This story was originally published on Sunlight Foundation’s website on Feb. 6, 2012.

House members depend on their staff to run their offices, to help them draft and pass legislation, and to advise them on how to vote. But some members do a much better job of holding onto staff than others.

For the first time, a new Sunlight Foundation analysis has ranked members of the U.S. House of Representatives by their staff retention rates. Our analysis, based on House disbursement data, is a snapshot that compares the staff of House offices between the third quarter of 2009 and the third quarter of 2011.

In this two-year period, office staff retention rates ranged from a low of 19% in the office of Betty Sutton (D-OH) to a high of 94% in the office of Michael Capuano (D-MA). Across all members who served in both 2009 and 2011, average retention rate was 64.2%. At this pace, we would expect the average House office to turn over fully within three sessions of Congress. (Within this sample, the average member will have served 7.3 sessions by the end of this year).

To find your member, click here. (For more details how we calculated these numbers, please see our methodology section at the end.)

Retention rates affect how well members of Congress can do their job, since they rely so much on their staff. And as any manager in the private sector knows, high turnover undermines organizational effectiveness. Hiring and training new staff takes substantial time, and institutional knowledge is frequently lost in the process.

Offices with less experienced staff and less institutional knowledge will generally be less competent. This makes it harder for members to execute their legislative priorities and makes them more likely to rely on lobbyists and special interests for guidance. It may also make it more difficult for offices to adequately serve constituent needs.

Our analysis finds that offices who wish to retain more staff can pay better. Additionally, more senior members have slightly higher retention rates than junior members, and Democrats have slightly higher retention rates than Republicans – although Republicans do hold into people in policy positions at a slightly higher rate.

Table 1 shows the offices with the 20 lowest overall retention rates. The list leans slightly Republican (12 Republicans, 8 Democrats), and includes one former presidential candidate, Rep. Michele Bachmann (R-MN). Bachman also shows up as a “Showhorse” and “Clueless” on the Washingtonian’s “Best & Worst of Congress” list. Often, though, it is difficult for outsiders to know who are good bosses and who are bad bosses, since many staffers are tight-lipped and loyalty is often a perquisite for being hired.

Table 1. Offices with the lowest retention rates

Table 2 covers the 20 offices with the highest retention rates. This list tilts Democratic (14 Democrats, as compared to 6 Republicans).

Table 2. Offices with Highest Retention Rate

Of course, in assessing these offices, we should note that this is based on only two years, for which House disbursement data is available. A longer time frame would likely reveal different patterns.

Committees

We would expect higher rates of turnover on committees, since the majority turned over from Democrats to Republicans in January 2011, and the majority party typically gets two-thirds of the staff on House committees. Given that staff are generally partisan (with the notable exception of the Joint Committee on Taxation and the Ethics Committee), we would expect a retention rate of at least 67% on all committees. With the exception of JCT (which held onto 82.4% of its staff), all other committees have retention rates of less than 67%.

Three committees (Natural Resources at 35.9%, Education and Labor at 38% and Intelligence at 38.9%) had retention rates of less than 40%. These committees are chaired by Rep. Doc Hastings (R-WA), Rep. John Kline (R-MN), and Rep. Mike Rogers (R-MI), respectively.

Table 3.Retention rates by committee

Determinants of Staff Turnover

Looking more closely at the data on turnover, a few patterns emerge. Offices that pay their staff more have higher retention rates. Senior members have higher retention rates. And Democrats have slightly higher retention rates than Republicans.

Figure 1 shows the relationship between average staff salary and staff turnover rate. As we can see, there is a positive relationship, and it is statistically significant. Our analysis (see below for the full model) estimates that for each additional $1,000 the member pays the average staff, the predicted retention rate goes up by 0.5%.

Figure 1. Retention rates and average annual salary

Figure 2 shows the relationship between member’s seniority and staff turnover rate. Again, there is a positive relationship, and it is statistically significant. In the full estimation (see below again), each additional session the member has been in Congress, the predicted retention rate increases by 0.4%.

Figure 2. Retention rates and member seniority

Table 3 (below) shows that Democrats have a 3% higher retention rate on average. It is small difference, though it is a statistically significant predictor of retention rate. Controlling for salary, seniority, and extremism (as measured by ideological voting scores), being a Republican predicts a 4.5% lower retention rate.

