## Tuesday, December 5, 2017

### Minimum Wage and Veterans at the State Level

In my last post on cities with living wage ordinances and the percentage of veterans in that city, the percent of the city who are veterans was negatively associated with the amount of the living wage that was passed.  For this post I thought I would take a look at what variables were associated with the amount of the minimum wage passed at the city or county level.

All but six states have their own minimum wage laws:  Alabama, Louisiana, Mississippi, New Hampshire, South Carolina, and Tennessee.   Georgia and Wyoming have minimum wages below the national minimum at \$5.15/hour.  Thirty states plus DC have minimums above the national minimum of \$7.25/hour.

First I looked at the univariate relationship between veteran percentages and the state minimums.  I coded the states without minimums as zero.  All of the zero states except Alabama had percentages of veterans above the national rate.  There was a weak negative correlation between these two variables of -0.214 which accounts for only 5% of the variability.  This relationship was not statistically significant at p=0.13.  The graph below shows the nature of the relationship.

Next I thought I would conduct an analysis of the 30 states plus DC with minimum wages above the federal minimum.  This was done to make it comparable to the analysis I conducted for the living wage cities and % veterans.  This time the univariate correlation was significantly negative with a value of -0.389 and a p-value of 0.03.  This relationship accounts for 15% of the variability in the state minimum wage.  The graph below shows this relationship.

The relationship at the state level is not as strong as it is for the living wage cities.  There veterans accounted for 28% of the variability.  To see which other variables might be at play I added the percents of the 30 states plus DC that voted for Trump.  When I did this the effect was no longer present for veterans but it was for Trump's % of the vote.  There was a fairly strong positive correlation of 0.55 for these states with suggests that states with higher veteran populations are more politically conservative.  For all states the correlation between veterans and Trump's % was slightly weaker at 0.46.

 Coefficients Standard Error t Stat P-value Intercept 15.71 1.84 8.52 0.00 Trump % -0.10 0.03 -2.91 0.01 % Veterans -0.15 0.31 -0.49 0.63

The District of Columbia was included in both the analyses for cities and states.  Removing DC did not radically alter the results.

**Related Posts**