analysis

We have data on candidates who ran for elections in 2000s. We want to figure out how long they live for after the election. One reason why we might doubt the data is because the life expectancy of people changes over time. It seems that on average, women live longer than men after an election.

# A tibble: 6 × 14
  state    year first_name last_name party    sex   died       status win_margin
  <chr>   <int> <chr>      <chr>     <chr>    <chr> <date>     <chr>       <dbl>
1 Alabama  1946 James      Folsom    Democrat Male  1987-11-21 Chall…       77.3
2 Alabama  1946 Lyman      Ward      Republi… Male  1948-12-17 Chall…      -77.3
3 Alabama  1950 Gordon     Persons   Democrat Male  1965-05-29 Chall…       82.2
4 Alabama  1954 Tom        Abernethy Republi… Male  1968-03-07 Chall…      -46.7
5 Alabama  1954 James      Folsom    Democrat Male  1987-11-21 Chall…       46.7
6 Alabama  1958 William    Longshore Republi… Male  2013-05-24 Chall…      -77.5
# ℹ 5 more variables: region <chr>, population <dbl>, election_age <dbl>,
#   death_age <dbl>, lived_after <dbl>

\[\begin{aligned} livedafter &=& \beta_0 + \beta_1{male_i} + \epsilon_i \end{aligned}\]