Truck and Barter

If you haven't seen it yet, I highly recommend the back-and-forth between Arnold Kling and J. Bradford DeLong. I trust these two highly intelligent men need no assistance from the likes of me in the larger debate. That, of course, doesn't mean I'm going to refrain from commenting. Apologies for the echo-chamber nature of the thing.

But, as J. Bradford mentions his own jaw does, mine dropped at this early point in the post:

Nobody I know has any complaints about WalMart's efficiency.

This, after mentioning the size of various other "large organizations". Clearly, this man has not spent much time in lunchrooms, coffee-klatches, or near people on smoke-breaks at such firms. That could be the only explanation for his not knowing anyone that has trouble with Wal-Mart (or IBM, or Toyota) efficiency. One need not even spend a full week in a large company to get an earful about slow HR, turnaround times on expense reimbursements, bizarre choices for management promotions, IT desk ineptitude (yes, I guarantee this happens at IBM as well -- in fact, if not more complaining there simply because of the sheer number of people who would be convinced they know how best to fix whatever the problem is), and so on and so forth.

Granted, this is not quite the same as the efficiency needed for getting products to a store on time, but I would challenge anyone to find even a Wal-Mart manager that doesn't have strong suggestions on better ways to order, deliver, and sell goods (let alone human resources policy). I would say, though, that these forms of inefficiency are among the most important to consider. The DeLong post doesn't have the most critical part of the question, namely, "...as compared to...". Considering the federal government, I also couldn't name a single person who would suggest Wal-Mart is on the losing side of any comparison.

The difference, of course, is in the ability of companies like Wal-Mart to correct their inefficiencies. Due to control over the staff, IBM, Toyota, GM, Disney, or most any private company can work to weed out inefficiency by removing the people who either created a problem or were incapable of fixing the problem. Firing someone for underperformance in the US federal government is, I daresay, near to impossible. Crude incompetence doesn't qualify one for termination. Actual harm to others, theft, or something similar is often the bar set for termination. Entire federal buildings could be filled with the people who have their jobs simply because their management is waiting for them to retire. The process of getting them transferred, let alone fired, is too exasperating, too time consuming and filled with administrative hassle to even consider.

Note that this is not simply anecdotal results from a few outdated agencies. I'd suggest a quick re-reading of section 2, chapter 3 of Oliver Hart's Firms, Contracts, and Financial Structure for a pleasantly quick-yet-theoretically-tractable discussion of the incentives between workers and managers. The ownership of complementary assets and the resulting investment in human capital towards those specific assets helps drive efficiency, addressing directly the point DeLong makes about Wal-Mart finding it more efficient to organize as a single company of 1.5 million. (As an aside, this seems like a needlessly superficial characterization of firm structure, since the various groups within a company will face a number of similar problems -- e.g. hold-up -- as we see working across separate firms.) Point being, government employ purposefully, often radically, divorces the incentives of workers from those things that would help drive efficiency. Case in point: in one section of the US Dept. of Agriculture where an acquaintance of mine worked, the person in charge of setting wages (choosing the dictated pay level) has no contact with the people who were managing the group in which my friend worked. The group management had no say, and could only appeal to another division for increases or reprimands. Reduction and firing is unheard of. What interest would there be, then, in working efficiently when it has absolutely no impact on salary?

I agree that there ought to be serious discussions on when certain types of organization should be adopted above others. But I do find it almost alarming that DeLong could suggest, as he seems to be, that a US federal agency could operate anywhere near on par with a private firm if only the right manager was in charge. Political management is answerable to those who appointed them (and, in the case of the Bush administration, often not even then), while those below -- those that do the vast bulk of the work -- are answerable to others who suffer the same lack of aligned incentives as the entirety of the bureaucracy.