Why do firms choose to be inefficient?

post by Nicholas D. (nicholas-d) · 2024-08-28T18:39:41.664Z · LW · GW · 4 comments

This is a link post for https://nicholasdecker.substack.com/p/why-do-firms-choose-to-be-inefficient

Contents

  i. Leibenstein’s thesis
  ii. Modeling it
None
4 comments

Firms choose inefficient production technologies for seemingly bizarre reasons. I examine the evidence, and show that this genuinely is worse for everyone, not merely the result of unobserved constraints. I think this can be explained by a model where information is cheap to attain locally but costly to acquire from afar, and where the relationship between money and utility is concave. Lastly, I emphasize the role of imagination in creating productivity.

Hope you enjoy reading it!


Harvey Leibenstein’s 1966 article on what he calls “X-inefficiency” is the forgotten forefather of an enormous literature on productivity. Why are there such big gaps in productivity between firms? Why do they persist so long, without inefficient firms being chased from the market? Why would managers leave simply enormous sums of money on the table? In this essay, I argue that productive inefficiency is real, and that it is largely the result of choices by managers. In a world lacking competition, managers are not pressured to find new ways of producing. 

i. Leibenstein’s thesis

Leibenstein first looks at the distortions from allocative efficiency. Skillfully, he begins by stressing the superfluity of monopolistic distortions. Harberger famously estimated the distortions at 1/13th of 1 percent – other contemporaneous studies find similar estimates. The distortions from tariffs approach, at most, 1 percent. Moreover, monopolistic distortions could never, given plausible elasticities of demand (that is to say, the slope of the demand curve) produce very large losses anyway. 

But the harms from monopoly may not lie in them being profit maximizing, and reducing the level of output to sell higher on the demand curve. Rather, it may lie in the inculcation of sloth and the tolerance of poor management which monopoly enables. Leibenstein thinks that firms systematically do not maximize profit when they don’t face much competition. This is a big claim, and I do not want us to take it too far. That firms may not profit maximize does not mean that neoclassical economics is all of a sudden worthless, or generates no knowledge. The world is kludgy. Information is expensive and non-divisible. Firms could produce things in a better way, if only they knew how. Competition empirically leads to greater productive efficiency, rather than merely greater allocative efficiency. How do we rationalize this commonsense observation in a rigorous way? 

ii. Modeling it

I think the answer is to focus on the externalities of innovation, in the presence of non-linear utility from money. It is cheaper to copy, than to act in a new way. As someone who invests in finding new ways will get only a small part of the total benefits, it will naturally be underprovided relative to the social optimum. As profits fall from the introduction of competition who, importantly, have ideas which are different from the local consensus, it forces local producers to adopt new ideas or perish. This is basically Banerjee’s model of herd behavior – people receive a signal on the right way to do things, and mistake valueless herd-following for a real signal of the best way to do things. An outside push toward better management practices has the opportunity to improve welfare, because things will be stable with too little investment into innovation. 

Second, we can assume that utility from income is non-linear. When firms are doing well enough, the marginal benefit of finding new practices that increase profits is less. People try to avoid risks, even at the cost of losing productive efficiency. When people are in danger of bankruptcy, things change  – it becomes worthwhile to take on more risk in finding new production techniques. This approach has its antecedent in Hart’s 1983 paper “The Market Mechanism as Incentive Scheme”, where satisficing managers only do as much as they are forced to by competition. My take is stronger than Hart’s – where he focuses on agents for the owner, who collect some sort of gain for keeping their job in existence but not necessarily for producing more, I include firms which are entirely owner managed. Even though they collect the whole return to their efforts, the correlation of money and utility isn’t the same over all levels of income.

Rest of the post: https://nicholasdecker.substack.com/p/why-do-firms-choose-to-be-inefficient 

4 comments

Comments sorted by top scores.

comment by habryka (habryka4) · 2024-08-28T18:49:46.938Z · LW(p) · GW(p)

I enjoyed reading this, so I added the first few paragraphs of the post to the LW version, since that tends to make a large difference in how much people actually read the content. Feel free to revert if you don't want that.

Replies from: nicholas-d
comment by Nicholas D. (nicholas-d) · 2024-08-28T23:01:24.167Z · LW(p) · GW(p)

That is an excellent addition — thank you very much for doing so. And I am glad you found it interesting!

comment by df fd (df-fd) · 2024-08-29T03:14:10.970Z · LW(p) · GW(p)

But wasn't bell lab the most innovative when they had monopoly? I recalled people telling me that monopoly has more money to invest in R&D.

Replies from: ChristianKl
comment by ChristianKl · 2024-08-29T09:45:33.030Z · LW(p) · GW(p)

Paying people to do Nobel Prize-winning physics research is inefficient from the perspective of focusing on business profits.