Much of the concern with government deficits in countries as unlike as the United States and Greece focuses on public employees, viewed as overpaid parasites who, being paid by the government, contribute directly to the public debt. And there are indeed good economic reasons to expect the public sector to be less efficient than the private sector. The principal reasons are four: the incentive provided by the profit motive is absent; public agencies tend to be monopolies; public employees are voters; and public employers tend to substitute nonpecuniary for pecuniary emolumens, such as tenure and generous retirement benefits, because the public notices and reacts adversely to high government salaries.
Therefore one might think that the larger the fraction of public employees in a nation’s workforce, the less efficient the nation’s economy, and so the lower per capita GDP would be. (Commonly for international comparisons GDP is translated into U.S. dollars on the basis of estimates of purchasing power parity, and I will do that.) I decided to examine that question empirically, with respect to 27 countries, including the United States and Canada from the Western Hemisphere, Australia, New Zealand, Japan, Taiwan, and Singapore from East Asia, Israel, and all the countries of Western, Northern, Central, and Southern Europe, plus Poland. The countries were not chosen at random, but instead selected as being at least roughly comparable to the United States in their economic system and political culture.
The percentage of public employees in the workforces of these countries ranges from 6.35 percent in Singapore to 33.87 percent in Sweden. Indeed the three lowest countries, and the only ones with fewer than 10 percent public employees, are Japan, Singapore, and Taiwan. The highest countries after Sweden are Denmark (32.3 percent) and Norway (29.25 percent). The remaining Scandinavian country, Finland, is fifth with 26.31 percent. In fourth place, just below Denmark, is Hungary. The other countries with public-employee percentages above 20 percent are Greece (22.3 percent), Canada, and Poland, Greece being the lowest in this group of eight countries, despite all the negative attention its public-employee workforce has received lately.
The rest of the countries in my list (that is excluding the above-20 percent and below-10 percent countries), are grouped pretty tightly between about 12 and 19 percent. The United States is in approximately the middle, with 16.42 percent. Surprisingly, it is well ahead of Israel, Spain, Italy, Germany, France, and Portugal. The European countries with the lowest percentage of public workers are the Netherlands and Austria, but Portugal is only slightly above the Netherlands.
Per capita income, in purchasing power parity terms, ranges from $17,537 in Poland to $53,748 in Norway; interestingly, both have very high percentages of public employees. Regression analysis reveals no systematic correlation between percentage of public employees and per capita GDP, except that the Scandinavian countries as a group exhibit a statistically significant positive correlation between those two variables, if the Asian countries are treated as a separate variable—Singapore has the second highest GDP per capita after Norway, yet the lowest percentage of public employees.
The upshot is that there does not appear to be a relation between a country’s prosperity and the number of public employees it has. (Or between population and the percentage of public employees, though one might expect that, given fixed costs of government, the percentage of public employees would be higher the smaller the population. Singapore is a dramatic refutation of the point, as it has a population of only 4.6 million, one of the lowest of the 27 countries, yet it has the lowest percentage of public workers.)
A more sophisticated analysis would cover more countries (there are 195 countries in the world) and correct for more variables; obviously there is much that affect a nation’s prosperity besides the percentage of its public employees. I am nevertheless surprised that my crude analysis should yield no correlation between per capita GDP and percentage of public workers in a nation’s workforce. The critical omitted variable may be the jobs the public employees do. Are they teachers? Bank examiners? Revenue agents? Food and drug inspectors? Air traffic controllers? Police officers? Medical workers? Or are they railroad workers or other employees of business enterprises owned by the government, politicians’ relatives, licensing officials taking bribes from small business, or beneficiaries of a spoils system of public employment? It does seem significant, though, that the Scandinavian countries should be as prosperous as they are (though Norway, with a very small population and huge oil reserves, may be a special case) despite having such a high percentage of public workers, and it is equally striking that the East Asian countries (though my sample of them is very small) should be so prosperous despite having such a small percentage of public workers. Perhaps the relation between a nation’s economy and the percentage of its public workers is determined by a political and social culture that determines what tasks are assigned to government, what incentives and constraints are placed on public workers, and who is attracted to public service. Maybe, with the right conbination, public service can be as economically productive as private enterprise.
