Quit rates of secretaries and other lower-level federal government employees are considerably below those of comparable workers in the private sector, while government officials quit at much higher rates than their civilian counterparts. What explains this difference, and is it good or bad?
The first part of the question is easy to answer: differences in quit rates are due to differences in the ratio of federal compensation to compensation in the private sector at low and high lob levels. Federal employees at lower level jobs may not make more than their civilian counterparts, but their economic situation is quite good when all other characteristics are taken into account. Government workers at these levels have great job security since they cannot be fired after a short probationary period, except for the grossest forms of misbehavior, such as frequent absences from work, racial or sexual remarks, etc. In addition, they get many holidays, good vacations, generous pensions and health benefits, and are usually not under much pressure at work. The full set of characteristics offered to these federal employees is very attractive, which is why lower level jobs attract many applicants, and the jobs must be rationed through tests and in other ways.
By contrast, federal employees at higher-level jobs-including senior executives in the Department of Homeland Security that Posner discusses- are paid considerably below that of comparable private sector executives. In order to attract and keep high quality employees, the federal government must provide enough compensating advantages in the form of prestige, power, working conditions, and in other ways. There is little turnover of federal judges presumably for that reason since most of these judges could earn much more as practicing lawyers, even judges who, unlike Judge Posner, are not particularly energetic.
For many high-level federal officials, government service is a short-term option that may provide interesting experiences, including learning about various policy issues. But after a while the much higher compensation in the private sector becomes too tempting-of course, their short stay in the government may have been anticipated- and many officials quit the federal government after only a few years (or less).
It would be possible to reduce turnover of federal officials by significantly raising their pay, so that it becomes closer to what they could receive in the private sector. As with federal judges, turnover might be low even with pay that remained considerable below that in the private sector, but not as much below as at present. Some talented men and women like working on public problems of great importance, the security of job tenure would appeal to some, and so on.
Members of Congress currently receive salaries of $165,200 per year-leaders in the House and Senate receives a little more- plus generous retirement and health benefits. That is a lot relative to the pay of most employees in the private sector, and there does not seem to be a shortage of men and women who want to be elected to Congress. Under present rules, it is not possible to pay senior officials in Homeland Security or other agencies more than members of Congress receive, even when higher pay is necessary to fight off the appeal of employment in the private sector. To pay top public officials much more than they already get would require a change in these rules, which would not be politically easy.
Yet for the sake of this discussion, suppose it were possible to cut down the turnover of federal officials in the Department of Homeland Security and elsewhere. Would that be desirable? I believe that having more experienced employees is as valuable to the federal government as it obviously is to the private sector. Turnover of executives with years of experience at the same private company is low- clearly much below that among federal employees- because these companies value that experience. Long time executives accumulate useful knowledge about the organization and practices of the firms they have worked for over the years-what economists call firm-specific capital. I see no reason why such knowledge should be much less important in the federal government. As Posner indicates, Great Britain and some other countries do manage to retain high-level officials of good quality over much of their working lives.
Turnover of federal officials is undesirable for another reason that is not really applicable to the private sector. There is much "rent-seeking" by private companies that try to get special treatment and subsidies from the federal government as it spends huge resources. Hiring of former federal employees to top executive positions at private companies helps them improve their rent-seeking position vies a vies competitors. In particular, if competitors all hire former top executives of the Department of Homeland Security, that may simply offset their rent-seeking positions without adding social value. At the same time losing many top executives weakens this department.
So I favor higher pay for top federal officials in sectors that experience heavy turnover. However, without major changes in pay scales, I am skeptical about the advantages of developing much more stringent rules that prevent federal officials from going to work for suppliers of services or products to the agencies that employed them. The risk of such rules is that they eliminate one of the major present attractions of federal employment at high-level jobs. Adding such rules to the low pay would then make it even more difficult for federal agencies to attract able and honest top-level executives.