adam sutton chia-jung liu grant volk yin chu can shen robert matarazzo andrew ratcliffe ruben bos
TRANSCRIPT
Adam Sutton
Chia-Jung Liu
Grant Volk
Yin Chu
Can Shen
Robert Matarazzo
Andrew Ratcliffe
Ruben Bos
Highest since Great Depression Effects well being of population Effects well being of economy Implicit effects on policy development and
reformation
To understand the correlation, if any, between unemployment rates and population characteristics
State Expenditures Percentage of population that is White, African-
American, and Hispanic Percentage of population with no health insurance Percentage of temporary jobs Percentage population over 25 with a bachelor’s
degree Income per capita Homeless per 1000 Number of citizens participating in the food stamps
program Crimes per 100,000 citizens Percentage of blue collar jobs Data taken across 51 observations including 50
states and the District of Columbia
Variable (1) Variable (2) R-squared F-Statistics PROB
Percent bachelors Income per capita 0.694573 0.000000
Percent blue collar Percent bachelors 0.747881 0.000000
Blue collar Income per capita 0.489143 0.000000
Food stamps Crime per 100,000 0.047541 0.124279
Food stamps State expenditures 0.771099 0.000000
Percent without health care Percent white 0.207299 0.000788
Percent without health care Percent black 0.043089 0.143844
Percent without health care Percent Hispanic 0.333601 0.000009
Crime per 100,000 per state Percent white 0.387852 0.000001
Crime per 100,000 per state Percent black 0.448505 0.000000
Crime per 100,000 per state Percent Hispanic 0.078482 0.046468
Percent without health care Crime rate per 100,000 people
0.111541 0.016613
Correlations between two signs of wealth Bachelors degree and income per capita
Correlations between two signs of poverty Lack of health care and food stamp
participants
Lack of significant negative correlations between a sign of wealth and a sign of poverty
Dependent Variable
Explanatory Variable
R-squared F-Statistics PROB
Unemployment Rate Percent of jobs that are blue-collar
0.040628 0.156083
Unemployment Rate Crime per 100,000 per state
0.200999 0.000969
Unemployment Rate Participation in food stamps program
0.185520 0.001604
Unemployment Rate Homeless per 1000 per state
0.146512 0.005568
Unemployment Rate Income per capita 0.007179 0.554416
Unemployment Rate Percent of adults with bachelors
0.004037 0.657805
Unemployment Rate Percent of African Americans per state
0.144186 0.005991
Unemployment Rate Percent of Hispanics per state
0.034391 0.192619
Unemployment Rate Percent of whites per state
0.094708 0.028029
Unemployment Rate Percent of pop with no health care
0.025767 0.260486
Unemployment Rate Percent of jobs that are temp
0.005882 0.592693
Unemployment Rate State expenditures 0.147193 0.005953
Percentage of population that is White and Hispanic
Percentage of population with no health insurance
Percentage of temporary jobs Percentage population over 25 with a
bachelor’s degree Income per capita Percentage of blue collar jobs
Dependent Variable
Explanatory Variable
R-squared F-Statistics PROB
Unemployment Rate Percent of jobs that are blue-collar
0.040628 0.156083
Unemployment Rate Crime per 100,000 per state
0.200999 0.000969
Unemployment Rate Participation in food stamps program
0.185520 0.001604
Unemployment Rate Homeless per 1000 per state
0.146512 0.005568
Unemployment Rate Income per capita 0.007179 0.554416
Unemployment Rate Percent of adults with bachelors
0.004037 0.657805
Unemployment Rate Percent of African Americans per state
0.144186 0.005991
Unemployment Rate Percent of Hispanics per state
0.034391 0.192619
Unemployment Rate Percent of whites per state
0.094708 0.028029
Unemployment Rate Percent of pop with no health care
0.025767 0.260486
Unemployment Rate Percent of jobs that are temp
0.005882 0.592693
Unemployment Rate State expenditures 0.147193 0.005953
Dependent Variable
Explanatory Variable
R-squared F-Statistics PROB
Unemployment Rate Percent of jobs that are blue-collar
0.040628 0.156083
Unemployment Rate Crime per 100,000 per state
0.200999 0.000969
Unemployment Rate Participation in food stamps program
0.185520 0.001604
Unemployment Rate Homeless per 1000 per state
0.146512 0.005568
Unemployment Rate Income per capita 0.007179 0.554416
Unemployment Rate Percent of adults with bachelors
0.004037 0.657805
Unemployment Rate Percent of African Americans per state
0.144186 0.005991
Unemployment Rate Percent of Hispanics per state
0.034391 0.192619
Unemployment Rate Percent of whites per state
0.094708 0.028029
Unemployment Rate Percent of pop with no health care
0.025767 0.260486
