Our method for industry association studies
An association study is usually a contribution analysis. It measures the economic activity tied to an industry as it exists today, rather than the effect of a new project. The direct effect is the industry's employment, labor income and output in the study area, built from member surveys and public data such as the BLS Quarterly Census of Employment and Wages.
The main pitfall is double counting along the supply chain. If members include growers and processors, or manufacturers and their distributors, a default model counts the grower's sales once as direct output and again as an indirect purchase by the processor. We prevent this by restricting the model's purchases from industries already in the direct effect, following IMPLAN's industry contribution method. Wholesale and retail members enter at their margins rather than their gross sales.
Scope is agreed in writing at the start: members only, or the whole sector defined by NAICS codes, including non-members. Legislative district results come from member locations or employment by ZIP code. IMPLAN publishes congressional district models, and we build state legislative districts by allocating ZIP-level results.
Typical inputs we ask for
- Member list with addresses and NAICS codes
- Member survey responses on employment, payroll and revenue, or permission to run the survey
- The industry definition you want to use, members only or the full sector
- The geographies you need: state, metro, congressional or state legislative districts
- Any prior industry studies the association has published
Questions clients ask
Can the study include companies that are not members?
Yes, and for advocacy it often should, since legislators care about the whole industry in their district. We define the sector by NAICS codes and estimate it from public employment and wage data, then use member surveys to check and refine the figures. If you need a members-only figure as well, we report it separately.
Can we get results for every legislative district in our state?
In most cases, yes. Congressional districts can be modeled directly in IMPLAN. For state senate and house or assembly districts, we allocate results using member addresses or ZIP-level employment. Sparse industries can produce very small district figures, and we will tell you where a number is too thin to publish.
Why not add up the impact studies our members have already done?
Each firm-level study counts its own purchases from suppliers, and some of those suppliers are other members. Adding the studies together counts those transactions twice, and the firm studies often use different years, regions and models. A single sector-wide model applies one method and removes purchases between members, so the total can stand up when an opponent's economist reviews it.