Posner is surprised that his "crude analysis yield[s] no correlation between per capital GDP and percentage of public workers in a nation's workforce." But his statistical analysis does not directly address the issue of whether or not having a large fraction of public employees translates into having a less efficient economy and, hence, lower per capita GDP (than it would otherwise be).
After all, a nation can be efficient yet still poor; inefficient, yet wealthy. Prosperity (in absolute terms) does not really follow from efficiency because there are so many other variables, so we can see that per capital GDP is an inadequate proxy for economic efficiency.
A comparison of percentage growth in per capital GDP versus changes in the proportion of government workers across different countries and over long periods of time might be more illuminating. Such an analysis uses changes in relative wealth as a proxy for economic efficiency rather than absolute wealth.
Posted by: Mitchell K. | 09/26/2011 at 04:35 PM
I'm not so sure that a comparison can be made between Public Sector employees and Private Sector employees regarding their respective "Efficiency" and "Productivity". Such that, the type of "work" and "work product" are vastly different and occur in vastly different "work environments". Any type of "organic" organization will stabilize into some form of operation, process or method. Depending on the environment in which it is called upon to operate in and the work product called upon to produce. Hence, the fundamental differences between the Public Sector and Private Sector.
As for their relative impacts at the Macro-Economic level, once again, I have doubts that any kind relationship or similarity can be established across the board. Due to the various and peculiar phenomenon that exist in the various multiple environments in which the Public and Private Sectors are called upon to operate in.
So the conclusion becomes, every Nation is essentially on its own and needs to come to grips with its own individual problems. In a peculiarly individual and selective fashion that can only fit itself. This just goes to prove there are no set "universal" solutions and trying to base them on a set "ideology" is one of the worst approaches...
Posted by: NEH | 09/26/2011 at 04:55 PM
If I understand the thrust of Posner's comment, his point is that it is a gross overgeneralization to term all public employees "overpaid parasites." He's right for the reasons stated, among others.
Posted by: TANSTAAFL | 09/26/2011 at 08:34 PM
Hmmm, there's even the question of "GDP" measuring efficiency or wealth.
I thought I'd look up the states, as by some measures Alaska is highly productive due to the nature of resource extraction, and it does rank high with $69,000 per capita despite 50% government employees and much of rural Alaska being a subsistence economy of little measurable production.
But then we're more than doubled by DC with a whopping $175,000 per capita. Whew! what IS that similarly sized population producing in that tiny foot print?
Which brings up the question of GDP measurement. "The sum of all goods and services sold?" In the case of comparing nations, I suppose the "overhead" or "enabling" done by government shows up in overall sales?? Well, even DC is not all government workers, so we're getting a reading on highly paid lobbyists? A lawsuit involving $1,000/hour attorneys being more productive than one involving %500/hour lawyers?
Then if the product, however measured, of government workers just happens to be near the average for the nation, as I suspect their wages may be, it would come as no surprise that their numbers (percentages of the work force) would not change things much.
It's and interesting topic and it would be good to know more as to government productivity vs wages paid in an era when some make political hay of government being nothing but overpaid folk of near zero productivity. Ha! to which I'd ask what would be the productivity of a GOOD team of "non-bank-bank" over-seers.
Posted by: Jack | 09/27/2011 at 12:22 AM
The main problem with public goods is one of measuring their value.
Public Value can be diffuse and permeate all but the very primitive activities in economic life such as law and order. Rule of law is fundamental to our way of life. It is obviously worth something but how much?
Other public goods keep on giving for a time horizon well beyond that capable of being measured by a yield curve. The US interstate road network was laid out in the Eisenhower era and it keeps on giving even as it deteriorates from lack of maintenance.
An educated workforce has qualities of being of value through-out the economy and of that value not being realised until years after the investment has taken place. Thus it is often a tempting target to be cutback by the short-sighted or in a fiscal emergency.
This makes justifying public expenditure something of an act of faith. Past experience tells us that public expenditure can pay dividends well into the future but where to spend and how much is much more difficult to measure. It calls for a degree of judgement and a willingness to make decisions in the face of uncertainty that is becoming absent from political discourse these days.
Posted by: Gordon Longhouse | 09/27/2011 at 03:15 AM
Hmmmm, I guess the recent quadrupling of oil prices would make that sector far more "productive" while those teaching our children, police, fire and those picking up our refuse have shown no such productivity gains?
If measured by final sales our financial sector must have improved the standard of living dramatically........... while efficiently steering someone's capital to building lots of homes for the efficient banking sector to carry on their books for years to come.
