I have been academically honest in all of my work and will not tolerate academic dishonesty of others, consistent with UGA’s Academic Honesty Policy.

Sign the academic honesty statement by typing your name on the Signature line.

Dorian Cushing:

We will not accept submissions that omit a signed Academic Honesty statement.

In this project, I look at earnings differences between full-time working men and women using data from the March 2024 CPS. The data set includes earnings, education, demographics, and household characteristics regarding this social group. I find an evidently clear gender wage gap that actually gets bigger over the course of a career and in controlling for things like education and demographics, the gap shrinks, but definitely doesn’t leave.

The March 2024 CPS ASEC surveys ~100,000 households. The ASEC includes data on annual earnings, total income, work experience, and health insurance coverage from the previous calendar year, making it especially useful for analyzing income differences like the gender wage gap.

cpsmar_e <- read_csv(here("data", "cpsmar_e.csv")) 

The extract was created using the script cpsmar_e.R. It selects key variables from the person file, including age, earnings, hours, weeks, race, gender, marital status, and education. New variables are created including indicators for gender, Hispanic origin, full-time status, union membership, and education levels. This is then restricted to full-time workers using filter. Overall contains 50,388 obs. and 20 variables.

cpsmar_a <- cpsmar_e %>%
filter(
age >= 23,
age <= 62,
earnings > 0
) %>%
mutate(
gender = case_when(female == 1 ~ "Female", TRUE ~ "Male"),
wage = earnings / (weeks * hours),
lwage = log(wage),
Black = case_when(race == 2 ~ 1, TRUE ~ 0),
south = case_when(region == 3 ~ 1, TRUE ~ 0),
married = case_when((marital == 1 | marital == 2 | marital == 3) ~ 1, TRUE ~ 0),
age_centered = age - 23
)

The CPS extract is narrowed to individuals between ages men and women: 23 to 62 with positive earnings. Final sample contains 44,614 observations.

figure1 <- ggplot(cpsmar_a, aes(x = earnings, group = gender, fill = gender)) +
  geom_density(alpha = 0.4) +
  labs(
    title="Figure 1. Distribution of earnings by gender",
    x="Earnings",
    y="Density"
    )+
  theme_minimal()

earnings_fvm <- cpsmar_a %>%
  group_by(gender) %>%
  summarize(avg_earnings = round(mean(earnings, na.rm = TRUE), 0))

avg_earnings_f <- earnings_fvm %>% 
  filter(gender == "Female") %>% 
  pull(avg_earnings)

avg_earnings_m <- earnings_fvm %>% 
  filter(gender == "Male") %>% 
  pull(avg_earnings)

Figure 1 shows that the earnings distribution for men is shifted to the right relative to women, indicating higher earnings on average. Average earnings for women are $75,026, compared to $94,269 for men. This implies a difference of $19,243, which is about 25.6% higher for men. This provides clear evidence of a gender earnings gap in the sample.

Table 1 shows that men earn higher log wages than women on average. Men also work more hours per week than women on average. Both differences are consistent with a meaningful gender gap in compensation.

cef_fvm_w <- cpsmar_a %>%
  group_by(gender, age_centered) %>%
  summarize(avg_lwage = mean(lwage, na.rm = TRUE))

figure2 <- ggplot(cef_fvm_w, aes(x = age_centered, y = avg_lwage, color = gender, linetype = gender, linewidth = gender)) +
  geom_point() +
  geom_line() +
  scale_linetype_manual(values = c("Female" = "longdash", "Male" = "solid")) + 
  scale_linewidth_manual(values = c("Female" = 0.7, "Male" = 0.5)) + 
  guides(linewidth = "none") +
  labs(
    title="Figure 2. Career log-wage profiles for women and men",
    x="Year",
    y="Average log wage"
    )+
  theme_minimal()

males <- cef_fvm_w %>%
  filter(gender == "Male") %>%
  rename(avg_lwage_male = avg_lwage) %>%
  select(!gender) 

females <- cef_fvm_w %>%
  filter(gender == "Female") %>%
  rename(avg_lwage_female = avg_lwage) %>%
  select(!gender)

diff_fvm <- inner_join(males, females, by = "age_centered") %>%
  filter(age_centered <= 30) %>%
  mutate(
    diff = avg_lwage_male - avg_lwage_female,
    age_group = cut(
      age_centered, 
      breaks = c(-1, 10, 20, 30), 
      labels = c("1-10", "11-20", "21-30"))
    ) %>%
  group_by(age_group) %>%
  summarize(mean_diff = mean(diff) * 100) 
  
table2 <- kable(
  diff_fvm,
  digits = 2,
  col.names = c("Year Range", "Avg Pct Difference"),
  align = "cc",
  caption = "Table 2. Percent wage differences, first 30 years",
  ) %>%
  kable_styling(position = "center")