Unemployment Rate Percent of jobs that are temp
0.005882 0.592693
Unemployment Rate State expenditures 0.147193 0.005953
Dependent Variable
Explanatory Variable
R-squared F-Statistics PROB
Unemployment Rate Percent of jobs that are blue-collar
0.040628 0.156083
Unemployment Rate Crime per 100,000 per state
0.200999 0.000969
Unemployment Rate Participation in food stamps program
0.185520 0.001604
Unemployment Rate Homeless per 1000 per state
0.146512 0.005568
Unemployment Rate Income per capita 0.007179 0.554416
Unemployment Rate Percent of adults with bachelors
0.004037 0.657805
Unemployment Rate Percent of African Americans per state
0.144186 0.005991
Unemployment Rate Percent of Hispanics per state
0.034391 0.192619
Unemployment Rate Percent of whites per state
0.094708 0.028029
Unemployment Rate Percent of pop with no health care
0.025767 0.260486
Unemployment Rate Percent of jobs that are temp
0.005882 0.592693
Unemployment Rate State expenditures 0.147193 0.005953
Dependent Variable
Explanatory Variable
R-squared F-Statistics PROB
Unemployment Rate Percent of jobs that are blue-collar
0.040628 0.156083
Unemployment Rate Crime per 100,000 per state
0.200999 0.000969
Unemployment Rate Participation in food stamps program
0.185520 0.001604
Unemployment Rate Homeless per 1000 per state
0.146512 0.005568
Unemployment Rate Income per capita 0.007179 0.554416
Unemployment Rate Percent of adults with bachelors
0.004037 0.657805
Unemployment Rate Percent of African Americans per state
0.144186 0.005991
Unemployment Rate Percent of Hispanics per state
0.034391 0.192619
Unemployment Rate Percent of whites per state
0.094708 0.028029
Unemployment Rate Percent of pop with no health care
0.025767 0.260486
Unemployment Rate Percent of jobs that are temp
0.005882 0.592693
Unemployment Rate State expenditures 0.147193 0.005953
Dependent Variable
Explanatory Variable
R-squared F-Statistics PROB
Unemployment Rate Percent of jobs that are blue-collar
0.040628 0.156083
Unemployment Rate Crime per 100,000 per state
0.200999 0.000969
Unemployment Rate Participation in food stamps program
0.185520 0.001604
Unemployment Rate Homeless per 1000 per state
0.146512 0.005568
Unemployment Rate Income per capita 0.007179 0.554416
Unemployment Rate Percent of adults with bachelors
0.004037 0.657805
Unemployment Rate Percent of African Americans per state
0.144186 0.005991
Unemployment Rate Percent of Hispanics per state
0.034391 0.192619
Unemployment Rate Percent of whites per state
0.094708 0.028029
Unemployment Rate Percent of pop with no health care
0.025767 0.260486
Unemployment Rate Percent of jobs that are temp
0.005882 0.592693
Unemployment Rate State expenditures 0.147193 0.005953
State Expenditures Percentage of population that is African-
American Homeless per 1000 Number of citizens participating in the
food stamps program Crimes per 100,000 citizens
State expenditures and food stamp participation In order to study the variables related to
state spending against unemploymentDetermined that food stamp participation is
the most significant variable while state expenditures is not
Dependent Variable: UNEMP_RATE Method: Least Squares Date: 12/01/09 Time: 14:25 Sample: 1901 1951 Included observations: 50 Excluded observations: 1
Variable Coefficient Std. Error t-Statistic Prob.
STATE_EXP -7.77E-06 1.96E-05 -0.396859 0.6933 FOOD_STAMPS 1.72E-06 8.29E-07 2.073140 0.0437
C 7.832522 0.407456 19.22297 0.0000
R-squared 0.218644 Mean dependent var 8.858000 Adjusted R-squared 0.185395 S.D. dependent var 2.240635 S.E. of regression 2.022296 Akaike info criterion 4.304468 Sum squared resid 192.2150 Schwarz criterion 4.419190 Log likelihood -104.6117 F-statistic 6.575919 Durbin-Watson stat 2.404541 Prob(F-statistic) 0.003034
0
2
4
6
8
-4 -3 -2 -1 0 1 2 3 4 5
Series: ResidualsSample 1901 1951Observations 50
Mean -1.19E-15Median 0.052733Maximum 4.919962Minimum -4.123167Std. Dev. 1.980595Skewness 0.523202Kurtosis 3.576950
Jarque-Bera 2.974647Probability 0.225977
-6
-4
-2
0
2
4
6
7 8 9 10 11 12 13
FITTEDEXPFOOD
RE
SID
EX
PF
OO
DNormal residualsNo heteroskedasticity
Crime rate and income per capita In order to study the effects of poverty and
unemploymentHighly significant regression did existCrime rate is highly significant while income
is not Introduction of dummy variable for DC
DC’s data was much higher than the 50 states
R-squared value increased because the absence of DC’s data decreased the sum of residuals
Dependent Variable: UNEMP_RATE Method: Least Squares Date: 12/02/09 Time: 14:06 Sample: 1901 1951 Included observations: 51
Variable Coefficient Std. Error t-Statistic Prob.
