2001 | 2009 | 2014 | |

10 | 14679.6 | 13986 | 13376 |

50 | 32935 | 32634 | 31680 |

90 | 72457 | 77700 | 77440 |

Question1.Measures on income inequality

Table1: Real income 10^{th}percentile (*P*_{10}),50^{th}percentile (*P*_{50}),90^{th}percentile (*P*_{90})

Table2: Logs of 10^{th}percentile (*P*_{10}),50^{th}percentile (*P*_{50}),90^{th}percentile (*P*_{90})

2001 | 2009 | 2014 | |

log( | 4.17 | 4.16 | 4.13 |

log( | 4.52 | 4.51 | 4.5 |

log( | 4.86 | 4.89 | 4.89 |

Table3: The three measures of income inequality across the three years

2001 | 2009 | 2014 | |

log( | 0.69 | 0.73 | 0.76 |

log( | 0.34 | 0.38 | 0.39 |

log( | 0.35 | 0.35 | 0.39 |

Table1 gives us the 10^{th}percentile, 50^{th}percentile and the 90^{th}percentile of income for the different years that is 2001, 2009, 2014and table 2 gives us the logs of the different percentiles in table 1for the three different years in the ACS waves.

Table3 gives us the three measures of income inequality

Forthe year 2001 the total income inequality of the population[log(*P*_{90})-log(*P*_{10})] was equal to 0.69 .This included the income inequality of the tophalf of the distribution [log(*P*_{90})-log(*P*_{50})]was equal to 0.34 and the income inequality of the bottom half of thedistribution [log(*P*_{50})-log(*P*_{10})]was 0.35.

Forthe year 2009 the total income inequality of the population[log(*P*_{90})- log(*P*_{10})] was equal to 0.73.This included the income inequality of the tophalf of the distribution [log(*P*_{90})-log(*P*_{50})]was equal to 0.38 and the income inequality of the bottom half of thedistribution [log(*P*_{50})-log(*P*_{10})]was 0.35.

Forthe year 2014 the total income inequality of the population[log(*P*_{90})-log(*P*_{10})] was equal to 0.76.This included the income inequality of the tophalf of the distribution [log(*P*_{90})-log(*P*_{50})]was equal to 0.39 and the income inequality of the bottom half of thedistribution [log(*P*_{50})-log(*P*_{10})]was 0.39.

Fromthe above we can be able to note that the 1^{st}measure of inequality, that is [log(*P*_{90})-log(*P*_{10})] increased through the years that is from 0.69 in 2001 to 0.73 in2009 and to 0.76 in 2014. The 2^{nd}measure of inequality, that is [log (*P*_{90})â€“ log (*P*_{50})]also increases through the ACS waves that is from 0.34 in 2001 to0.38 in 2009 and finally to 0.39 in in the years 2014. The 3^{rd}measure of income inequality, that is, [log (*P*_{50})-log(*P*_{10})] remained the same at 0.35 in 2001 and 2009 but increased to 0.39 in2014.

Question2: Measures of Wage inequality

Table4: Wage 10th percentile (*P*_{10}),50^{th}percentile (*P*_{50}),90^{th}percentile (*P*_{90})

2001 | 2009 | 2014 | |

10 | 6.77 | 6.57 | 6.09 |

50 | 14.38 | 14.94 | 13.81 |

90 | 30.76 | 33.62 | 33.85 |

Table5: Logs of 10^{th}percentile (*P*_{10}),50^{th}percentile (*P*_{50}),90^{th}percentile (*P*_{90})of wage

2001 | 2009 | 2014 | |

log( | 0.83 | 0.82 | 0.78 |

log( | 1.16 | 1.17 | 1.14 |

log( | 1.49 | 1.53 | 1.53 |

Table6: The three measures of wage inequality across the three years

2001 | 2009 | 2014 | |

log( | 0.66 | 0.71 | 0.75 |

log( | 0.33 | 0.36 | 0.39 |

log( | 0.33 | 0.35 | 0.36 |

Table4 gives us the 10^{th}percentile, 50^{th}percentile and the 90^{th}percentile of wage for the different years that is 2001, 2009, 2014and table 2 gives us the logs of the different percentiles in table 1 for the 3 different years in the ACS waves.

Table6 gives us the three measures of income inequality

Forthe year 2001 the total income inequality of the population[log(*P*_{90})-log(*P*_{10})] was equal to 0.66 .This included the income inequality of the tophalf of the distribution [log(*P*_{90})-log(*P*_{50})]was equal to 0.33 and the income inequality of the bottom half of thedistribution [log(*P*_{50})-log(*P*_{10})]was 0.33.

Forthe year 2009 the total income inequality of the population[log(*P*_{90})-log(*P*_{10})] was equal to 0.71.This included the income inequality of the tophalf of the distribution [log(*P*_{90})-log(*P*_{50})]was equal to 0.36 and the income inequality of the bottom half of thedistribution [log(*P*_{50})-log(*P*_{10})]was 0.35.

Forthe year 2014 the total income inequality of the population[log(*P*_{90})-log(*P*_{10})] was equal to 0.75.This included the income inequality of the tophalf of the distribution [log(*P*_{90})-log(*P*_{50})]was equal to 0.39 and the income inequality of the bottom half of thedistribution [log(*P*_{50})-log(*P*_{10})]was 0.36.

