Table 7.
Linear models with CPUMA fixed effects using total opioid-overdose death rates to predict being employed, household income, and living in a household with income below the poverty line, by race-ethnicity, among never married less educated women and never married less educated men
| Women Aged 20-34 |
Men Aged 20-39 |
|||||
|---|---|---|---|---|---|---|
| White | Black | Hispanic | White | Black | Hispanic | |
|
|
|
|||||
| Opioid-Overdose Death Rate (per 100,000 pop.) in CPUMA (year y-3) | ||||||
| Employed | ||||||
| ODR | 0.0001 | 0.0001 | 0.0012 | −0.0003 | −0.0002 | 0.0013 |
| [0.0003] | [0.0008] | [0.0013] | [0.0003] | [0.0010] | [0.0011] | |
| CPUMA-level controls (year y-2) | Yes | Yes | Yes | Yes | Yes | Yes |
| CPUMA Fixed Effects | Yes | Yes | Yes | Yes | Yes | Yes |
| Year Fixed Effects | Yes | Yes | Yes | Yes | Yes | Yes |
| State*Year Fixed Effects | Yes | Yes | Yes | Yes | Yes | Yes |
| Household Income ꓕ | ||||||
| ODR | 2.9487 | −11.9756 | 37.3391 | −12.2588 | −101.578 | 49.7555 |
| [35.0447] | [72.5272] | [136.8263] | [32.1129] | [72.0312] | [123.2651] | |
| CPUMA-level controls (year y-2) | Yes | Yes | Yes | Yes | Yes | Yes |
| CPUMA Fixed Effects | Yes | Yes | Yes | Yes | Yes | Yes |
| Year Fixed Effects | Yes | Yes | Yes | Yes | Yes | Yes |
| State*Year Fixed Effects | Yes | Yes | Yes | Yes | Yes | Yes |
| Living in a Household with Income below Poverty Line | ||||||
| ODR | −0.0003 | 0.0006 | 0.0013 | −0.0001 | 0.0017 | 0.0004 |
| [0.0005] | [0.0010] | [0.0014] | [0.0003] | [0.0013] | [0.0010] | |
| CPUMA-level controls (year y-2) | Yes | Yes | Yes | Yes | Yes | Yes |
| CPUMA Fixed Effects | Yes | Yes | Yes | Yes | Yes | Yes |
| Year Fixed Effects | Yes | Yes | Yes | Yes | Yes | Yes |
| State*Year Fixed Effects | Yes | Yes | Yes | Yes | Yes | Yes |
| Observations | 462,970 | 117,398 | 130,849 | 777,434 | 146,507 | 202,143 |
|
| ||||||
Notes: Analyses were restricted to women aged 20-34 and men aged 20-39. Controls for individual characteristics include respondent’s age, number of own children living in the household (for women), and whether respondents were born outside the U.S. For less educated respondents, we also control for whether they had completed high school. All of these individual controls were measured at the time of interview (year y). Controls for sociodemographic characteristics at the CPUMA level include the percentage of the population that has less than high school, percentage of the population that has high school, average household income, the percentage population that is unemployed, the percentage that receives public assistance income, and the percentage Non-Hispanic black, Hispanic, and foreign born. We also included controls for the population density in the CPUMA (total population per square mile), and the sex ratio (number of men aged 25-39 per women aged 20-34). Local controls were measured in year y-2. Data obtained from the 1-year samples of the American Community Survey (Ruggles et al. 2018) and the National Vital Statistics System (NVSS 2018). Standard errors are clustered by CPUMA.
Household income given in 1999 constant dollars.
p<0.05
p<0.01
p<0.001.
Standard errors in brackets.