The log wage gap grows over the first years of a career. The average percentage gap is smallest in years 1-10 and largest in years 21-30. The gender wage gap isn’t just starting salaries, but also career progression.

formula <- avg_lwage ~ age_centered + I(age_centered^2)

figure3 <- figure2 +  
  geom_smooth(
    method = "lm", 
    formula = y ~ x + I(x^2), 
    aes(group = gender), 
    se = FALSE
  ) +
  stat_poly_eq(
    aes(label = after_stat(eq.label)),
    formula = y ~ x + I(x^2),
    parse = TRUE
  ) +
  labs(
    title="Figure 3. Career log-wage profiles with quadratic fits",
    x="Year",
    y="Average log wage"
  ) +
  theme_minimal() +
  theme(legend.position = "bottom")

Table 3 shows that women in the sample are more likely than men to have a college degree and less likely to have only a high school diploma. Table 4 highlights that men in the sample are more likely to be Black, live in the South, and to be union members, while women are more likely to live in cities. These differences in observable characteristics could help explain part of the gender wage gap, but they do not seem to fully account for it.

singles <- cpsmar_a %>%
  filter(
    married == 0,
    child_u6 == 0
    )

models <- list(
  "Baseline"      = lm(lwage ~ female +
                       age_centered + I(age_centered^2),
                       cpsmar_a),
  "Add Education" = lm(lwage ~ female +
                       age_centered + I(age_centered^2) + HSGrad + SomeColl + CollDeg,
                       cpsmar_a),
  "Add Person"    = lm(lwage ~ female +
                       age_centered + I(age_centered^2) + HSGrad + SomeColl + CollDeg +
                       Black + hisp + south + city,
                       cpsmar_a),
  "Add Household" = lm(lwage ~ female +
                       age_centered + I(age_centered^2) + HSGrad + SomeColl + CollDeg +
                       Black + hisp + south + city +
                       married + child_u6,
                       cpsmar_a),
  "Only Singles"  = lm(lwage ~ female +
                       age_centered + I(age_centered^2) + HSGrad + SomeColl + CollDeg +
                       Black + hisp + south + city, 
                       singles)
)

cm <- c(
  'female'            = 'Female',
  'age_centered'      = 'Age',
  'I(age_centered^2)' = 'Age$^2$',
  '(Intercept)'       = 'Constant'
)
gm <- tibble::tribble(
  ~raw, ~clean, ~fmt,
  "nobs", "$N$", 0,
  "r.squared", "$R^2$", 2
)
rows <- tribble(~term, ~Baseline, ~Add_Education, ~Add_Person, ~Add_Household, ~Only_Singles,
  'Education controls',   ' ',         'X',           'X',          'X',           'X',
  'Demographic controls', ' ',         ' ',           'X',          'X',           'X',
  'Household controls',   ' ',         ' ',           ' ',          'X',           'X'
)
attr(rows, 'position') <- c(9, 10, 11)

table5 <- modelsummary(
  models,
  add_rows = rows,
  coef_map = cm,
  gof_map = gm,
  vcov = c("HC1", "HC1", "HC1", "HC1", "HC1"),
  title = "Table 5. OLS estimates of the gender wage gap",
  notes = "Heteroskedasticity-robust standard errors in parentheses.",
  escape = FALSE
  )

In the baseline model, the female coefficient is -0.165, implying women earn about ~16.5% less on average. Adding education controls increases the gap to 23.9%. Demographics slightly reduce the gap to 23.0%, and including household characteristics reduces it to 21.8%. Restricting to singles without young children decreases it to 13.7%. Observable characteristics explain some of the gap, but a substantial difference persists.

The CPS data on M/F workers from ages 23 to 62 shows a clear gender wage gap, with men earning more on average. The gap increases from about 8% early in careers to over 20% later on. Controlling for education and demographic characteristics reduces the gap but does not eliminate it. Regression results show that women earn between 13% and 24% less, even after accounting for observable differences. These findings suggest that the gender wage gap is driven by observable and unobservable factors.

This is a list of the main variables used in this project with their definitions.