Wednesday, March 06, 2013

Good Enough for Government Work

Less than fifty percent of new start-up businesses last beyond five years before failure.  (  The reason: the new business fails to generate sufficient benefits for its customer base.  The successful business is successful if: 1) it recognizes an unmet need of consumers; and 2) it meets that need in an efficient way.  Failure occurs due to trying to enter a market that is already well served, or in failing to economize on delivery costs.  In 2010, 4.1 percent of the employed workforce lost their jobs due to business failure.  (  This number is historically low – trending down from 6.1 percent in 1977.  But the number is still rather startling – 4 out of every 100 people will lose their job this year because some new business owner has over-estimated the net benefits offered by his business to consumers.

The beauty of business failure is this – it is the ultimate cost-benefit test – and it doesn’t rely upon theoretical debates about its costs and benefits.  It is the job of the business owner to control his costs and to convince his potential customers that the benefits of the product or service he offers meets or exceeds the price - he cannot compel customers – he must convince them to spend their own money.   Fifty percent of new business owners, who lay money on the line, and subsequently a good deal of sweat, miscalculate the cost-benefit comparison and in the process lose a good amount of their own investment.

Now consider any given government program or agency.  What is the likelihood that it doesn’t also share the same 50 percent failure rate in terms of delivering benefits that exceed its costs?  In fact the likelihood is very low that it shares the same 50 percent failure rate – it is more likely a failure rate that goes well beyond 50 percent.  Why?  Because unlike a private business, which relies upon convincing each and every customer that the benefits exceed the costs, a government program is financed out of general tax revenues, its costs are hidden, its benefits are in many cases amorphous and a matter of debate, and there is no competitive pressure to deliver more with less over time.

There is the old line that it is not difficult to convince Paul to rob Peter to pay Paul, which refers to the political salability of redistribution in a democracy – if Pauls out number Peters 51-49, you’ve got yourself a winning policy.  Now the reality of course is that there is an administrative cost to robbing Peter and paying Paul – if we’ve stolen $100 from Peter, less than $100 will go to Paul.  How much less depends upon the administrative efficiency of the government.  Even if Peter agrees that he should be robbed to pay Paul, so that the idea of government redistribution is universally accepted, there is still the issue of the fiduciary duty of the government to both Peter and Paul to do so at minimal cost.  If it costs $10 to administer, this is surely worse than a $5 administration fee, as it reduces the amount that goes to Paul.  At some level of administrative fee, the program would also presumably lose the support of Peter.

Let’s say that the Peters actually outnumber the Pauls, but enough of them still vote to tax themselves (and other recalcitrant Peters) to pay the Pauls.  Will administrative excess ever get to the point where Peter pulls back?  Probably not, for various reasons.  One, Peter doesn’t pay a different tax for each different government program – the same taxes go toward the military, law enforcement, the courts, and welfare, so while he may express dissatisfaction with the inefficiency of this or that agency, he views it as a bundle, and may be satisfied with the aggregate bundle even if he knows he is getting screwed on this or that item.  Two, he has very little visibility to the inefficiency of any given agency, and efforts to learn about it are more likely to just add to his frustration, as he has little ability to do anything about it.  Three, ultimately the admin costs may only be pennies – more pennies than are necessary – but still not worth bothering with at the end of the day.  And because it is only pennies, our old friend “Expressive Voting” comes into play – although Peter may oppose a policy because he knows the costs exceed the benefits, his relative powerlessness as one vote among 200 million leads him to espouse favor on the policy so as to be perceived as being compassionate. 

