Privileges: Market failure in salary fixing and the agency problem (draft)

As the Economist remarks this week concerning the salaries of bankers, supply and demand don’t seem to play a role in fixing the level of salary (direct and indirect) of several collectives. The demand for bankers is far smaller than qualified supply, and so is that for higher corporate management, politically appointed government officials, higher civil servants, university tenured personnel, and a surprising array of other jobs. All of those accrue higher salaries and perks than seem warranted by their scarcity and value generation.

Given that those distorted markets generate higher costs, inequalities and social resentment, it seems reasonable to delve a bit into the probable causes. Agent theory gives a very helpful insight that only reinforces common sense.

Most of those markets have two common characteristic:

– the roles involved are agents managing the assets of a different set of people (shareholders, customers or citizens) who do not scrutinize the decisions.

– the people in charge of setting those salaries are part of the collective into which those new people are being hired (i.e. they are agents of the same kind).

A third characteristic is common but not necessary: the market is usually very small in proportion to the workforce.

Typically (as based on extensive anecdotal observation), the farther the decision is removed from the actual asset owners, the larger the extravagancies of the price or salary. The less transparency or visibility the negotiation has, the higher likelihood of divergence. Large quoted companies without controlling activist shareholders attest to eye-watering differences in salary between managerial and productive roles. European Union officials or parliamentaries have been notorious for opaque and disproportionate remuneration. Parlamentarians and even city council members frequently fix their own salaries and retirement perks beyond the general level for their tasks and qualifications.

In short, it seems there is a simple process in operation: the interest of the agent and the interest of the customer are not aligned. Those agents are (observably) competent in seeking market equilibria for roles that are different or unrelated to their own. But the closer the role gets to their own position (and thus directly or indirectly affects their own remuneration), and the more the decision process can be made opaque, the more they become incompetent due to the distortion created by their own interests.

Putting serious econometric backing into this would doubtlessly prove interesting, but already the weight of anecdotal evidence is simply overwhelming. Plus, those apparently similar situations where those price divergences don’t appear can be directly explained by the same two factors. For instance, management salaries in family-controlled companies are famously tighter, and rocket up when control becomes diffuser. City councils where decisions are subject to hostile scrutiny by an effective opposition rarely call attention. Public companies work in the same way; the recent BBC management appointments in Britain illustrate what happens when there’s a powerful opposition and a driven coalition government, while the current TVE salary ranges in Spain illustrate what happens when a cosseted collective fixes their own wages for a long time without a shareholder powerful enough to call them to task.

Opacity is thus key, if not enough, in bringing about the market inefficiencies. Some of the agent collectives have managed to build hiring mechanisms that help entrench their benefits and in some instances put managers and troop in the same boat (and against the owner). Two examples of those processes are the executive search agencies and the public service selection and guarantees laws in many countries.

Executive search agencies typically earn a commission on the new hiring’s first year of salary; the effect of that on their hiring counsel should be self-evident. Also, they don’t actually try to search for the best value for the job, but for the highest achievements (needed or not) below the customer’s spending limit; again, the effect is not an efficient market price.

Public services have traditionally built a shell around them to insulate them from politician’s (presumedly) undue pressure. Those include an oppositions or entry exams system that is often centered on knowledge of applicable regulations, is easily skewed by the organisers, and does little to ensure relevant experience. A bleeding recent example has been the disproportionate weight of regional language qualifications for technical and medical roles in the Basque government. A typical result is a core civil service that requires several layers of outside, competent contractors to actually get things done (thus inflating the cost of administration). And an all-too-frequent corollary is that the jobs of members of that core are bullet-proof, regardless of productivity or the opinion of the asset owners about their execution.

The investment side of banking is closer to that model than would appear. The people doing the selection and hiring are professionally very close to the new entrants and their remuneration criteria are related. Thus the agent is both empowered and incentivated to reach agreements that reinforce his own remuneration.

Any way out?

From a practical and policy point of view, those distortions are evidently relevant. There have been many attempts to eradicate them.

Public companies have attempted to put executive hiring and salaries decisions out of the hands of management and into special committees answering to the shareholders. The results are not perfect, not least because they routinely use executive search agencies incentivated to find candidates in the highest price ranges. That kind of incentives should be eliminated, both in the specific process and as a practice, at least for the higher reaches of management: C-level and US-style corporate vicepresidencies at the very least.

In those lower-down, specialist selection processes where market inefficiencies are observable, puncturing the bubble would require taking the hiring, and the design of remuneration, out of the hands of related specialists and into those of the people in charge of a higher level of the organization. It would also require a brave person and a firm hand, or a weak labor market: reforming the incentives for banking salespeople has taken a crisis, induced largely by those same incentives, which in turn were set by people who benefited by them (i.e. selling mortgages to anybody for as high a percentage of the house value as possible was driven by targets and incentives to salespeople set by the commercial managers, who in turn had their own earnings defined by short-term sales and not by the quality of the bank’s balance sheet. The very same applies to sub-prime derivatives. In both cases, huge remunerations resulted of value destruction. The lower-level agent defining the incentives should have been kept on a shorter rein).

Parliamentarians (at any level) and other elected officials are routinely criticised for setting their own salaries and adding less transparent perks. Finding a way of weeding that out is harder since there is no higher, independent, transparent, elected, non-partisan instance in most systems (Britain is lucky, at least for a while: the House of Lords should be a good proxy while it lasts). Putting the decision in the hands of another branch of government only reinforces the current incentives for collusion at the expense of the taxpayer. Putting it to any hired expert committee risks enshrining current practices and (if the names of the experts are known) can be subject to pressures. Direct democracy might work, but risk shearing the sheep too close to the bone and making it ruinous to work in politics. A solution might be a body made up of retired leaders of different branches and levels of government and nominally-elected independents, themselves vetted for any party connection.

Privileged professions (see public service) could do with complete disenfranchisement. Avoiding undue political influence can be achieved without either life-long contracts or arbitrary barriers to entry, not to mention salary escalators or (worse) undue bindings between the decision makers’ interests and those of the hired.

Conclusion

Bubbles in salaries (and wider remuneration) are brought about by distortions in the markets that set them. Most often those distortions arise from agents’ interests interfering with those of their customers. That closeness can be less than apparent, for instance through belonging to the same specialist group of having mutually dependent incentives, or quite evident, such as an interest in setting the highest possible price.

Transparency can be a big help in puncturing the bubbles, but some require harder measures: business practices and whole institutions have been built to entrench many of those distortions.