Allison Schrager writes
. . . people must rethink the social contract between state workers and taxpayers. As health care gets more expensive and people live longer, the old model simply isn’t sustainable. This means that either benefits must be cut (which, given legal hurdles, is unlikely) or state residents must pay more taxes.
An issue I have with the popular discussion of public sector pay and unionization is that on all sides there is a temptation to frame this as a moral question. What are worker’s rights? What are tax payer’s rights. What is the social contract and is one group cheating the other.
Some of this is unavoidable since public sector pay is influenced by the democratic process. Still we should not encourage it.
The public sector isn’t a stage on which to air our perceptions of the just society, either from the point of view of workers or tax payers. The public sector is a labor market.
In a labor market the greater the total value of the compensation offered the greater the size of the applicant pool. The question facing policy makers is at its heart, do we have too many applicants or not enough? Are our best applicants over qualified or under qualified?
If they are over qualified then you are paying too much. This is bad, but not primarily because it raises costs for tax payers. Obviously tax payers would prefer to pay less, but so would any customer for any service.
If the tax payer is buying a higher quality service than he or she needs then the problem is that we are wasting human talent. That public sector worker could be employed somewhere else in the economy and produce more value there.
On the other hand if public sector workers are under qualified then they are being paid to little. Again this is bad but not primarily because the workers are getting too little pay. All workers, public and private would prefer to be paid more.
It’s a problem because there are workers somewhere out in the private sector who could be creating more value as a public sector worker. This may strike some more libertarian readers as crazy. Yet, consider the following simplistic example.
A talented driver might face a choice of whether to to drive a fire engine or drive a tractor trailer. Both involve a special set of skills that is rewarded by higher pay. It could be the case that human welfare is higher if the best drivers all drove private tractor trailers, but this is not obviously the case. There is a lot on the line in getting a fire engine to a fire quickly and safely. Having the worst or even a mediocre set of drivers could easily be more costly to the public than offering higher pay.
Now, we might imagine that this simplistic analysis is mucked up by the presence of labor unions and public pay scales. Some of this is possible if the union prevents the firing of certain workers or doesn’t allow differences in pay based on differences in ability.
Simply driving up the salaries and benefits for public sector workers, however, will not cause basic labor market mechanics to collapse. This generates inefficiencies but the basic forces of supply and demand still assert themselves. We still have an upward slopping labor supply curve and unless the union is actively keeping people out, that doesn’t change at all.
Unlike in a private firm there is no profit maximizing relationship that matches labor demand to the marginal benefit of customers. This implies that there is no force to mitigate inefficiencies on the customer side. It doesn’t change the nature of the inefficiencies emanating from the labor market, though. Either the workers will have pay in excess of the marginal benefit of quality or the marginal benefit of quality will be in excess of pay.
In a private market the very same inefficiencies would occur. Constriction on the supply of workers would drive up the wage which would drive up prices for customers. The customers would respond by buying less and that would mitigate the inefficiency to some extent. Yet the failure in the labor market would be the same, the bar for quality is set too high.
We shouldn’t think that simply because public workers are employed by the government that the basic rules of supply and demand fall apart. Labor supply curves still slope upward and the quality of workers will still be a function of their compensation.
18 comments
Comments feed for this article
Tuesday ~ February 22nd, 2011 at 12:38 pm
Adam Ozimek
I would add that unions also change the production function. They also distort inputs, in particular the utilization of labor versus capital. Ever see a government worker doing something a machine could and should be doing? Prices aren’t just raised for schools hiring teachers, so to is the range of human resource management policies available to administrators.
Tuesday ~ February 22nd, 2011 at 3:13 pm
Hal
I would really love to see the support for your arguments in this area. If things were as cut and dried as you seem to make out, then I would think that there would be supporting data.
Given the massive fail I’ve witnessed in the private sector, I do wonder what you can come up with that shows such a bright line demarcating the fail of unions.
