CEO Pay and the Lake Wobegon Effect

This post comes from Rachel Hayes and Scott Schaefer at the University of Utah.

In our paper, CEO Pay and the Lake Wobegon Effect, which was recently accepted for publication in the Journal of Financial Economics, we analyze a common explanation for the recent increase in CEO pay at US firms. Our model, which is based on asymmetric information in financial markets, is motivated by an observation made by former DuPont CEO Edward S. Woolard, Jr. speaking at a Harvard Business School roundtable on CEO pay: “The main reason (CEO) compensation increases every year is that most boards want their CEO to be in the top half of the CEO peer group, because they think it makes the company look strong. So when Tom, Dick, and Harry receive compensation increases, I get one too, even if I had a bad year…. (This leads to an) upward spiral”

We present a game-theoretic model of this phenomenon, which is known in the business press as the Lake Wobegon Effect. We study a firm and a manager who privately observe a parameter that affects the productivity of their match. Stock market participants cannot observe this parameter, but attempt to infer it from observing the manager’s wage. The firm maximizes a weighted sum of short-run and terminal firm value, and so may wish to distort the publicly observable wage contract in order to affect the market’s beliefs regarding firm value. Our model has three key features, all of which have been noted as important assumptions of the Lake Wobegon Effect: (i) there is asymmetric information regarding the manager’s ability to create value at the firm; (ii) the pay package given to the manager must convey information about the manager’s ability to create value at the firm; and, (iii) the firm must have some preference for favorably affecting outsiders’ perceptions of firm value.

We present three main results. First, we show the Lake Wobegon Effect can occur. That is, there are instances of our model in which the full-information benchmark is not an equilibrium, and firms distort pay upward to affect market perceptions of firm value.

Second, we characterize the settings in which the Lake Wobegon Effect does occur. We show that our three key assumptions — asymmetric information, managerial rents, and corporate myopia — are not sufficient to guarantee upward distortions in pay. In the basic version of our model, two additional conditions are necessary: (1) the marginal effect of increases in managerial ability on match surplus — defined as the difference between the parties’ output when working together and when pursuing their outside options — is positive, and (2) the weight placed by the firm on short-run share prices is greater than the fraction of the match surplus that is captured by the manager.

Third, we find that the temptation to distort pay upward is stronger when the information asymmetry pertains to characteristics of the firm rather than characteristics of the manager. When the manager’s ability is uncertain, increases in the manager’s pay do boost the market’s assessment of managerial ability; however, the manager — not the firm — captures rents associated with increases in managerial ability. Thus, the marginal increase in firm value associated with a dollar increase in managerial pay is low when the uncertainty pertains to a characteristic of the manager.

Our analysis suggests that greater shareholder involvement in the pay process — so-called “say on pay” — might be counterproductive. A potential solution to the problem of shareholder myopia is to delegate decisions to individuals with a longer-term view. That is, if shareholders care only about short-run share prices then they may be tempted to raise CEO pay in a (futile, in equilibrium) attempt to affect short-term market valuations. If instead, shareholders delegate pay decisions to directors and motivate those directors (using contracts) to take a longer-term view, then pay decisions can be insulated from myopia.

The full paper is available for download here.

Both comments and trackbacks are currently closed.

One Comment

  1. Rob
    Posted Wednesday, January 21, 2009 at 3:25 pm | Permalink

    Very enjoyable experiment to read and interesting theoretical game. Maybe, the opposite could be tried to see if the results are inversely related in any way?