David Brooks has written a terrific column on the challenges facing state governments, and the expenditures that are crowding out investment in infrastructure. Ezra Klein has written a thorough and detailed reply.

Ezra suggests that David’s column blames public sector workers for the problems facing state governments:

Wondering why New Jersey isn’t finishing its much-needed commuter tunnel to New York? Easy, says David Brooks. Blame public employees, their pensions and their pay. “States across the nation will be paralyzed for the rest of our lives because they face unfunded pension obligations,” he writes.

Note that David is making a claim about the long-term burden of unfunded pension obligations. My own reading suggests that David laid the blame primarily at the feet of elected officials, with the leadership of public sector unions playing a secondary role. But read the column for yourself and draw your own conclusion. I’ll focus on Ezra’s reply, as it illuminates many interesting public policy questions.

State pension systems are a problem. But they’re not the problem right now. And they’re certainly not what’s standing between New Jersey and its tunnel. After all, Christie didn’t fund the pension system this year. He simply skipped the $3.1 billion payment, saying he wouldn’t add money to a “broken” system. If I didn’t buy lunch today, you can hardly blame the cost of my lunch for the fact that I don’t have bus fare.

Money is fungible, as the president and his allies have reminded the U.S. Chamber of Commerce. And it is certainly true that there is no one program standing in the way of the tunnel. Rather, as we have discussed, the state’s Transportation Trust Fund is out of money. Moreover, the broader state budget is under severe pressure due to the economic downturn and, of course, the planned withdrawal of federal stimulus funds.

But it is important to remember that New Jersey’s commitment to the ARC tunnel would be a multi-year commitment. To make the ARC project work, New Jersey taxpayers would have to widen trackage leading to the tunnel itself, purchase new rolling stock, train new personnel, and much else besides.

Then, of course, there is the fact that the ARC project will take many years to complete. The completion date was penciled in for 2017. That is, New Jersey taxpayers are committing themselves to many, many, many years of bus fare, and bus fares are pretty darn high these days. It doesn’t seem entirely unreasonable to factor in the future cost of lunch. Indeed, if bus fare is a priority, one might consider switching from filet mignon to, say, steamed broccoli.

Ezra spends the next several paragraphs suggesting that pensions are not currently a major driver of public spending (we disagree on what constitutes major), and he notes, to his credit, that this burden will grow over time. He also writes:

If you look forward 10 years, of course, pensions become a much bigger problem. But not the biggest one. “Assuming 30-year amortization beginning in 2014, [pensions] would rise to only 5.0 percent [of total state spending] and, even assuming a 5 percent discount rate, to only 9.1 percent.”

The fact that pensions aren’t the biggest problem doesn’t seem like a case for not moving to defined contribution pensions like those offered by the federal government. The extraordinary speed with which pension obligations increase by this account would alarm most observers. Josh Barro of the Manhattan Institute is the expert on state pension obligations, so I’m hoping he’ll address the notion that pensions aren’t seriously constraining state governments.

Drawing on a report by Jeffrey Keefe of Rutgers, sponsored by the Economic Policy Institute, Ezra writes:

Brooks also attacks public-sector employees from another angle: “Nationally, state and local workers earn on average $14 more per hour in wages and benefits than their private sector counterparts,” he writes. This relies on a loose definition of the word “counterpart.” It doesn’t mean “compared to the workers like them.” It means “compared to the average worker.”
If you compare workers (pdf) of like education, experience, etc., you’ll find that blue-collar workers in the public sector do a bit better than blue-collar workers in the private sector, but white-collar workers do a lot worse. The calculations that compare average wages neglect to mention that public-sector jobs tend to be white-collar positions, while most positions in the private sector aren’t. More than half of all public employees have college degrees, compared to about 35 percent of private-sector employees.

First, I’m not sure it’s fair to say that David is attacking public sector employees, but of course honest people will have different readings of what struck me as a fairly measured column that took potshots at the antigovernment right.

Second, Ezra writes that blue-collar workers do a bit better in the public sector. If more than half of all public employees have college degrees, that leaves a substantial number of non-college workers who are presumably earning an above-market wage. One assumes that cost adds up over time.

Economist Jeffrey Keefe offers some insight into this question:

A full-time worker on average employed by state and local government received an 11% lower annual earnings compared to the private-sector employees. However, when compared to total compensation, the public employment penalty declines to 2%.

This full-time number includes both college-educated and non-college workers. Given that David was focused on aggregate numbers, it seems useful. Under a fairly favorable set of assumption, the blue-collar compensation edge seems to all but close the college-educated compensation anti-edge. Perhaps college-educated public sector workers should be compensated much more while blue-collar public sector workers should be compensated much less. I tend to think this is a bad idea, as there are non-pecuniary attractions of public service. But we’ll touch on that later.

We’ve discussed Jeffrey Keefe’s paper on many occasions. It is full of interesting insights. And we’ve raised some questions about whether we’re dealing with an apples to apples comparison, as the growing role of pay-for-performance in the private sector has implications for productivity. But arguendo, let’s uncritically accept Keefe’s analysis.

Ezra writes that white-collar workers “do a lot worse.” And Keefe’s analysis seems to bear this out:

When we examine total compensation costs, college-educated public employees cost more than 20% less than similarly educated private- sector employees. The earnings differential is greatest for professional employees, lawyers, and doctors. These earnings differences may create opportunities for cost saving by reviewing professional outsourcing contracts to examine what work might be performed by lower cost public employees.