It’s a bit rich to see consistent assertions without even a nice post laying out the evidence that led to them. Or am I missing ones you’ve laid out in the past? The power of blogging, after all, is within your grasp.
Tuesday ~ February 22nd, 2011 at 4:01 pm
Adam Ozimek
See Barry Hirsch, “Unions, Dynamism, and Economic Performance” for an overview of the literature on how unions impact HRM practices.
Tuesday ~ February 22nd, 2011 at 4:34 pm
Hal
Interesting. The first thing that struck me was the long section detailing the paucity of actual data and how incredibly difficult it is to draw any conclusions based on the little empirical data there is. The second thing that struck me is how long winded the “soft sciences” papers can be talking about stuff they can hardly measure – I’m far more used to actual hard data, falsifiable theories and the papers only exceed 30 pages when it’s a survey of the field or a thesis.
In any event, doing a simple search on my google reader revealed that you’ve been pushing this particular idea about unions for a while and have arguments already lined up for the obvious rejoinders from my side. This seems like a pet peeve of yours, which is not surprising coming from an economist the right.
Still, I doubt that there’s going to be a lot of people convinced of the point of view you’re aggressively pushing without quite a bit more evidence that what appears to be out there. Requiring some rather interesting inferences just to even see the effect doesn’t actually help the case, either, regardless of whether they’re justified or not.
Further, it seems that given the literal mountains of money we’re dumping into things like wars, corporate subsidies, defense contracts with the private sector (no union scum there!), and the wealth sucking black hole caused by the the pinacles of capitalism in the las t couple of years, well… Well, it seems like you’re focussing on a rather small problem if there’s any problem all.
It certainly takes like one requires a large amount of effort to look for the needles one is certain are there. Pretty sure that far more egregious examples of waste that’s perfectly obvious to all just littering the place. Would that would become a pet peeve with the right wing economists…
Tuesday ~ February 22nd, 2011 at 12:52 pm
Psychohistorian
This assumes accurate, meritocratic hiring and monitoring in the public sector. This probably doesn’t even happen in the private sector. Higher wages only result in a slight improvement in worker quality, if that, without effective screening mechanisms.
People in charge of hiring public sector workers are often not really accountable for the quality of said workers. Employees aren’t fired for being less competent than expected. This is particularly true of union members. These jobs are ultimately much more attractive to low-skill workers, who command lower wages in the private sector.
This would also explain why people throw such a fit when pay is reduced: they don’t have the capacity to earn a comparable living in a private sector job. In all fairness, this may be more related to the inflexibility of human capital and the limited inter-relatedness between many public and private sector jobs (teaching, for example, has very limited analogs in the private sector).
Tuesday ~ February 22nd, 2011 at 2:24 pm
Jeff
There’s a recurring theme here: Karl alludes, “the union prevents the firing of certain workers”, and you assert “Employees aren’t fired for being less competent”. Moreover, this was one of the major themes of the recent educational documentary “Waiting for Superman”, in which an economist suggests if we could just fire the worst 5% of teachers and replace them with average teachers we would be the best in the world. This is very intuitively appealing, and it resonated with me. When I was in college, I tutored poor kids in local public schools. In low-SES schools, there really are some terrible teachers who appear to be damaging the prospects of their students and to not really care. It was these memories and the movie that led me to start investigating the issue.
The data were surprising, however, and changed my mind. First of all, as the number of teachers who are unionized in a state goes up, the rate at which teachers are fired for incompetence does not fall: r=-0.04, p>0.10. Second, as the firing rate for bad teachers goes up, educational outcomes do not improve: r=0.01, p>0.10.
data:
Teacher unionization
http://teachersunionexposed.com/state.cfm
Firing rates for incompetence
http://nces.ed.gov/surveys/sass/tables/sass0708_2009320_d1s_08.asp
Educational outcomes
http://nces.ed.gov/nationsreportcard/naepdata/
Tuesday ~ February 22nd, 2011 at 5:38 pm
Jeff
I tried commenting before, but it doesn’t seem to have ‘taken’. Hopefully, this won’t show up twice. My apologies if it does.
