To read Section III, click here
Introduction
While the methods of conducting an IO analysis are straightforward, there are several common pitfalls that tend to overstate impacts, either because they fail to account for monies that leak outside of the region, or because they ignore monies that may be withdrawn or lost from the economy.
This is not meant to be a complete list, but three of the most common pitfalls in economic impact analysis are:
- Expressing impacts in terms of sales, not earnings
- Using total sales rather than trade margins for retail and similar industries
- Failing to calculate true net impacts
A. Expressing impacts in terms of sales, not earnings
Sales comprise the total dollar amount collected in return for goods and services, while “earnings” refers to total wages, salaries, and other employment-related compensation. At the regional level, sales figures tend to be larger than income, thereby making impacts appear to be greater than they really are.
Consider the following example. Two visitors each spend $40,000 in a region. One visits a local auto dealer and purchases a new automobile. The other undergoes a medical procedure at the local hospital. While both transactions amount to $40,000, they have widely different impacts on the local economy. Of the $40,000 spent for the automobile, perhaps $3,000 remain in the region as salesperson commissions and auto dealer income, while the other $37,000 leave the area for Detroit or Tokyo as wholesale payment for the new automobile. The hospital expenditure, on the other hand, retains perhaps $37,000 as wages for physicians, nurses, and assorted hospital employees (part of the region’s overall income), while only $3,000 leave the area to pay for hospital supplies or help amortize building and equipment loans.
In terms of sales, both the automobile purchase and the medical procedure have the same impact. When expressed in terms of income, however, the former has a much smaller im-pact than the latter. As a true evaluation of regional impact, therefore, the use of earnings rather than sales is a far more accurate measure.
B. Using total sales rather than trade margins
This pitfall is related to the previous one, because it deals with output in the retail, wholesale, and related industries that primarily “re-sell” things they’ve purchased. When I buy a television from a local store, the store keeps a small percentage of the purchase price while the majority goes to pay the product’s wholesale and transportation costs. The TV is not a product of the store; rather, the store’s major “product” is actually a service—the service of selecting TVs, buying them in large quantities, transporting them, and displaying them for purchase by consumers like me. So, we measure the total output of such industries by their trade margins—or the difference between the cost of their products from the wholesaler and the cost to the consumer—not by their gross receipts. This is reflected in the model as well. In EMSI’s model, which shows estimated sales by industry, the “sales” shown are actually only the trade margin portion of sales.
So, when you are modeling purchases from these industries and want to enter the direct effect in terms of an increase in sales, you actually should enter the total increase in the trade margin. If automobile dealers have a 5% trade margin, for example, you would multiply their gross sales increase or decrease by 5% to get the trade margin amount to enter into the model. Note: if you enter direct effects in terms of jobs or earnings, you do not run into this issue.
Finally, if in the rare case that any manufacturers of the type of product sold are also located in the region, you could allocate the remainder of the increase in sales (after trade margins are taken out) to the manufacturer’s industry.
C. Failing to calculate true net impacts
While an economic event may generate direct regional impacts, failure to account for the money withdrawn from the economy as a result of that event results in overstatement of the impacts. As such, gross impacts must be adjusted to account for impacts that would have been generated had the event not taken place, and/or an alternative event had taken place.
Economists often explain this with the “Broken Window Fallacy.”* When an errant baseball breaks a window pane on a house, the good news is that the local building supply store benefits from the sale of a replacement window—a plus side for the economy. From one point of view, therefore, this visible transaction has not only contributed positively and directly to economic growth, but also indirectly through a multiplier effect as the supply store and its employees spend the added income from the sale of the replacement window. Given this reasoning, the careless batter could be naively regarded as a public benefactor, having created additional economic growth by breaking the window.
* The name of the fallacy, and the whole explanation in this section, is summarized from Chapter 2 of Henry Hazlitt’s Economics in One Lesson (2nd edition; NY: Arlington House, 1979).
What’s missing is the invisible impact. The biggest loser is the owner of the broken window and the biggest beneficiary is the building supply store. The store benefits but the homeowner loses because he now has less money to spend on other things. For example, the local clothing retailer loses because he didn’t receive an order for that new suit the homeowner had wanted but now can’t afford. Likewise, all the other vendors who would have benefited from the money earned and spent by the clothing retailer if the suit had been procured are also losers because the money was instead spent on replacing the broken window. In the end, the errant baseball incident does not really add to economic growth at all. The clothing retailer and all of the vendors not benefiting from his business are the “invisible” losers. So at best, there’s no net effect; at worst, economic growth is actually hindered because the invisible loss could be greater than the visible gain.
The broken window fallacy illustrates the difference between gross and net impacts. Measuring only the gross—the transaction between the homeowner and the building supply store—is only half the story. The other half is the adjustment needed to account for the money withdrawn from the economy to make that transaction, money that could have been spent elsewhere had the window not been broken. For the homeowner, that money is what he would have spent at the clothing retailer’s for a new suit, and the subsequent ripple effect that would have occurred as the clothing retailer spent the money earned. These effects should be subtracted from the effect created by the broken window to determine its net impacts.
So, make sure that the regional project you are evaluating is not a giant broken window. Always ask, what are the alternatives to this scenario, and what would their impacts be? An IO model alone cannot tell you this. It requires additional insight and knowledge of the local region and the specific project(s) involved.
And of course, pure dollar-value economic impact is not the only way to evaluate a project, especially public works projects. For example, a public park may have a negative impact because it costs money to maintain and supports only one or two jobs. However, citizens expect government to fund such projects to maintain quality of life in their communities, since the private sector would not undertake them given the conditions put on them by the public (e.g., free use of the park). And quality of life has indirect economic consequences as well—even if those consequences cannot be captured in an IO model.
To read Section V, click here
- Expressing impacts in terms of sales, not earnings
- Using total sales rather than trade margins for retail and similar industries
- Failing to calculate true net impacts