But again, this does seem to be about the distribution of benefits and burdens between blue-collar workers and professionals, and it’s not obvious that an egalitarian should object to the status quo. Here is the first paragraph of Keefe’s conclusion:

The earnings equation estimates indicate that public employees, both state and local government employees, are not overpaid and may be slightly undercompensated. When we make comparisons controlling for education, experience, hours of work, organizational size, gender, race, ethnicity, and disability, the public employment compensation penalty is relatively small, but there remains a significant difference between private and public employee compensation costs. [Emphasis added.]

“Relatively small” and “a lot worse” are both pretty subjective. The professionals are doing a lot worse, according to the narrow criterion of cash compensation if you accept that Keefe is making an apples to apples comparison. The blue collar workers are doing so much better that they are wiping out what had been a 20 point gap to become a “relatively small” gap.

As far as aggregate numbers are concerned, and aggregates were David’s focus for the obvious reason that he’s talking about the crowding out of spending priorities that he likes, I do think Keefe’s “relatively small” gap is the headline number.

How do we reconcile Keefe’s findings regarding compensation after we correct for hours and skills with the following?

Nationally, state and local workers earn on average $14 more per hour in wages and benefits than their private sector counterparts.

It’s easy. The large majority of public sector workers tend to work shorter hours. The skilled Stakhanovites who work private sector hours suffer. This sounds like a very strong case for pay-for-performance. Indeed, a better system of pay-for-performance might reduce quits of the most effective performers, thus allowing public bureaucracies to do more truly valuable work with fewer employees, to invoke David Simon’s hated phrase “do more with less.” This strategy would be fully compatible with David’s implicit agenda of reducing the burden of public sector compensation on state budgets.

After making the important point that there is plenty of blame to go around — Ezra is definitely right about this, and I’d be much harder on Republicans than Democrats because Republicans are more likely to loudly proclaim their devotion to delivering value for taxpayer money — Ezra writes:

There are real challenges facing states in the coming years. But these are serious, complex challenges, not a simple morality play in which we just need to get tough with state employees and all will be well.

We have good reason to believe that “getting tough,” if by getting tough we mean addressing the compensation premium for blue-collar workers and moving towards more pay-for-performance, more flexibility and autonomy for public sector workers, and more aggressive weeding out of the weakest performers, would yield significant dividends. Getting tough is a serious and complex challenge, but that doesn’t make it any less worth doing.

In the short-term, the problems facing states — the problems that are forcing them to stop construction of ongoing projects — are not driven by pensions or public employees, but by the financial crisis.

I’d agree with this statement. I’d also add, however, that state and local spending has been growing at a far faster rate than private consumption since 1950, long before the financial crisis. The imperative of balancing state and local expenditures against the private sector economy has been here for a while, but it’s been masked by various asset bubbles.

The challenges facing them later are not nearly so simple as greedy state employees or runaway pension spending.

I don’t think state employees have to be greedy to want and expect more compensation. (For the record, both of my parents worked for many years in the public sector.) The issue is that private employees who want and expect more compensation are often disappointed, while state employees have had more success in pursuing a goal that, when framed in the context of growing families and the need to pay for college and braces, etc., is very noble indeed. Just as private vice can lead to public virtue, private virtue can lead to public vice.

Health care is a major player here, as it is everywhere else. And while states are facing serious pension problems because they still offer defined-benefit pensions, we’re also going to see the retirement of millions of private-sector workers who don’t have defined-benefit pensions, and either haven’t contributed enough to their 401(k)s or saw their wealth wiped out in the recent turmoil. We’re facing a pension crisis in the public sector, but we’re also looking at a retirement crisis in the private sector.

That or we’re looking at childless American retirees forced to economize for some period of time, and at multi-generational households that will be stretched by past decision on the part of elderly relatives to emphasize the present — including, of course, invest in children rather than retirement accounts — over the future. This is very vexing, and it is one of the many reasons why I think we should do more to recognize, support, and encourage parental investment in children.

Ezra is dead on when he says that medical expenditures a big part of the problem facing the states. That is one reason why Governor Christie has pressed for public employee contributions to insurance costs comparable to those made by private employees. This is hardly a panacea, but it’s easy to see how it would make a difference to state budgets.

Last thing: after citing Keefe’s research re: the compensation of college-educated professionals in the public sector, Ezra writes:

This shouldn’t be too much of a surprise. Ever talked to a lawyer who works for the government, or a teacher? Have they ever told you they signed up with the public sector because the money is so good? Meanwhile, would that we’d offered more money, and attracted more talent, to the regulatory agencies in the run-up to the crisis. That investment would have paid for itself a thousand times over.

I actually think that many people go into public service because they are public spirited. This is good news for taxpayers. Now, we could decide that, like Singapore, we should pay top civil servants much more. Or we could decide to offer public spirited public employees more autonomy and more scope to do their jobs well, e.g., by doing a better job of handling school discipline and personalizing instruction so that teachers don’t have to struggle with problematic students who draw attention away from the non-problematic majority. I tend to think that this strategy would yield much bigger returns that increasing lockstep compensation, but of course this is an empirical question.

I’m grateful that David and Ezra have raised so many important questions.

I’m flying to California tomorrow for a long-overdue visit to my second favorite coast. And I think I’ll head back there in December. Posting will resume a bit later than usual.