Congress Is Aging Too, But It's Not Frozen
Congress Is Aging Too, But It's Not Frozen
In my last post, I showed that the presidency has been locked on a single generation since 1993. The birth-year slope flatlined. Is Congress doing the same thing?
Partially. Congress is aging, but 535 seats refresh faster than a single office. Members retire, lose primaries, die in office. The two chambers tell very different stories, though. The House has stabilized. The Senate hasn't.
Median Age Over Time
The House held remarkably steady from 1927 through the early 1980s, median age bobbing between 51 and 55 for over five decades. Then it dipped to a low of 49.7 in 1985 as a wave of younger members entered after the post-Watergate and Reagan-era elections. From that floor, it climbed steadily to about 59 by 2011, then plateaued. The ascent wasn't one dramatic election. It was a grinding ramp from roughly 1997 to 2007, gaining about a year of median age every two calendar years.
The Senate has a wilder story. It held in the 57-61 range from 1919 through the late 1960s, then dropped sharply, from 61 in 1959 down to 53 by 1983. That was the Watergate-era revolution. A wave of young senators replaced the old guard: Biden elected at 29 in 1972, Gary Hart at 37 in 1974. The generation that had held power since the New Deal got swept out.
Then the line reversed. From 53 in 1983 to 66.2 in 2025, a climb of 13 years in four decades. The 119th Congress Senate is at its highest median age in the entire dataset.
Piecewise regression puts the breakpoints at 1983 and 2007 for the House, and 1959 and 1987 for the Senate. The House broke its upward trend around 2007 and stabilized. The Senate broke its downward trend around 1987 and has been climbing ever since.
Generational Composition
Median age tells you something is happening. Generational composition tells you what.
House of Representatives
Senate
The boomer generation (purple) dominates both chambers. It peaked at 63% of the House in 2015 and 68% of the Senate in 2021. But the next generations are entering at very different rates.
In the House, Gen X (pink) has already reached 41% and Millennials (coral) are at 13%. The purple band is clearly past peak and narrowing steadily.
In the Senate, Gen X only just crossed 26% in 2025 and Millennials are at a mere 5%. The Senate is where the generational story from the presidential analysis is repeating: a single cohort holding disproportionate power for longer than the historical pattern would predict.
The Silent Generation (teal) is the baseline. They peaked at roughly 60% of the House around 1985-1989 and about 70% of the Senate around 1991-1993, then faded steadily. Normal generational turnover: a generation rises, dominates for 15-20 years, then yields. The question is whether boomers will follow the same curve or hold longer.
The Senate Is the One to Watch
The presidential analysis found a birth-year slope of 0.01 since 1993, effectively frozen. The conversion is straightforward: birth_slope = 1 - age_slope. When the median age rises one year per calendar year, birth years are frozen (slope 0). When median age stays flat, birth years advance at full speed (slope 1).
The Senate since 1987 has a birth-year slope of 0.73. Not frozen, but in the same neighborhood as the presidential second era (1901-1961, slope 0.69). Generational turnover is happening, but slowly. Each new Congress brings in members born only about 8-9 months later than the cohort they're replacing, when it should be closer to 12.
The House since 2007 has a birth-year slope of 0.97, nearly full generational turnover. Whatever was driving the age ramp from 1985 to 2007, it stopped. New members are entering at roughly the expected rate.
The Senate hasn't stabilized. Its median age of 66.2 is the highest in the dataset. Remember the 65-year ceiling that held for presidents from Washington through Carter? The Senate just broke through it too.
Methodology: Data from unitedstates/congress-legislators (public domain), covering the 66th through 119th Congresses (1919-2025). Age computed on the first day of each Congress. Piecewise linear regression fitted via dynamic programming; breakpoints selected by minimum AIC. Generational boundaries: Lost 1883-1900, Greatest 1901-1927, Silent 1928-1945, Boomer 1946-1964, Gen X 1965-1980, Millennial 1981-1996, Gen Z 1997+. Members with unknown birth dates excluded. See the previous post on presidential aging for the methodology on birth-year slope conversion.
Comments
Post a Comment