There’s a recurring theme here: Karl alludes, “the union prevents the firing of certain workers”, and you assert, “Employees aren’t fired for being less competent than expected”. Moreover, this was one of the major propositions of the recent education documentary “Waiting for Superman”. The movie featured an economist who suggested that if we could only fire the worst 5% of teachers in America and replace them with average ones, we would be the best in the world. The idea resonated with me quite a bit. When I was in college, I tutored poor kids in local elementary schools. In low-SES schools, there really are bad teachers who appear to be damaging the prospects of children without much to fall back on and who really don’t seem to care. It’s terrible. As a result of the movie, and my memories, I decided to look into the issue.
The data, however, surprised me and changed my mind. First, as the percentage of teachers who are unionized goes up, the rate at which teachers are fired for incompetence does not fall: r=-0.04, p>0.10. (That is, states where no teachers are unionized, e.g. SC, have firing rates approximately equal to states where 100% of teachers are unionized, e.g. NH, 0.02 and 0.03 respectively.) Second, as dismissal rates for poor performance goes up, educational outcomes do not increase: r=0.01, p>0.10. This just doesn’t seem to be the problem.
Hal, you might appreciate some real data:
Teacher unionization
http://teachersunionexposed.com/state.cfm
Dismissal rates for under-performance
http://nces.ed.gov/surveys/sass/tables/sass0708_2009320_d1s_08.asp
Educational outcomes
http://nces.ed.gov/nationsreportcard/naepdata/
Tuesday ~ February 22nd, 2011 at 6:55 pm
Karl Smith
The list of links was marking you as Spam
Tuesday ~ February 22nd, 2011 at 7:52 pm
Hal
mmmm… me likes data. thx.
Tuesday ~ February 22nd, 2011 at 4:46 pm
Adam Ozimek
Hal,
There is indeed a paucity of good data on unions. But if you want more evidence that they affect HRM you can just read a newspaper. Honestly, unions don’t constantly and with frequent success fight to limit human resource management practices in schools? And you need data to prove this? Everyone but you seems to agree on this, be it union critics or defenders. I’m baffled, really.
I’d also point out that while I am a frequent union critic, I’m also a fan of the EITC, a generous safety net, much more immigration, carbon taxes, and congestion pricing. I care more about better government than small government. So I don’t think you can put me in a box labeled “economist from the right” and explain away my beliefs like that.
Tuesday ~ February 22nd, 2011 at 5:49 pm
Hal
Well, I work with teachers a lot in the mentoring of high school students that I do. Some of them have become good friends over the years. I can honestly say that I have seen zero evidence for what you’re saying wrt HRM. That doesn’t prove jack, obviously, as the fallacy of biased sample informs me, but it does mean that yes, I do need evidence for such.
Further, HRM in the private sector is a nightmare. I mean, a literal nightmare. And so I’m much more inclined to believe in the sociopathic organizational issues in that realm than I am of the employees that they “serve”. I can only imagine what a true horror such an organization must be in the public sector. But again, that’s just pure bias on my part.
I find that as a general rule, humans are lazy, incredibly inefficient prima donnas who’s apparent purpose is to absorb all available resources while loudly complaining about the comfort of the sofa they’re planted on while doing so. That one finds this in union collectives doesn’t surprise me any more than the hilarious hijinks of Enron’s cadre of the smartest people in the room or the government fattened bonuses of Goldman Sachs.
I just think that employees have as much right to collectively organize as employers have to form alliances, industry lobbying groups and other market distorting corporate collectives that is their wont. That such collective action makes them harder to deal with than a pack of individuals is obvious. It is, after all, the strategic reason why “Divide and Conquer” works so gosh darn well. But given that we seem to believe this a fundamental rights of artificial persons – even when they’re bleeding the government dry in the process – means that we should at least allow it for real persons.
WRT the box, I apologize for using it as a pejorative. I’ll read up on some of your other positions, and expand the model of you that I have.
