Multipliers Explained

When you read an economic impact study and see a sentence like "this project will generate $X million in direct, indirect, and induced economic activity," you are looking at the result of a multiplier calculation. The "multiplier" is at once the most useful concept in input-output analysis and the most abused. Here is what it actually measures, where it comes from, and how to think about it sensibly.

The Three Effects

Direct Effects

Direct effects are the most concrete part of the analysis. These are the jobs, wages, and spending that happen at the project itself. If a hospital employs 500 people and spends $40 million on salaries and benefits, those 500 jobs and $40 million in wages are the direct effects of the hospital. They are observable, countable, and rarely disputed.

Indirect Effects

Indirect effects are the supply-chain ripple. The hospital does not just pay its own staff — it also buys medical supplies, food for patients, laundry services, electricity, and a hundred other inputs. Each of those suppliers, in turn, has its own employees and its own supply chain. The indirect effect captures the economic activity at every tier of the supply chain that supports the original project.

Crucially, the indirect effect only counts spending that stays within the region being analyzed. If the hospital buys MRI equipment from a company in Germany, that purchase generates no indirect effect in the local economy because the money has left. This is what economists call leakage, and it is why a study at the city level shows much smaller indirect effects than the same project modeled at the national level.

Induced Effects

Induced effects capture what happens when employees — both directly and indirectly employed — spend their paychecks. Hospital nurses pay rent, buy groceries, eat at restaurants, and put gas in their cars. Each of those purchases supports yet more jobs. The induced effect models that household-spending ripple.

Induced effects are real, but they are also where input-output models are most fragile. The model assumes that an additional dollar of household income produces a fixed pattern of spending across categories. In reality, household behavior varies — by income level, by household type, by region, by economic conditions. Most of the time, the standard induced effect calculations are reasonable approximations. But analysts should not treat them as precise.

Where Multipliers Come From

A "multiplier" is the ratio of total effects (direct + indirect + induced) to direct effects alone. If a project has $10 million in direct output and the indirect plus induced effects add another $8 million, the output multiplier is 1.8. For every dollar of direct activity, the analysis estimates an additional 80 cents of activity elsewhere in the regional economy.

Multipliers are calculated from the underlying input-output matrix. They are derived, not assumed. The matrix data comes from federal economic accounts (the Bureau of Economic Analysis, the Bureau of Labor Statistics, and others) and is updated regularly. IMPLAN, REMI, and RIMS II all generate multipliers from these underlying data, though they make different methodological choices and produce slightly different numbers.

What Multipliers Do Not Tell You

A multiplier is a ratio, not a proof. It does not tell you:

• Whether a project is a good idea. Multipliers describe the size of activity associated with spending. They say nothing about whether the spending is well-spent or whether the alternatives might be better.
• Whether the activity is incremental. An impact study can correctly calculate multipliers and still wildly overstate the impact of a project if it counts substituted spending as if it were new.
• The fiscal balance. Big economic activity does not mean a positive fiscal impact for any specific government. See our article on fiscal vs. economic impact.
• Long-run effects. Input-output models are static snapshots. They do not predict how the economy will evolve, how prices will respond, or how labor markets will adjust.

Reasonable Multiplier Ranges

For most industries in most U.S. regions, output multipliers fall between 1.4 and 2.2. A few industries with very long, locally-rooted supply chains can reach 2.5 or slightly higher. Multipliers above 3.0 are rare and should be examined skeptically. Multipliers above 5.0 are almost always a sign that something is wrong: the wrong industry code, double-counting, or a region defined too broadly to capture realistic leakage.

The lowest multipliers are typically in retail and services, where margins are thin and supply chains are short. The highest are in manufacturing, where deep supply chains pull inputs from many other industries. Construction is variable, depending on how much of the structural steel, concrete, and equipment is sourced locally versus imported.

A Sanity Check

When you read a study, the simplest sanity check on the multipliers is to divide the headline total by the direct figures. If the result is between 1.5 and 2.2, the analysis is probably in a normal range. If it is much higher, ask the analyst to explain where the extra activity is coming from. The answer might be defensible — but you should hear it.