BTW........... how are the books of the LESS than transparent banks of the EU looking these days? Mebbe five? or ten trillion of junk not marked to market?
Posted by: Jack | 09/27/2011 at 05:39 AM
Gordon, excellent points, well put!
Posted by: Jack | 09/27/2011 at 05:41 AM
As an Argentine citizen, my country is a lab of bad economics, I can hint that cultural factors may explain the different figures accross countries. The idea is not new, Max Weber found a religious/ethical explanation of successful market economies. Guy Sorman partially agrees, is not the religious but the culture, he says. Dr. carlos Nino has blamed anomy and corruption as responsible for the Argentine malaise. All of them may be right. I celebrate Posner's turn, he doesn't sound so canonical neoclassical anymore. The challenge is how do we integrate cultural factors to economic models? Mathematical models and regressions won't fully explain the real world.
Posted by: Jenízaro Negro | 09/27/2011 at 12:57 PM
Measuring the public sector's contribution to society would be easier if government agencies engaged in meaningful cost/benefit analysis for their activities and rules they issue. While not conclusive, such cost-benefit analyses would at least establish some kind of benchmark. Unfortunately the Obama Administration, despite its empty rhetoric to the contrary, shows no inclination toward encouraging or compelling government agencies to weigh costs against benefits. See, e.g., the EPA.
Posted by: TANSTAAFL | 09/27/2011 at 09:00 PM
My experience is that some employees in the public sector and in the private sector are helpful and efficient and some are not but overall in both sectors the culture of helpfulness and efficiency is diminishing. Life in the US is getting difficult and unpleasant economic measures aside.
Posted by: Jim | 09/28/2011 at 10:37 AM
Tans: Most likely we all agree with the concept of "getting our money's worth" and have been so for a long time. But, it's hardly simple, and not to be expected in a short time frame.
For example one area to consider is that of our locking up folks at rates seven times that of the EU and other civilized nations, in many cases for outrageously long terms; young kids, "punks" if you like for a couple decades over some "drug king pin" deal or similar youthful transgression for which a year or two, probation, an effort to rehab and educate may well have served better at far less cost.
But who, especially of what I assume to be your party, is likely to chance being "soft on crime?" So great! let's study the ROI. Surely, maybe? we've some states trying better means and could try more experiments..... but for half or more of the incarcerations being dealt out under federal mandatory minimums.
"Funny thing" those who'd normally favor local control and management by those closest to the problem are most often those who continue to favor hamstringing the judge and neuter his/her insight after presiding over what is often a rushed, even shoddy example of a trial, that triggers a long mandatory sentence, and that sentence being impossible to shorten by the insight and review of a parole board.
At a rate of sever times that of the civilized nations while "the product" is a general crime rate being about the same, but with FIVE times their rate of homicide, mostly with guns, shouldn't this be a particularly ripe arena in which to improve return on investment?
And, taking the biggest chunk first? While it's sacrosanct for the US to maintain the biggest military of all the superlatives, often pork-seeking pols can come up with, are we so poor at defending ourselves, even with geographical advantages of having no nearby "enemies" that we have to outspend the next 14 biggest, national, military budgets combined?
A quick rough suggestion of the value of our huge military expenditures in terms of return on investment? Given the natural advantages the US has over nearly all EU nations our per hour GDP output seems hardly better, France for example has a higher GDP per hour worked than does the US.
While "radical" to be sure, we'd ask ourselves "Do we benefit by "being able to fight 2.5 wars in two theatres simultaneously?"
Suppose we were "weaker" by half in this example and could handle only 1.25 wars at a sitting? If such had been the case when we rushed into Iraq, might things have been better had we not wanted to devote most of our war making capability there unless we had the real support of a number of those 14 next biggest military forces?
But, assuming my "radical" suggestion were right, or at least in the right direction, who would step forth with a decade long, or longer plan to significantly reduce the size of the military? Even the Base Reduction Commission efforts were stymied by changes in the White House and the "best efforts" of pork-seeking Congressmen.
While ROI may be a good measure in many areas, it's a lot less useful in running a nation.
For example, perhaps I'm wrong about warehousing our "criminals" who seem, regardless of race to share most often being functionally illiterate, being a costly waste and economic drag on our nation and is instead the "most efficient" means of cleansing our streets of our educational failures? and even "creates" a fine prison industrial complex of private sector profits and the jobs of prison admin and guards. If it did prove "cheapest" is it still good policy?