Tuesday ~ February 22nd, 2011 at 8:04 pm
govt_mule
Data? Data? We don’t need no stinkin data! Everyone knows unions are the problem.
If that doesn’t work for you, here’s a study where an economist did more than download a bunch of coarse-grained data from the internet and run a regression analysis predicated on the same old assumptions – he actually went out in the field and collected data on teacher salaries in every district in three states before and after unionization. Different data, different conclusions.
http://www.nctq.org/docs/effect_of_teachers'_unions.pdf
Tuesday ~ February 22nd, 2011 at 8:31 pm
Hal
That’s a very interesting paper. On page 18, it appears that he has invalidated Adam’s primary thesis on the subject at hand, implying that empirical data invalidates the premise that there’s a positive wage premium for public union employees. And on page 19, another assertion falls by the wayside by relegating the fear that public sector unions exert undue political influence.
Would be interesting to hear Adam’s take on this study and its conclusions.
Thanks for the link.
Tuesday ~ February 22nd, 2011 at 10:20 pm
Adam Ozimek
It’s an interesting study, thanks for bringing it to my attention. Notice that the authors concede it is inconsistent with the past literature. The least persuasive is the estimates of the impact on dropouts, which they take as a proxy for the production function. They use census data for 3 decades, a is long time period yet not enough to employe DiD estimators. Even here they find unionization increases drop-outs for the first 7.7 years after unionization. Also this is but one educational outcome.
With respect to their salary estimates, am I correct they they only report the effects on average salaries? I find it quite plausible that the average effect of unionization is zero, and that the unionization is used to benefit senior teachers at the expense of new teachers. The use of years experience as the determinant for pay strikes me as an example where this happens today.
Another important drawback is the timeperiod: 1972-1991. This pre-dates the modern education reform movement. Would a non-unionized district back then have even known to try paying teachers based on something other than experience? Was value added an option at all?
All that said, and while I haven’t had the chance to read it closely, this does look like a very carefully done study, and it’s results are interesting. One must weigh these results against the evidence provided by the success of charter schools in many states which more often then not operate outside unions, and of the past literature that this study is in contrast to. Here, providing even more recent support, is a more recent study using spatial econometric techniques that shows a large wage-premium.
http://mpra.ub.uni-muenchen.de/21202/1/MPRA_paper_21202.pdf
Tuesday ~ February 22nd, 2011 at 10:23 pm
Adam Ozimek
You also have the judge the study against common sense. Look around you: are school districts policies impacted by what unions will allow? The answer is clearly yes, and in newspapers every day there are examples of things the district or states want but unions block.
Wednesday ~ February 23rd, 2011 at 1:04 am
Jeff
Based on a preliminary read of the paper, it appears that Lovenheim reaches different conclusions primarily in that he’s mostly asking different questions. I don’t see how the paper shows that greater freedom to fire teachers leads to better educational outcomes in the real world. Maybe you’re thinking about his finding of a small, short-term increase in the high school dropout rate following unionization, but he states that “in the long-run [unionization] is associated with a decrease in dropout rates”.
Wednesday ~ February 23rd, 2011 at 9:51 am
Adam Ozimek
I wouldn’t say it shows that. My point is that they look at but one educational outcome, and even here they find short run impacts. I’m also skeptical of the ability to reliably estimate short run and long run impacts using three decennial census time periods.
Wednesday ~ February 23rd, 2011 at 12:10 pm
Hal
From my admittedly outsiderish POV, it seems like you’re arguing over the result of compromise. You don’t like the compromises being made and _seem_ to be looking for reasons not to like it. You claim they’re non optimal and buggy and I’m sure they are. But what, pray tell, isn’t?
Again, it just seems like you’re focussing on a small issue at the expense of #fails that literally swamp any effect you’re seeing. That’s why it causes me to wonder why you’re focussing on this rather than the things that are literally orders of magnitude worse in precisely the same ways you’re claiming this one is.
Cartels? Sure. But as they say, motes and boards. Usually you go after the boards before the motes.