Bryan Caplan  

Monetizing Job Security

Jason Furman at CEA... Solow on Friedman...
Federal workers' total compensation far exceeds that of private sector workers.  The CBO says so, and so does a large academic literature.  One glaring omission, though, is federal job security.  High job security is nice to have during normal times, and a godsend during recessions.  But since it has no explicit budgetary cost, researchers ignore it.  At least all the researchers I've seen.

Question: If you had to monetize job security, how would you do it?  The obvious Google Scholar searches yield no good hits.  (Anything I'm missing?)  Presumably you'd want to start by multiplying (a) the difference between your probability of being unemployed in a normal job and being unemployed in your actual job and (b) the difference between your total compensation and your reservation wage.  But that's clearly inadequate.  You'd also want to factor in risk aversion, shame, and more. 

Personally, you'd have to pay me at least $50,000 extra per year to get me to renounce my tenure.  It's a terrible system, but I love it. 

Comments and Sharing

COMMENTS (21 to date)
Bryan Willman writes:

One way to think about it - what level of investable assets would allow a person to retire at the same standard of living at which they lived while employed?

That is, of course, a lot of money for most employed people, and it gives some hint to the value of job security.

Hana writes:

I would do it on risk analysis or ROI. Unusually, since the risk is lower and pay is higher than comparable jobs, the analysis would indicate that the alternative job would have to either significantly overpay or offer a large (founder's shares IPO)future payday, in order to justify the risk. If you add in the retirement benefits, the analysis becomes even more skewed.

Hmmm. Everyone should quit their jobs and go work for the government, or at the least get tenure. I almost sense Hayek's blessing. Damn I wish I wouldn't have been an entrepreneur and paid taxes.

David Friedman writes:

The obvious way to measure it would be by finding otherwise closely comparable jobs, one with and one without tenure, and seeing how much employers had to pay to fill each. You might manage by comparing the pay of schoolteachers in otherwise similar districts in different states, since I believe states differ a good deal in how quickly they give teachers tenure.

Meg Kedrowski writes:

Because of the risk aversion you mention, you might need to calculate the value differently in the short term vs. the long term. Using tenure as the example, the current academic, tenured/tenure-track workforce is probably significantly more risk-averse than the general population. (Of course you love tenure --- you chose to follow a career path in which it is a significant component of your total compensation!) In the near term, many people qualified for tenured positions place a disproportionately high value on job security. This shouldn't be ignored, given the years of graduate school/"apprenticeship" required to secure a tenure-track job. In the longer term, if tenure were eliminated, I imagine that the population qualified for academic jobs would gradually come to more closely resemble the general population, at least in terms of risk aversion. The time required to reach the new equilibrium should be less (though still potentially significant) for lower-skilled government jobs.

A related question: How does recruiting people with above-average risk aversion affect the quality of output for government and academic institutions? I suspect the impact is negative and large. Indeed, at least in the humanities and social sciences, this tends to produce the very groupthink that tenure is in theory designed to avoid, reducing the quality of undergraduate education and academic input to policy design.

(This assumes the optimum risk aversion is no higher than the level in the general population --- I'd be interested to hear arguments that this is not the case.)

Hadur writes:

Damn you David Friedman! I was going to suggest the approach you did.

However, I would use the salary of consultants and contractors who work for the federal government. Booz Allen, PWC, and a few other firms have thousands of employees in the DC area who work for the federal government, doing work that is similar to what many federal employees do. They're brought in as contractors because the work is short-term, or there's nobody at that agency with that skill, or more cynical reasons you might imagine.

Compare their wage to that of comparable fed employees. Difference could be job security.

Chris writes:

I would simply look at the incremental compensation per year one would require over time to forego the job security (all future years of employment) and then discount those back to the present value. Per individual, that is the present value of the job security.

Using contracted wages from consultants or contractors would not be a good proxy because consultants in almost any industry earn different wages than their salaried counterparts due to their willingness to travel and do project work that ends after a period of months, i.e. that does not have long-term job security.

MG writes:

So far, I believe Brian and others have been describing ways to value job security. David Friedman's experiment goes one step further by actually establishing the amount a tenured employee would have to give for such security.

To me "monetize" means how to make "liquid" or "capture", the benefits of having job security. I think the credit risk differential between the job secure workers and to the job insecure ones could provide a way to value this factor. All else equal, to capture this value, "all" you would have to do is lend to a pool of "job secures" at the market rate, which can not assume job security. Of course, a "few" other factors -- net worth, etc --would have to be controlled for.

EclectEcon writes:

Bryan, I can understand why you say you'd have to receive an extra $50K to give up your tenure. I assume that is conditional on no one else giving up their tenure.

But what if suddenly ALL universities got rid of tenure? Then if a person lost a job at one place, finding another would likely be much easier. In this latter scenario, I'd place a lower value on my own tenure.

lupis42 writes:

In what industries is unemployement insurance available? What are the cost/benefit ratios like?

A Murricun writes:

I'm fairly certain that the meat-packing, auto and other industries have priced this out from an actuarial approach. It began as the "guaranteed annual wage" circa 1950 in the meat-packing industry, mainly as a means for employers to assure a labor pool for seasonal demand. It is now known as "supplemental unemployment benefits", supported by employer contributions.

That will approach the aspect of cost to the employer.

The value to the employee might be measured as an alternative to an individual insurance plan.