It's "handy" of course for those in positions of power and generally supported by the middle classes, as the problem weighs most heavily on "those people" half who are racial minorities and half who are not but share similar demographic and educational positions and a number of our people equal to the number who staff our military is locked up and "no longer a problem".
But as a democracy, and one greatly influenced by Christian values, is this good policy even IF it is "the most efficient?"
As you can see the task of measuring productivity in the public sector is not simple.
Posted by: Jack | 09/28/2011 at 04:23 PM
Jack, evidently you disagree that government agencies should weigh costs against benefits in choosing what they do and what kind of rules to issue. Help us understand why government agencies should be immune from considering costs and benefits -- matters that households all over America struggle with at their kitchen tables each day. For those unable to decipher your post above and understand the precise reasons behind your viewpoint, would you please try again in 100 words or less? (Omit the link to your favorite wealth distribution graph, if you would be so kind.)
Posted by: TANSTAAFL | 09/28/2011 at 07:56 PM
Geez Tans that is a helluva a conclusion to arrive at after my fairly long response regarding the fine intent, absolute difficulty and political impossibility of making decisions based on "cost benefit".
Sorry for using perhaps 500 words that most can read in two minutes, but no thanks, I'm not planning to spend a long time creating a Reader's Digest version for one reader.
As for the link depicting the soaring consolidation of wealth in the hands of fewer and fewer, of which I admit to repeated warnings that it won't work, can't work, and has lead to the greatest Depression in the fairly recent past, I saw no need to include it while discussing the difficulties of even determining ROI on public endeavors, much less the near impossibility of implementing policy on a cost benefit basis.
BTW......... did you happen to note in the knee-jerk, lie based "decision" to go of half-cocked and largely alone into the Iraqi quagmire that the COSTS were spread around many including the very hefty burdens born by those doing back to bank two year "tours" and those losing their young lives, while the "BENEFITS" accrued to a quite small group of war profiteers and those who'll likely divvy the spoils? Can you think of other areas and schemes were it's one group's COST and sacrifice and another group's BENEFIT?
Posted by: Jack | 09/29/2011 at 12:45 AM
Norway needs to be taken out of the analysis. The per capita income from oil extraction in Norway is of the order of $30,000 per capita, similar to Saudi Arabia. Thus deriving any connection between politics and prosperity in Norway is similar to proving the superiority of monarchy by pointing to Saudi Arabian wealth.
Posted by: Pekka | 09/29/2011 at 01:18 AM
Jack asks: "Can you think of other areas and schemes were it's one group's COST and sacrifice and another group's BENEFIT? [sic]"
Thanks for the softball.
Answer: Chris Dodd, Barney Frank, Fannie Mae, Freddie Mac.
How can a popularly elected government declare war on honest people who pay their debts by setting up a regulatory scheme that encourages deadbeats to cheat their creditors, then charge the cost of that ill considered decision to the taxpayers at large, their children, and their grandchildren?
Posted by: TANSTAAFL | 09/29/2011 at 05:40 PM
tanstaafl, And where's your "free lunch" coming from? In the Law it's called Bankruptcy and most of those taking "advantage" of it these days would rather be working a full time job with benefits or running their business's and paying their bills on time as most of them did throughout their lives. Are you advocating the return of Debtors Prisons or better yet, Debt Slavery?
Posted by: NEH | 09/29/2011 at 07:01 PM
Tans -- I ummmm, see where you are coming from, and it's "talk radio?"
"How can a popularly elected government declare war on honest people who pay their debts by setting up a regulatory scheme that encourages deadbeats to cheat their creditors, then charge the cost of that ill considered decision to the taxpayers at large, their children, and their grandchildren?"
........... in THAT vein all I can do is ask if that is about the $12 trillion of the $25 trillion of D E B T being rung up while "conservatives" held the veto pens and said ever so cute things about "deficits not mattering".
And Fannie and Freddy? they went along happily for decades after LBJ sold 'em off to help hide the costs of the Viet fiasco....... ah but those were the days when REGULATED banks with actual LENDING STANDARDS sold them the paper....... ONE whole loan at a time and ONLY those that met their criteria.
