October 10, 2001

The accounting profession and the federal government clearly have different views of what is extraordinary.

The cancellations and subsequent ticket refunds caused by the attacks of September 11 were sufficiently severe upon the airline industry to generate a Congressional relief package that included instant cash infusion and long term government loan arrangements.

By contrast, the accounting profession has ruled that losses incurred as a result of these attacks are not "unusual," and therefore do not require unusual accounting treatment.

These decisions raise several questions. First, if the terrorist attacks are not extraordinary, what is? Second, if the attacks were not "unusual", why should the government provide financial support for an industry. Third, if the airline industry gets support, what are the principles that exclude support for the auto plants that closed because they could not get parts because the Canadian border was congested, or the hotels who also suffered cancellations requiring refunds, or the expense account restaurants, Broadway shows, and entertainment destinations that had no customers.

Frankly, that first question must be difficult for the accounting profession to answer. I understand that it is difficult to differentiate between normally weak performance and the extraordinary performance on September 11 and beyond. There is a danger that some companies will use the terrorist event to hide customer weakness or bad management.

Nevertheless, other "extraordinary" items are booked following some company presented justification. The same can, and should have been done for the terrorist attacks. It would be just as wrong to assume that a decline in earnings caused by the closing of malls is the same as a drop in earnings because of the absence of customers.

One should be reflected in stock values to indicate that management cannot use capital as effectively as previously thought. The other should hardly have an impact on the future value of companies. The accounting profession is trying to simplify a difficult problem by coming up with the wrong answer.

The second and third questions really require one response.

Economists like to talk about a multiplier effect. If a paycheck is eliminated, the recipient of that check cannot buy groceries, go to restaurants, buy entertainment, etc. As a rule, every paycheck lost because of non-economic events causes an additional paycheck worth of economic activity to be lost.

There is no appreciable multiplier if the paycheck is lost because a better competitor takes the business. For example, if a big discount store opens on the fringes of the town and takes business away from downtown retailers, the total economic gain is small (or even can be negative if the downtown deteriorates, causing people to leave the community).

However, there are some industries that provide infrastructure that allows other industries to be competitive. Development economists have argued that these industries have super multipliers. For example, if you build a port, you not only get the paychecks from the new activity plus its multiplier effects; but you also get the opportunity to create manufacturing and distribution that could not be justified before the port was built.

I believe the airline industry is part of that infrastructure that creates a super multiplier. If we kept all our planes grounded, the economic impact on hotels, tourism, entertainment, business transactions, and other activities would compound the impact of lost paychecks.

We must support our infrastructure. Thus, providing special economic aid to prop up a portion of our infrastructure such as the airline industry is justified. Support for an auto plant that was shut because parts were delayed by congestion at the Canadian border is not.

Of course, if enough empty hotels, stalled auto plants, and empty malls creates a large multiplier effect, some support for the economy may be in order. But no special bailout should be provided unless we are protecting the infrastructure.

 

mbar.jpg (9380 bytes)