Table 3. Republican and Democrat staff retention rates

Table 4 (below) shows the results of a multivariate linear regression that estimates the retention rate based on four variables: Salary (in $1,000s), Member Sessions in Congress, whether the member was a Republican, and member’s extremism (a measure of ideological polarization squared). As described above, each additional thousand in average salary increases the predicted retention rate by 0.5%, each additional member session in Congress increases the predicted retention rate by 0.4%, and being a Republican reduces the predicted retention rate by 4.5%. Extreme members are no more or less likely to retain staff than moderate members.

Table 4. Multivariate regression estimating retention rate by office

N = 328; Adjusted R-squared = 0.1024

These factors are not significant in explaining variation across House committees. Higher salary is associated with higher retention rates for committees, but since the sample is much smaller, the relationship is not statistically significant.

Policy Staff Positions

To the extent that we care about Congress’s policy capacity, we might care especially about the turnover of policy positions in member offices. These are senior staffers who can often be crucial in helping lawmakers draft legislation and formulate positions. We also calculated the retention rate for policy staff positions (for a list of positions, see the methodology selection below)

The (slightly) good news is that the retention rate for policy positions is slightly higher than non-policy positions. The average retention rate is 76% (as compared to 64% for all positions).

There were, however, a dozen offices where at least two-thirds of the policy staff departed between 2009 and 2011.

Table5. Offices with the lowest policy staff retention rates

There is some overlap with the highest overall turnover offices (Reps. Sutton, Terry, McDermott, Lee, Hensarling show up in both lists). Interestingly, Rep. Kucinich shows up in this list, even though he has one of the highest overall retention rates.

None of the same factors that predict overall staff turnover predict policy staff turnover. There are no statistically significant correlates for the variation we observe in our data. The only pattern we see is that Republicans on average hold onto to their top policy staff at a slightly higher rate.

Table 6. Republican and Democrat staff retention rates

Conclusions

As far as we know, this is the first time anybody has calculated turnover rates for House offices. We believe this is important data for citizens to have. Members of Congress cannot do their jobs without staff, and members who preside over high-turnover offices are likely to be less effective as legislators and may have a more difficult time performing efficient constituent service. Moreover, offices with higher turnover are likely to be more reliant on lobbyists to help them analyze and draft legislation, since there will be fewer experienced staff.

Some turnover in all offices is certainly healthy and natural. New blood is always good. But too much turnover can be a dangerous thing, too.

To the extent that members might wish to improve their retention rates, the data do tell us that offices that pay more do a better job of retaining staff, and that more senior members retain staff better. Democrats do a slightly better job of retaining staff over all, but Republicans do a slightly better job of a retaining policy staff.

We also note that this analysis covers a time when the House reduced its own budget by 5% in 2011 and will be reducing its budget again by 6.4% this year. This has already led to a 7.4% reduction in staff positions across the board. The House will be facing an additional 6.4% budget reduction this year, which will likely lead to further cuts in staff.

Methodology

The results are based on a comparison of House disbursement data from third quarter of 2009 and third quarter of 2011. To get a roster of staff in the office in third quarter of 2009, we limited our search to only individuals who received more than $4,000 in the third quarter ($16,000 annual salary) reasoning that anybody who got less money was probably an intern or in a temporary position.

To construct our list of policy staff, we added up anybody with the following as part of their title: Legislative Director, Legislative Assistant, Senior Legislative Assistant, Senior Policy Advisor, Policy Advisor, Policy Director, Legislative Counsel, Counsel, Senior Counsel. Since some offices may have other names for their policy staff positions, it’s possible we undercounted the number of policy staff for some offices.

To find your member, click here.

Our data come from the Office of the Chief Administrative Officer of the U.S. House of Representatives. That means we are dependent on what the House reports. The biggest challenge in aggregating the data is that different House offices classify expenses in different ways. We must in good faith disclose that the underlying data are messy. At best, the data are approximate, and higher levels of confidence in it can only come when the House of Representatives makes a better effort with respect to how it normalizes and releases the data to the public. To dig through the data yourself, visit our House Expenditure Reports Database.

Special thanks to Eric Mill and Daniel Schuman for their help in preparing this piece.