All of which brings me to the sequester, which requires all of a 2 percent cut to the current budget, which is itself an increase over the prior year’s budget.  If you don't bemoan the potential loss of jobs caused by the sequester, you're a heartless conservative who wants to starve children.  It is a problem that the cuts are more or less indiscriminate across agencies – surely there would be a better way to allocate the same cost reductions – but this is not to say that the aggregate bundle of cuts would not be hugely beneficial to the American taxpayer. (i.e. the costs to the taxpayers vastly exceed the benefits).  It is already a well established fact that federal government employees are paid more than comparable workers in the private sector.  (  Workforce reductions in government, as rare as they may be, more often than not occur due to attrition rather than layoffs – it is very hard to get fired from a government job.  If the private sector find that 4 percent of its workforce is simply not cutting the mustard each year (and this figure is due only to business failure, not inclusive of firings for lousy performance), is it at all likely that the number in government is zero? 

In my work, I get to see a lot of contracts between businesses for the delivery of products or services, and these contracts almost invariably require productivity improvements over the life of the contract, delivered as price reductions to the customer.  While it is certainly true that a vendor or service provider who performs well under a contract can expand its business with its customers, it cannot generally raise its price for delivering the same product or service (the one exception is in cases where the product being delivered has a cost that is affected by commodity prices – but even here, there is a formula for pricing that allows for increases due to underlying commodities but which still builds in an expectation of productivity improvements).  Every mature business, after experiencing a growth period where it is scrambling to meet growing demand, and is therefore more afraid of austerity measures than it is of wasteful costs, eventually sees its demand stabilize, at which point in time it starts to trim the fat.  This most recent recession, it is often noted, has been marked by good growth in labor productivity – i.e. companies have laid off the fat without losing output.

This suggests two possibilities – 1) the same exact services provided by government bureaucrats could be provided with a 5 percent reduction to their numbers, as efficiency gains should allow for such reductions to the workforce; or 2) five percent of the services performed by these agencies have costs that exceed the benefits they deliver, and should be eliminated or scaled back in some way; or 3) the same exact services could be provided with a 5 percent reduction to the compensation of the government employees.  We already know that 4 percent of the entire workforce (which is a higher percentage of the private sector workforce) loses their jobs as a result of not delivering sufficient value.  We know two things from the private sector – 1) the company that fails to achieve these types of productivity gains will lose business, ultimately fail, and every person in the organization will lose his job; and 2) the same level of employee, even without the same job security offered by government, is willing to work for less.  If a government agency were to take the first approach, and were to fail in delivering the same “output,” this would clearly be an indication of its lack of efficiency.

The labor market in government is a real problem.  It is hampered by the lack of ability to objectively measure the performance of an agency and the lack of competition forcing the agency to improve its efficiency over time.  It is only due to the lack of competition that it is able to get paid more than private sector counterparts for doing less, with almost no risk of being laid off.  I know and respect many government workers, so this is not a critique of all government workers – many are very capable and hard working people.  But as a group they are overpaid. 

Lord knows there is plenty of dead wood in the private sector, but this is not an argument for letting it persist in government.  Ultimately, the deadwood in the private sector is on borrowed time – especially the white collar dead wood.  I never really have any sympathy for a white collar worker who is part of a workforce reduction (as opposed to those who lose their job because the business fails).  There are plenty of people who feel entitled to a job and a good salary due to various reasons unconnected to their work performance, and many of them are permitted to float along without adding value by managers who prefer not to deal with the mess of firing someone until they are told they have to.  But the market often imposes the discipline that leads to ownership putting the boot to these people.  And when it doesn’t, such inefficiency is at least partially constrained by competition.  To the extent that it is not, that there is always some dead wood, it is a cost born by both consumers and the business owners, but no consumer is compelled to pay for it.  Not so with a taxpayer and his government – there is no market discipline, the excessive costs are borne by the taxpayer alone, and he has no choice in the matter.