I HOPE you mention of them has nothing to do with giving "non-banks" who were ladling out "loans" at over 30 times underlying assets, a pass or "get out jail free" card.
BTW and getting back to the science or craft on economics, I'd remind you that much of it is based on some understanding of human behavior and be it your clubiest of venues with all sorts of reiterations of "accountability" or on the poorest of street corners if you open up a "bank" and begin making "loans" with zip down, on No Job, NO income based on phony assets, you'll do a land office "biz" (as did Ponzi) until A. the payments are due or B. there is the slightest shrinkage in the value of the asset. Want to test it out?
Posted by: Jack | 09/30/2011 at 01:26 AM
typo -- $15 trillion of D E B T.
Posted by: Jack | 09/30/2011 at 01:27 AM
Perhaps the blight of public employees has contributed to why sentencing in courts have become more creative and less jail oriented as the overhead for the sentences of old become more and more expensive: http://lawblog.legalmatch.com/2011/09/26/judges-experimenting-punitive-approach-juvenile-cases/
Just a thought is all...
Posted by: Radical Loo | 09/30/2011 at 10:37 AM
Radical Loo, This is a little off topic, but here goes anyway, in order to solve the problem of of the high cost of incarceration (approx. 35K dollars per inmate per year) we could always return to the use of the Gallows as the sentence of choice. At one time, in the not too distant past, stealing a loaf of bread to feed your hungry family was a Capital offense. Interested in the new Sentencing guidelines (it will save money, but what about Justice)? Perhaps that "blight" of "public employees" has actually had a good effect...
Posted by: NEH | 09/30/2011 at 12:08 PM
NEH - Ha! so, as I've long suspected "an eye for an eye" was a plea for lesser penalties in keeping with the crime.
Texas continues to "use the gallows" with both Bush and Perry seeming to pride themselves on executing those who are criminals, those of serious mental defects and low IQ, and those who might have been among the over 200 now? exonerated by DNA testing had it not been "too late".
Texas appears to have rates of crime similar to those of the more civilized states along with higher rates of gun related homicide. Admittedly, lacking evidence, I'm not adjusting for any "bad seed" factor.
Aaah, yes and what are we to do with the words regarding the holding of life, and liberty as being inalienable? Liberty, of course, can be suspended on behalf of the safety of society and the individual, but life once taken is difficult to restore with current tech.
Posted by: Jack | 09/30/2011 at 05:26 PM
I think your missing variable is the amount of succesful rent seeking going on in the private sector. Public employees aren't the only ones that can be parasites.
Posted by: steve | 10/01/2011 at 01:11 AM
Are there any economic studies out there which correlate institutional size versus efficiency for various nclasses of institutions? For example, hospitals over 500 beds are probable less efficient than smaller ones. Another variable in that mix might be the level ofn autonomy in the various divisions of a larger organization.
Posted by: Jim | 10/01/2011 at 11:24 AM
Jim... I think there are tons of it.. but it varies. Consider one GOOD plumber goes into biz for himself with perhaps one helper and is very efficient. Add two more hourly plumbers? Management costs go up and it's VERY likely that the employees as less efficient..... and wasteful of materials.
On the other hand (of which economists have two or three) efficiency in the auto biz likely begins at a million units/year.
And........ ha, the "efficiency" of our largest operation, the US military is subject of a LOT of discussion just now. Try.... flying fuel to Iraq for costs in the 500/gallon range to our Abrams tanks and other equipment built with a ZERO cost of fuel assumption. That's now being reconsidered, but in terms of security and being ABLE to supply them. The Abrams tank uses a fuel hogging turbine and as heavy as it is, they didn't bother with the small auxiliary you'd expect in a motor home to run the A/C and electronics while standing by....... so the turbines are run all the time.
Posted by: Jack | 10/02/2011 at 06:36 PM
Jack, ;) Remember, in the "New Economics", "Justice", is secondary to or must be made too fit the cost cutting requirements of the new budget requirements...
You're wondering about the operational expenses? Try the new anti-IED scout/recon. cars. They weigh in at about forty to fifty thousand pounds each. Stand about as high as a semi-tractor trailer rig and use about as much fuel as an M-1 perhaps a lot more. In fact, reports are coming back from the Rural areas, the roads and bridges can't support them and are collapsing at an alrming rate and will require trillions of dollars for rebuild and upgrading... ;0
Posted by: NEH | 10/03/2011 at 02:04 PM