GLM Welfare Spending Effect on Crime Rates
Project Overview
Over the last 60 years, welfare spending has increased at rates that taxpayers may be concerned about how much of the money spent is advantageous for the community both at a federal and local level. This report uses FBI crime data and city spending data over the last nine years. The project goal was to model the relationship between non-violent crimes and welfare spending in cities with populations over 100,000 while comparing their crime rates in those cities from FBI crime data. Our model results show that welfare spending does not have a statistically significant effect on non-violent crime.
Problem Definition & Significance
Our target audience is municipalities and their city council members. Cities have a limited budget every year, and to take on new initiatives, they either have to optimize their current spending or increase revenue by taxation. Cities like to keep in favor with their current residents; no matter what party one belongs to, either Republican or Democratic, no one is a massive fan of taxation. Thus it is essential to optimize spending to free up capital for new initiatives. Every city is interested in learning how spending over time affects its goals. To be more specific, every city wants to know the marginal effect of certain areas of spending on specific key city initiatives. There are three types of cities tier-1, tier-2, and tier-3 cities, and they all compete against each other within their respective tier. For a city to increase its attractiveness to the desired residents, it must perform well on three metrics - crime rate, cost of living, and high-paying jobs. An ideal situation for a future resident is high-paying jobs with low crime and below-average living expenses. High-paying jobs and increased cost of living are correlated. The only metric a city can have a marginal effect on is the crime rate regarding increasing city's attractiveness to prospects.
Looking at violent crimes over the years creates somewhat of a U-shape where 2013 and 2014 were the lowest years, but 2016 was extremely high. It even has an outlier whereby 2016 had the highest violent crime rate.
Non-violent crimes with the expenditure to see if there is any relation. There does not seem to be any relationship. However, there is a fanning shape.
Looking at the scatter plot to see if there is any relation between expenditure and crimes. It does not seem to have any homoscedasticity; it has a fanning shape.Looking at the scatter plot to see if there is any relation between expenditure and crimes. It does not seem to have any homoscedasticity; it has a fanning shape.
Looking at violent crimes over the years creates somewhat of a U-shape where 2013 and 2014 were the lowest years, but 2016 was extremely high. It even has an outlier whereby 2016 had the highest violent crime rate.
Model Quality Checks
Recommendations
Our recommendation is to re-evaluate the relationship between city crime rates for nonviolent crimes and welfare spending programs. Our model results show a nonsignificant relationship between nonviolent crimes and welfare spending programs. This model result could be because the welfare amount is not significant enough to reduce crime, as crime is still needed to supplement household income. Cities could increase allocated amounts of welfare to decrease the reliance on crime for additional household income.