As an illustration of the problem, I am aware of one government agency that not too long ago posted openings for economists, and had something on the order of 100 applicants per opening.  Admittedly, not all of these applicants were likely qualified, but let’s be conservative and say that only 40 such applicants were qualified.  Let’s also concede that they may have been applying for other open positions which they may get and choose over the government role.  So now maybe we have only 10 people qualified and available at the end of the day to take each position.  If I see that in the private sector, my first reaction is that I am offering too much money, and I scale back the salary so that the demand for the job is equal to the supply.  I probably do not go to my existing employees at the same level and give them a pay cut, but this is due only to my goodwill – the fact that there are multiple people lining up for the job at a lower salary indicates I can safely fire my overpaid employee and replace him with someone equally qualified.  If you were selling your house, and posted a price that drew in 100 contracts meeting that price, you would inform everyone that they are competing with 99 other contracts, and give them all one last chance to bid.  If you had a real estate agent who randomly accepted one such offer and didn’t ask them to re-bid, you would rightly consider yourself ripped off by your agent.  The same holds true when the government draws 100 applicants per opening and doesn’t lower the offering salary.

How big is the problem?  Federal salaries to civilian employees (i.e. excluding military) were $271 billion in 2011, compared to $3,539 billion for the total federal budget in 2011.   (I don’t mean to exclude the military under the assumption that they are paid correctly, I just couldn’t find the stat inclusive of military salaries.)  This amounts to slightly over 7 percent of the federal budget, not necessarily large in the scheme of things, but shaving $27 billion (10 percent) off of it should be entirely possible without losing much if any benefits, and certainly it is not likely we would lose benefits in excess of $27 billion.  Bottom line – we are probably eliminating jobs that would never survive in the private sector, where no one is willing to pay prices to justify them.  When the taxpayer is put on the hook for such expenses, it is pure theft.  It may be pennies in the ultimate scheme of things, but this just makes it sophisticated theft (per Office Space, it is much smarter to steal pennies at a time across a base of millions of people or transactions, rather than trying to steal $1 million off of one person or one transaction). 

But why limit this discussion to federal government salaries?  If you extend this to state and local government, you are probably talking about real money.  So much of the federal budget is transfer payments unconnected to salaried employees, but especially at the local level, we are primarily talking about salaries for teachers, police, firefighters, etc.  The same labor market problems exist here, except they are limited to some extent by the ability to move to a different town that is more efficient in its provision.  When the flap over teacher’s collective bargaining rights in Wisconsin took place, I had remarked that the fact that there was no shortage of teachers in Wisconsin probably indicated that they were, on average, overpaid as a group.  If I had the power, I would cut their salaries 5 percent across the board, and see if a shortage arose; absent a shortage, rinse and repeat until you see one, and then increase as necessary so that there is no excess supply of teachers.  (It’s a little more complicated than this, as there are no doubt individual teachers who are underpaid relative to their abilities, which is the fault of their unions.  These might be the ones most likely to walk due to a salary reduction, but the point remains that if you could eliminate the union you could lower the average pay and still retain the best teachers). 

As an aside, there is nothing that bugs me more than people whining that such and such a profession deserves to be paid more than it is, because those professions most prone to such whining are the ones most likely to be overpaid.  Teachers are the best case in point.  If you think you are underpaid, leave and find another job, otherwise shut up.  It’s a noble profession, true.  But there is an ample supply of noble people who would do it for less.  (In fact, there are about a gazillion hours that go into voluntarily teaching kids to play various sports – not reading or writing, granted, but not entirely devoid of value).  The only way to maintain premium salaries in such a profession is to limit the ability of would be teachers to compete against each other for available jobs.  This is done by erecting bogus barriers to entry (education degrees), and by unionizing to restrict the ability of local governments to contract directly with individual teachers on mutually agreeable terms.  I know of a guy who has a law degree from Yale, experience in investment banking, significant experience as a youth wrestling coach after competing as a Division 1 wrestler, who has to jump through certain hoops in order to get a job teaching high school in Texas, one of the most conservative states.  Now, he may be a better teacher for going through these hoops (though he may not), but there are plenty of people who have probably jumped through these hoops who are known to still be lousy teachers. 

Sign up for my Notify List and get email when I update!

powered by