Michael Kinsley wrote the following on campaign finance disclosure:
The ability to raise money has become an independent test of a candidate's prospects, completely apart from money's traditional role as a way to buy things. Candidates raise money not to purchase TV time and hire political consultants. They raise money to prove that they can raise money. All the major candidates have rejected federal subsidies in order to avoid the contribution limits that go with them. This includes McCain, whose name is on McCain-Feingold, the most recent failed attempt to curb money's role in politics. When the press started reporting that his campaign was in trouble, McCain hired a major corporate lobbyist as his finance chairman. The press approved. This showed that he was serious.
All of this parallels a development in the larger economy. For most people, the point of money is that you can buy things with it. But at the top, where people already can buy whatever they want, the purpose of money is keeping score: making sure that you don't slip down in the Forbes 400 list.
This prompted Time business and economist columnist Justin Fox to think about executive compensation:
The solution here is so simple: We just have to keep campaign finance data secret. Then candidates would worry far less about how much money they'd raised, and would spend far less time trying to raise it. Hushing up executive pay data would have a similar effect. If nobody outside the board of directors and the head of HR knew how much the CEO was making, there'd be far less keep-up-with-the-Home-Depots upward pressure on pay.
I'm kidding about this, I think. Although when you think about the impact that increased disclosure has had on both campaign spending and executive pay over the past couple of decades, you've got to wonder.
Indeed, there is a widely shared view that executive compensation disclosure has had the unintended consequence of pushing up compensation. Former DuPont CEO Edward S. Woolard, Jr., for example, reportedly stated that: "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 in 2002, I get one too, even if I had a bad year....(This leads to an) upward spiral." This phenomenon is known as the Lake Wobegon effect or the "keeping up with the Jones" problem. Jay Lorsch explains:
... compensation committees rely on surveys by compensation consultants about CEO pay in similar companies but without regard to company performance. These surveys report compensation by quartiles, and no compensation committee wants to admit that its CEO is below the median. In fact, most want to place their CEO in the upper quartile. As a result, CEOs are like the children of Lake Wobegon—all are above average.
The supposed effect is a favorite of financial populists like Gretchen Morgenson, who routinely invokes it in her screeds against CEO pay. (E.g., here and here). It's also a favorite of shareholder activists who claim that pay levels reflect managerial power rather than efficient market contracts.
Curiously, however, the Lake Wobegon effect seems to be more a matter of lore than hard evidence. An interesting analysis of the issue was conducted by economists Peter Swan and Xianming Zhou. They noted that Ontario in 1993 enacted the first Canadian rules requiring disclosure of individual executive compensation. They examined pay practices before and after the regulation went into effect. They found:
By examining CEO compensation of 461 listed Canadian companies over three fiscal years pre-disclosure (1991-1993) and five years postdisclosure (1994-1998) we document that the link between CEO wealth and shareholder value substantially increased after the disclosure regulation was enacted. This observation is reinforced when we use comparable U.S. firms as a control group. In a comparison in which Canadian firms are matched one-to-one with U.S. firms according to size and industry, we obtain two findings: (i) before disclosure, the pay-performance sensitivity for Canadian CEOs was only a small fraction of that for U.S. CEOs, and (ii) the sensitivity increased dramatically for Canadian CEOs after the disclosure requirement in Canada. Our findings are statistically significant and economically strong, as well as robust to model specification. They support the belief of the SEC and legislators that, by removing information asymmetry in executive compensation schemes, disclosure helps make executives more effective agents of shareholders. ...
Our results show that after disclosure, the total compensation to Canadian CEOs increased relative to U.S. CEOs. However, the increase resulted from higher incentive pay due to a stronger pay-performance relation. Thus, our results do not support the view that public disclosure helps either increase or reduce their compensation levels as such, but disclosure improves pay-performance sensitivity....
... after controlling for firm size and performance, we find no meaningful differences in the fixed, i.e., base pay, component of pay between Canadian CEOs and U.S. CEOs, either before or after the disclosure requirement in Canada. This observation does not support the compensation consultant’s view that pay levels are determined by the “keeping up with the Jones” effect, nor the shareholder activists’ view that, due to a lack of shareholder scrutiny, managers were excessively paid and that public disclosure of executive pay helps reduce compensation levels.
Of course, there are valid arguments against the current plethora of executive compensation disclosure requirements (e.g., overloading investors or inherent ineffectuality), but the Lake Wobegon effect isn't one of them. Instead, it seems to be mainly a myth.
Update: In an update to his post, Justin Fox implies that the effect may be limited to Canada. Maybe. What the researchers did was to identify a real world laboratory to test the Lake Wobegon effect. They found a jurisdiction where the law had changed to require significantly greater pay transparency. What effect did that regulatory change have? It increased pay but only because pay became more sensitive to performance. Hence, this is not mere economic modelling; it's real world experimental data. In addition, the paper cites a number of reasons to think the results would hold true in the US.