CRIME_PER_100000 0.005314 0.001599 3.323136 0.0017 INC_PER_CAP -1.26E-05 5.35E-05 -0.235172 0.8151
DUMMYDC -2.494080 2.984445 -0.835693 0.4076 C 7.146291 2.092222 3.415647 0.0013
R-squared 0.219285 Mean dependent var 8.917647 Adjusted R-squared 0.169452 S.D. dependent var 2.258646 S.E. of regression 2.058404 Akaike info criterion 4.356924 Sum squared resid 199.1403 Schwarz criterion 4.508439 Log likelihood -107.1016 F-statistic 4.400400 Durbin-Watson stat 2.622396 Prob(F-statistic) 0.008272
Introduction of dummy variable for DCDC’s data was much higher than the 50
statesR-squared value increased because the
absence of DC’s data decreased the sum of residuals
-4
-2
0
2
4
6
6 8 10 12 14
FITTEDCRIMEINC
RE
SID
CR
IME
INC
0
1
2
3
4
5
6
-3 -2 -1 0 1 2 3 4 5 6
Series: ResidualsSample 1901 1951Observations 51
Mean -2.56E-16Median -0.076828Maximum 5.509507Minimum -3.422640Std. Dev. 2.010469Skewness 0.547344Kurtosis 3.090094
Jarque-Bera 2.563729Probability 0.277519
Normal residualsNo heteroskedasticity
Crime rate, food stamps, and homeless per 1000
In order to test the three most highly significant variables against unemployment
Determined that a significant relationship exists with participation in the food stamps program being the most significant variable
Dependent Variable: UNEMP_RATE Method: Least Squares Date: 12/01/09 Time: 14:49 Sample: 1901 1951 Included observations: 51
Variable Coefficient Std. Error t-Statistic Prob.
HOME_PER_1000 0.303980 0.197584 1.538488 0.1306 FOOD_STAMPS 1.11E-06 3.75E-07 2.959184 0.0048
CRIME_PER_100000 0.002413 0.001352 1.785036 0.0807 C 6.483981 0.577719 11.22343 0.0000
R-squared 0.350128 Mean dependent var 8.917647 Adjusted R-squared 0.308646 S.D. dependent var 2.258646 S.E. of regression 1.878012 Akaike info criterion 4.173489 Sum squared resid 165.7656 Schwarz criterion 4.325005 Log likelihood -102.4240 F-statistic 8.440629 Durbin-Watson stat 2.361724 Prob(F-statistic) 0.000136
Normal Residuals No heteroskedasticity
0
2
4
6
8
10
-3 -2 -1 0 1 2 3 4 5
Series: ResidualsSample 1901 1951Observations 51
Mean -2.49E-16Median -0.093738Maximum 5.403755Minimum -3.316906Std. Dev. 1.820800Skewness 0.658761Kurtosis 3.650791
Jarque-Bera 4.588710Probability 0.100826 -4
-2
0
2
4
6
6 8 10 12 14
HCSFITTED
HC
SR
ES
ID
Introduction of dummy variable for DC Increased R2 by decreasing sum of
residualsDependent Variable: UNEMP_RATE Method: Least Squares Date: 12/01/09 Time: 15:04 Sample: 1901 1951 Included observations: 51
Variable Coefficient Std. Error t-Statistic Prob.
HOME_PER_1000 0.527074 0.251541 2.095379 0.0417 FOOD_STAMPS 8.76E-07 4.06E-07 2.155693 0.0364
CRIME_PER_100000 0.003564 0.001567 2.274129 0.0277 DUMMYDC -4.754245 3.371999 -1.409919 0.1653
C 5.807296 0.746482 7.779548 0.0000
R-squared 0.377048 Mean dependent var 8.917647 Adjusted R-squared 0.322878 S.D. dependent var 2.258646 S.E. of regression 1.858581 Akaike info criterion 4.170398 Sum squared resid 158.8989 Schwarz criterion 4.359793 Log likelihood -101.3451 F-statistic 6.960497 Durbin-Watson stat 2.339112 Prob(F-statistic) 0.000181
Percent of population with no health insurance does not correlate with unemployment rate
Percent of adults with bachelors does not correlate with unemployment rate
Unemployment positively correlated with:Participation in the food stamps programHomeless rate per 1000 peopleCrime rate per 100,000 people
Should be the goal of the government to decrease unemployment by a reallocation of fundsWould lead to decrease in crime, homeless
rate, and poverty
(End)