How Not to Shorten Government

ice President Gore has called for broadening the span of control in government, leaving the average manager supervising 15 employees, rather than the current 7. The edict is based on the simple notion that there are too many layers between the top and bottom of government. The President's agenda gets weakened on the way down; ideas for improvement get sidetracked on the way up.

Alas, a 1:15 span of control is not the way to shorten government. For starters, spans are much narrower at the very top of government, where a small number of senior political and career employees work, than they are near the front lines. However, managers of front-line workers, being more numerous, make a more tempting target in the kind of "body count" climate that a span-of-control mandate creates. Unfortunately, it is at the senior-management levels that the effects of excessive layering are most insidious.

Another problem is that span-of-control figures are easy to manipulate. The quickest way to alter a span-of-control ratio is merely to reclassify managers and supervisors as team leaders, project facilitators or special consultants. Though this method streamlines the organizational chart, it may leave layers of supervision intact. No matter how long or empowering their titles, no matter how much they swear by team-based management, some former managers just can't resist getting in the way.

If across-the-board mandates and quick fixes were the way to lessen the distance between the top and bottom of government, the three ideas below would merit strong consideration. In the end, however, none of them is much more useful than mandating a "one-size-fits-all" span of control. Eliminate Alter-Ego Deputies. One reason government has so many layers is the nearly insatiable demand for "alter-ego" deputies. Having assistants who fill in for the principal during vacancies of one kind or another has created a series of 1:1 spans of control all the way down the hierarchy.

In theory, only two 1:1 spans of control can be justified in the federal system. The first is between the President and the Vice President -- a span rooted in the constitution. The second is between department secretaries and deputy secretaries, a span often coded in statutes. The rest of the government's 1:1 spans should be reviewed closely.

It would be relatively easy to mandate the elimination of all alter-ego positions -- chiefs of staffs, principal deputies, senior deputies, etc. -- thereby reducing the height of government rather painlessly. The problem, of course, is that the principals to whom these alter-ego deputies report will find it hard to give up the buffering and status that comes with having someone ready to fill in at a moment's notice.

Cap Title-Riding.

One of the growth industries in the layering of government is what I call "title-riding." The quickest way to add a new layer of management is to add one or more of the following modifiers to an existing title: deputy, associate, assistant, senior or principal. These "title-riders" exist in every department, and have grown steadily since the 1960s. The first associate deputy secretary was established in 1960, the first deputy undersecretary in 1952 and the first assistant deputy undersecretary in 1968. Deputy undersecretaries are now found in 11 departments, deputy undersecretaries and assistant deputy undersecretaries in 6 each. Principal deputy assistant secretaries are everywhere.

If a cap is implemented, one can imagine the need for a title czar to process titles and second-guess departments. Departments, in turn, might need to create whole new title units to foster compliance with policy. It's just the kind of control function that Gore rightly seeks to eliminate.

Create a Height Limit. Just as management writer Tom Peters has called for a five-layer maximum in any corporation. Congress and the President could set a maximum height limit for government. The limit would have to be much higher than five, however. Counting from the top, most departments have at least five layers before they even reach the assistant secretary level.

A height limit would create two problems. First, departments would have to figure out how tall they actually are. No one really keeps track, leaving span of control as the best surrogate measure of top-to-bottom distance. A whole new industry would have to be invented merely to measure the height of departments and their sub-units. Second, Congress and the President would likely have to bargain over ideal heights. Just as there is no way to establish an ideal span of control, there is no way to create a perfect height.

The problem with all three options is that the federal government is not a monolith. Just as there can be no ideal span of control either across government or at the various levels of departments, neither can there be a uniform height limit, a ban on alter-ego deputies or an end to title-riding.

The only good way to shorten government is to look at each layer and ask what's necessary and what's not. That means finding out the true distance between the top and bottom, winnowing layers with care and redistributing resources downward as needed.

It may also be time to reconsider the idea of a government reform commission, proposed recently by Sens. John Glenn and William V. Roth but set aside to make room for Gore's National Performance Review. It took decades for government to become as thick as it is today. Perhaps we should take the time to ask exactly how to make it shorter, then use the muscle of the base-closing-type mechanism proposed in the Glenn-Roth bill to make the flattening stick. Merely converting a government of managers and supervisors into a government of team leaders and special consultants who meddle equally fearsomely can't be the answer.