This is one of the several daily shiny toys that I have trained myself not to be distracted with, leaving the work to others.

egd writes:

I think you undervalue your own tenure. Without tenure, given your job performance and expertise, it is probably unlikely that you would lose your job for cause.

You work in a field that has a high demand and lots of long-term stability. Even in a tough economic climate, Universities are still going to get funded and people will still want college degrees. There is a low likelihood of you losing your job for not-for-cause reasons.

Unemployment insurance is a good metric for measuring not-for-cause termination. But how do you measure the benefit of not being fired for cause?

I think we could measure this by assuming that the difference in productivity between secure (public) and unsecure (private) employees is due to concerns for job security.

Sure you'll capture some people who really love their jobs and work because they enjoy it, but I suspect they are statistical outliers.

Bob Knaus writes:

The approach suggested by Hadur, comparing contract to permanent staff cost, seems most promising.

I did a lot of IT labor cost analysis for state and local governments (not fed!) a few years back. In that field, contractors had roughly twice the cost per FTE as permanent employees. Transitioning these to permanent was typically a featured cost-saving recommendation in our reports.

One caution... few of these contractors worked for themselves. Their cost usually included an agency fee or other overhead. So while the government might cut its costs by 50% in converting them to permanent staff, their actual pay cut would be far less... 20% to 25% would be a rough guesstimate.

Brian writes:

David Friedman has the right idea, but comparing compensation with and without tenure of PUBLIC employees (like public school teachers) won't do, since they are already skewed by the governmental nature of the job. The right comparison would be, for example, the compensation of tenured employees at PRIVATE schools compared with the compensation of private-sector employees with comparable credentials.

If I remember correctly, the difference is usually about a factor of 2 in salary (I don't know about overall compensation), so there's a 50% penalty for job security. That suggests that Bryan's $50,000 level is substantially understated.

Tom West writes:

Ms. Kedrowski, I'm not sure how anyone can call attempting to enter academia low-risk.

The odds of finding a tenure track position and then achieving tenure rival that of the odds of a would-be actor making a middle-class living in their profession.

As for Libertarian academics, it scarcely surprising that those who no longer have to worry about the primary insecurity in modern adult life (will I have a job tomorrow?) may not fully internalize the attraction that government policy that offers greater stability has for the majority of the populace.

Much, I imagine, that I, who lives in a society where the idea of needing a gun for self-protection borders on the ridiculous, don't really internalize the importance for others of being able to physically protect oneself against violence.

KnowPD writes:

I don't know what the literature says or how to factor for government workers but for professionals in the private sector a good rule of thumb is the difference in wage between contractors vs full time employees. Many contractors have lower status (sit in cubes,project rooms) but make much more than high status managers sitting in corner offices. Bonus and long-term incentive pay are factored in to the hourly bill rate.

Hazel Meade writes:

One could probably calculate it using a model of the likelihood of being laid off, and typical length of time spent unemployed, and thus compute the expected lost earnings per year for the private sector vesus public sector. Probably start by surverying employees in various fields public and private sector and gathering data to get a good estimate of the likelihood of being laid off.

One possibility is that laid off government employees might be less likely to be laid off but may end up on the UI rolls for longer because government job openings are rarer.

Finch writes:

Bryan's financial risk from unemployment is effectively nil. People who can get employment as economics professors can get other employment in any economic clime.

The benefit of tenure is that it lets him choose what he wants to work on, and can therefore concentrate on things he finds interesting and important even if other people don't value them for reasons of applicability, correctness, taste, politics, whatever. In this job he's free to make controversial points about immigration, and he doesn't have to, say, help PIMCO and Goldman Sachs resolve a dispute over the economic value of something.

Tom West writes:

People who can get employment as economics professors can get other employment in any economic clime.

Is this really true?

I'd be hard pressed to see how adding them as an additional an employee would increase most company's bottom line (or decrease their costs) enough to warrant the hire. Is the demand for research economists in private industry really that high?

Tim V writes:

A great experiment would be to freeze government workers pay and not adjust it until 20% leave for the private sector. None of them would probably ever leave.

Finch writes:

> Is this really true?

If you can get tenure in economics you can consult at decent rates if you have vaguely useful skills. You generally won't have a problem getting in the business at a billing rate most people would consider high, and keeping most of it. You will have some trouble filling enough hours and getting that career to last. But, on the other hand, you don't really need to do too much of that to fund retirement. Call a friend at McKinsey, tell them you'd like to travel, and say you're interested in international trade or something. You'll be cashing $800 per hour checks and redeeming frequent flyer miles from all those trips to Kuala Lumpur pretty soon.

Further, he doesn't need to move to private industry, he could move to public service or another university pretty easily. There's plenty of demand for economists at places like the Fed.

But I really meant that Bryan could choose to do something useful to other people with his brain. He need not be employed as a research economist in private industry. A smart, quantitative person can get all sorts of interesting jobs in private industry, and replacing a professor's pay is no big trick. He might have to do a little reinventing, but that is something most people do a couple of times in their life.

Floccina writes:

You might be able to get some hints by looking at how much AFLAC charges for the benefits that they offer and scaling it up. I would guess that job security is very valuable especially for the less ambitious less hard working. That brings in another factor with working for Government, it is widely believed that there is much less dis-utility to work in government jobs.

Comments for this entry have been closed
Return to top