The Senate is older than the House, and it shows up in the completion odds. Senators average 65 years old and carry an 84% probability of completing their current term. House members average 58 years old, seven years younger, with a 96% completion probability. That is a 12-point gap in the odds of a member simply finishing the job voters sent them to do. Age alone does not explain every point of that gap, but it is the most visible difference between the two chambers, and the data lines up accordingly. Two legislative bodies, elected under different rules, on different cycles, with different constitutional minimum ages, produce two very different risk profiles once you actually run the numbers. This piece walks through what those numbers are, where they come from, and how a reader should actually use them when comparing the two chambers side by side.
The Senate holds 100 members. Average age: 65. Average completion probability: 84%. The House holds 437 members. Average age: 58. Average completion probability: 96%. Two chambers, two different risk profiles. The Senate's smaller size means each individual member's age and health carry more statistical weight on the average. The House, with more than four times the membership, spreads that risk across a much larger pool, so a handful of older members has less pull on the overall number. Fewer members in the Senate also means each seat is a larger share of the chamber's total power, which makes the completion-probability gap more consequential than the raw percentage might suggest at first glance. A six-year Senate term is also a longer window for anything to happen than a two-year House term, which is part of why the completion-probability model treats the two chambers separately rather than pooling them into one average. Put the two data sets side by side and the pattern is simple: fewer members, older average age, lower odds of a full term completed without interruption. More members, younger average age, higher odds of a full term completed. Neither chamber is uniform internally, but these are the current averages across every sitting member of each body, computed the same way for both.
Completion probability is not a prediction about any single person. It is a statistical measure, built from age and other factors, that estimates the likelihood a sitting member finishes their current term rather than leaving office early through retirement, resignation, or death. When you see the Senate's average sitting at 84% against the House's 96%, that is a signal about structural risk in how the chamber is currently composed, not a comment on any individual senator's fitness to serve. Readers using Too Damn Old should treat these averages as a starting point for comparison, not a countdown clock. The value of the number is in the comparison. It tells you the Senate, as a body, is carrying more age-related turnover risk than the House at this moment, and that has downstream effects on committee continuity, institutional knowledge, and how often voters may find themselves choosing a replacement mid-term rather than at a scheduled election. It also matters for expectations. A chamber with a lower average completion probability is a chamber more likely to see special elections, gubernatorial appointments to fill vacancies, or shifts in committee leadership between now and the next general election. None of that is a judgment on any one person's capacity to serve. It is simply what the aggregate math says about the group as currently composed, and it is worth knowing before you draw conclusions from a single member's profile page rather than the chamber-wide picture.
The averages are drawn directly from our live dataset of all 100 Senate members and all 437 House members currently in office. Age is calculated from public birth records. Completion probability is a statistical estimate built from age and term-related factors, applied consistently across every member in the dataset so that Senate and House figures are directly comparable. Nothing here is rounded dramatically or estimated beyond what the dataset itself produces. The number attached to each chamber is the actual computed average across every sitting member, not a sample or a projection drawn from a subset. Because the underlying data updates as membership changes, these figures reflect a snapshot of the chambers as they stand, not a fixed historical average. That is also why the site treats every number as provisional in the sense that it will shift as terms end, seats change hands, or new members are sworn in, even though the method used to calculate it stays constant.
For the complete breakdown of how completion probability is modeled, how age data is sourced, and how the two chambers are weighted in our comparisons, visit toodamnold.com/data.