, the adjusted R-square value raised up to 0.643, meaning that an increase of . 7%in the variability can be explained by the data. The F-statistic of 16,561,095.283 shows an improvement over model two, so the estimated line is a better fit of the data. The coefficient for the variable named independent is a -0.339, meaning that being independent has a -33.9% effect on earnings. The addition of this variable shows omitted variable bias, but its most significant effect is on those with less than a high school degree.

Conclusion

This article has explored the role of parental aid in determining the transfer of wealth along generations by alleviating the burden of being independent and making use of the parents resources through staying at home. Human capital theory explains the increase in productivity of the worker, making them more valuable to employers. While the amount of parental aid received increases the gains of the effects of education by providing individuals the opportunity to concentrate in education. Though it is possible that there are many other factors involved in making this determination, the rise in income through education, supplemented by parental aid can explain a majority of these cases. The results are a positive insight into this particular phenomenon.

References: Blanden, J., Gregg P., Macmillan L. 2007. Accounting for Intergenerational Income Persistence: Noncognitive Skills, Ability and Education. The Economic Journal, 117 (March), C43-C60. The Royal Economic Society 2007. Blackwell Synergy Publisihing. Maurer-Fazio, M., Dinh N. 2004. Differential Rewards To, And Contributions Of, Education in Urban Chinas Segmented Labor Markets. Pacific Economic Review, 9: 3 (2004), pp.173-189. Blackwell Synergy Publishing. Behrman, Jere R., and Rosenzweig, Mark R. 2005. Parental Wealth and Adult Childrens welfare in Marriage. The Revue of Economics and Statistics, August 2006, 88(3):496-509. Liu, Haoming, and Zeng Jinli. 2007. Genetic Ability and Intergenerational Earnings Ability. Journal of Popular Economics (2009) 22:75-79.