Objective Scarcity and the Paradox of Productivity

In general, if you care about equality, you ought to be passionate about scarcity. As long as there’s not enough of some valuable commodity to go around, then whoever’s richest is going to end up getting it even if the income distribution is relatively flat. By contrast, when you make some category of goods plentiful, you necessarily end up curbing inequities. These days all kinds of Americans can afford a good television. Tragically, though, many Americans can’t afford a house in a safe neighborhood with a decent school that’s within a convenient commute of the central business district of a major city.

That’s Matt Yglesias again, pointing out (correctly) that just because Americans have plenty of food and cheap entertainment relative to just about any period in the past, that doesn’t mean there’s no such thing as poverty.

His solution to the problem of scarcity of good schools and safe neighborhoods, though, is to make more slots available for both – build more housing in good neighborhoods and make good schools educate more children. And this implicitly assumes that opening up more slots won’t change the character of either the schools or the neighborhoods – an assumption that sounds wrong on its face and that many would argue has been disproven in practice repeatedly (most dramatically with the failure of busing).

Here’s the thing. You can improve performance at a single school by attracting a higher caliber of student and/or a higher caliber of teacher. Similarly, you can make one neighborhood safer by attracting a higher caliber of resident and/or improving policing. But the only way to make all schools better in aggregate is to improve the caliber of students and/or teachers generally. The only way to make neighborhoods safer in aggregate is to improve the caliber of residents and/or improving policing generally.

And increasing the number of available slots in good schools or safe neighborhoods can only do this indirectly if at all. The implicit assumption seems to be that new students and residents will acculturate to their new schools and neighborhoods. But it’s at least as likely that the schools and neighborhoods are what will change – for the worse. If you recognize this, then what you have to be arguing for is not simply increasing slots, but giving greater scope and authority to good institutions. Give management of a high-performing network of schools the opportunity to take over a failing school. Give the precinct commander who’s cleaned up a tough neighborhood the job of police chief in a small city. And so forth. But even as you do these obvious things, you’re going to discover that these highly-effective managers succeed in part by selecting good personnel, personnel they select from a larger pool. And if they do a better job of getting rid of lousy teachers or cops, those people go elsewhere. And some other school or precinct gets just a little bit worse.

Public policy has limited leverage over the overall caliber of the student body or the resident population (immigration policy is one of the few levers it does have). It obviously does have some leverage over the quality (and quantity) of teachers and police officers, but even this leverage is relatively limited, and not only by obstructive unions or whatever your particular bugaboo is. It’s limited because talent is the scarcest commodity there is. And the price of talent is driven, most fundamentally, by productivity.

The paradox alluded to in the title of this post is that the cheap television sets are, in a very real sense, the cause of objective scarcity in schools. Productivity growth in big swathes of the economy has made our society objectively richer. In many, many professions, a single individual can generate much more output per hour than he or she could a generation or two generations ago. Teaching and policing, though, have experienced much less productivity growth than the economy as a whole. As you would expect, competition for employees with more productive sectors of the economy has resulted in rising costs combined with declining quality.

We could improve the quality of the teaching and/or policing pool nationwide by the combination of higher spending and more intelligent recruitment, training and management techniques. The result would be an overall better pool for these occupations, which should improve the quality of the service. On the other side of the ledger, we’d experience a deterioration in personnel quality in some other sector in the economy – which might be a good tradeoff depending on what that sector was. But even that will only buy us time. Costs will rise inexorably, and the more concerned we are with maintaining the quality of the labor pool in these essential services, the faster they will rise.

There are only three ways to avoid this trap. One is to cease pursuit of, or even reverse, productivity gains elsewhere in the economy. If the economy as a whole became less productive, professions like teaching would become more attractive – even at relatively lower wages. (It is worth noting that the Soviet Union had a very highly regarded schools system.) On the other hand, productivity is the most important driver of increases in national wealth, and so long as other countries are pursuing national wealth, our national power depends very directly on successfully competing with them. (It is worth noting that the Soviet Union no longer exists.)

The second is to reduce the quantity of the service. Quality of live theater, for example, or of live orchestral music, has arguably remained high in spite of Baumol’s cost disease because there is simply less live theater and fewer orchestras around than there used to be, and they charge high enough prices to maintain high quality. Similarly, we could just teach fewer kids. Police fewer neighborhoods. To some extent, this is precisely what has occurred. This approach amounts to the secession of the more productive sectors of society from the rest of society, which will get progressively worse-educated and less personally secure. It is not an impossible outcome, but it is one that should be abhorrent to anyone with remotely progressive sentiments, or, really, any sense that we ought to have some degree of solidarity with our fellow citizens.

The third alternative is to pursue productivity gains in professions like teaching that have not experienced many such gains historically. There’s some low-hanging fruit here – better teaching and classroom-management techniques can increase the quality of instruction in a given period. Similarly, good collection and use of data can significantly improve the impact of policing by putting cops where they are most needed. But once that fruit is plucked, it’s not clear where you go for additional gains. Recall what productivity means: it means more output per hour. If we’re talking about teaching, that means getting more instruction out of each individual teacher/hour. This is much more difficult to do with teaching than with manufacturing – that’s why we haven’t done it. But that’s the question we have to be asking. (Yglesias has suggested in the past that smaller classes may make schools worse in aggregate because it means hiring more teachers which means, on average, worse teachers. But in the absence of productivity-enhancing innovations, there’s no reason to think that bigger classes would be better – rather, there’s some class size beyond which performance starts to deteriorate for any teacher.)

I’m on the board of a charter school. I believe that letting strong personalities create institutions, giving them scope to achieve their goals and holding them accountable if they don’t, is the way to grow strong institutions, and that strong institutions will deliver better services than weak ones. (And I believe that public policy has a variety of levers to make sure that these institutions genuinely serve everyone, as public institutions must, and should use those levers.) But I also recognize that “do more of what works” is a much, much more difficult mandate than boosters seem to realize.