Wow! 46.3 Million People in U.S. Receive Food Stamps!
The following graph, depicting 46.3 million people in the U.S. receiving food stamps, motivated my interest to identify odds and probabilities of receiving food stamps.
To identify the odds and probabilities of receiving food stamps I utilized the 2010 IPUMs-USA dataset. The following table depicts the variables in the analysis. Only individuals greater than 18 years old were included in the analysis. Sample size equals 2,326,721 individuals.
The dependent variable, a binary variable identifying food stamp recipients and non-recipients, is regressed on labor force participation, educational achievement, white race, marital status, gender and age. The logistic regression model is statistically significant as well as all regression coefficients ( p = 0.000).
Findings
Odds and Probabilities
The following represents a summary of the findings. In each summary statement the implied statement is that the outcome occurs with all other variables in the analysis controlled.
Labor Force Participation
Being in the labor force decreases the odds of being a recipient of food stamps by 47%. Individuals in the labor force have a .35 probability of being a recipient of food stamps.
High School Diploma or More
Individuals without a high school diploma increase their odds of receiving food stamps by a factor of 3.36, compared to individuals with a high school diploma or higher educational achievement. Individuals with a high school diploma or greater level of educational achievement have a probability of of being a recipient of food stamps equal to .23.
‘White’
Being white reduces the odds of being on food stamps by 47.6%, compared to all other races. Individuals identified as ‘white’ have a .34 probability of receiving food stamps.
Marital Status
- Compared to married individuals with a spouse present, individuals who are ‘Married, spouse absent’ are 88% more likely to receive food stamps. Probability equals .65.
- Compared to married individuals with a spouse present, individuals who are ‘Separated’ are 373% more likely to receive food stamps. Probability equals .83.
- Compared to married individuals with a spouse present, individuals who are ‘Divorced’ are 231% more likely to receive food stamps. Probability equals .77.
- Compared to married individuals with a spouse present, individuals who are ‘Widowed’ are 101% more likely to receive food stamps. Probability equals .67.
- Compared to married individuals with a spouse present, individuals who are ‘Never married/single’ are 70% more likely to receive food stamps. Probability equals .63.
Gender
Compared to females, males are 24.3% less likely to receive food stamps. Males have a .43 probability of receiving food stamps.
Age
Each additional year of age decreases the odds of receiving food stamps by 3%. The elasticity of ‘age’ is – 1.33, meaning that a 1% change in age is associated with a 1.33% decrease in food stamps. The age semi-elasticity equals -.027, meaning a 1 year increase in age yields a 2.7% decrease in food stamps recipients.
Predictive Margins Vary By Age
It should be noted that the predicted probability margins vary be age. For example, here’s a graph of the predictive margins of educational achievement by age and gender.
It’s clear that individuals with ‘No high school diploma’, especially females, increase the probability of being food stamp recipients. For each category in the above graph the probabilities of being a recipient of food stamps decrease with age.
Reference
Ruggles, J. Steven, Alexander, Trent, Genadek, Katie, Goeken, Ronald, Schroeder, Matthew B., and Matthew Sobek. (2010). Integrated Public Use Microdata Series: Version 5.0 [Machine-readable database]. Minneapolis: University of Minnesota.