Fromthe above we can be able to note that the 1^{st}measure of inequality, that is, [log (*P*_{90})-log(*P*_{10})] increased through the years that is from 0.66 in 2001 to 0.71 in2009 and to 0.75 in 2014. The 2^{nd}measure of inequality, that is, [log (*P*_{90})â€“ log (*P*_{50})]also increases through the ACS waves that is from 0.33 in 2001 to0.36 in 2009 and finally to 0.39 in the years 2014. The 3^{rd}measure of income inequality, that is, [log (*P*_{50})- log(*P*_{10})] increased all through the years from 0.33 in 2001 to 0.35 in 2009and finally to 0.36 in 2014.

Question3: College Income premium

__Table7__

Log_income | 2001 coefficient | 2009 coefficients | 2014 coefficients |

College | 0.5712 | 0.6138 | 0.6182 |

constant | 10.2172 | 10.1932 | 10.1365 |

Fromtable 7 we can see that the difference in log income between collegegraduates and those without a college degree was 0.5712 in 2001,0.6138 in 2009 and 0.6182 in 2014.This shows a successive increase inthe difference across the years from year 2001 to the year 2014.

Thedifferences in log income between college graduates and those withouta college degree seem to be consistent with the trend of total incomeinequality [log (*P*_{90})- log (*P*_{10})]and the total wage inequality [log (*P*_{90})- log (*P*_{10})]of the distribution which also increased across the years from theyear 2001 to 2009. This shows that as income inequality and wageinequality increases across the years, the difference in log incomebetween college graduates and those without a college degree alsoincreases.

Thisregression might suffer from omitted variables bias because collegeonly accounts for a small percentage of the variance in log incomeacross the years, that is, 15.53% in the year 2001, 17.04% in 2009and 16.34% in 2014.

Question4: Free Response

__Regressingreal income and the total number of hours worked in a year__

Todetermine how the total number of hours worked in a year determinesthe income of an individual. I was motivated to ask this questionbecause I believe that the number of hours a person has worked in ayear plays a great role in determining the income of that particularindividual. I expect that the total number of hours worked willaffect the income of an individual positively(Pouwels, Siegers and Vlasblom 73).

Table8: Regression of real income and total number of hours worked for theyear 2001

Real income | Coef. | Std. Err. | t | P>|t| |

Workhoursperyear | 22.21638 | 0.1537379 | 144.5 | 0.000 |

_cons | -9419.334 | 360.2965 | -26.14 | 0.000 |

*F(1,354196)=20882.6 Prob>F=0.0000 R squared=0.0557 Adj R squared=0.0557*

Fromtable 8 In the year 2001 total hours worked per year significantlypredicted Real income, b=22.22, t (1)=144.5,P=0.000. Total hoursworked per year also explained a significant proportion of variancein Real income, R^{2}=0.0557,F(1,354196)=20882.6,P=0.000.

Thisindicates that a unit increase in the work hours per year lead to anincrease in real income of 22.22 dollars

Table9: Regression of real income and total number of hours worked for theyear 2009

Real income | Coef. | Std. Err. | t | P>|t| |

Workhoursperyear | 25.58535 | 0.1046317 | 244.53 | 0.000 |

_cons | -14767.12 | 242.7235 | -60.84 | 0.000 |

*F(1,894167)=59793.78 Prob>F=0.0000 R squared=0.0627 Adj R squared=0.0627*

Asindicated in table 9,In the year 2009 total hours worked per yearsignificantly predicted Real income, b=25.56,t(1)=244.53, P=0.000.Total hours worked per year also explained a significant proportionof variance in Real income, R^{2}=0.0627,F (1, 894167) = 59793.78, P=0.000.

Thisshows that a unit increase in the work hours per year lead to anincrease in real income by 25.58 dollars

Table10: Regression of real income and total number of hours worked forthe year 2014

Real income | Coef. | Std. Err. | t | P>|t| |

Workhoursperyear | 23.30393 | 0.1015499 | 229.48 | 0 |

_cons | -10873.05 | 236.9586 | -45.89 | 0 |

*F(1,908861)=52662.21 Prob>F=0.0000 R squared=0.0548 Adj R squared=0.0548*

Fromtable 10, in the year 2014 total hours worked per year significantlypredicted Real income, b=23.30, t (1) =229.48, P=0.000. Total hoursworked per year also explained a significant proportion of variancein Real income, R^{2}=0.0548,F (1, 908861) =52662.21, P=0.000. This shows that a unit increase inthe work hours per year lead to an increase in real income by 23.30dollars.

Generally,from the analysis above we can conclude the increase in real incomeas a result of an increase in one unit of work hours per year ishighest in 2009 followed by in the year 2014 and finally in 2001.

Theregression might suffer from omitted variable bias because in all thethree years. Work hours per year accounts only for a small portion ofvariance in real income, that is, 5.57% in 2001, 6.27% in 2009, andonly 5.58% in 2014.

WorksCited

Pouwels,Babette, Jacques Siegers, and Jan Dirk Vlasblom. "Income,working hours, and happiness." Economics letters 99.1 (2008):72-74.