Starting a new business that uses fixed capital (plant and equipment), labor, and other variable inputs, to produce some output creates economic activity. The "impact" of this economic activity can measured by examining the revenue generated by selling the output, the wages paid to workers, the jobs created, or the total money spent on other variable inputs. New tax revenue is also sometimes considered an impact.

Economists sometimes use "input-output analysis" to analyze how these direct economic impacts ripple through the economy to generate additional indirect and induced impacts. Conceptually, indirect impacts measure the extent of the ripple effect that results from linkages with other businesses, while induced impacts capture spending by the firm’s labor force and owners as well as the wages and dividends (or "patronage refunds") they earn.

To accurately estimate indirect economic impact from a given business it is necessary to know the input expenditure profile (i.e., source and quantity of inputs) of the given firm. Induced impacts are estimated by applying wage and dividends generated by the firm to an average household expenditure pattern (i.e., destination and quantity of expenditure), and then by estimating the ways in which these expenditures produce further economic activity. For example, a law partnership, which uses principally a labor input, will generate a large induced effect, but almost no indirect effect. Alternatively, an ethanol plant, which uses significant capital and non-labor variable inputs, but very little labor input, will generate large indirect effects, but a small induced effect. 

For a large-scale study of many firms, collecting detailed information on each firm’s input expenditure profile, or even on total input expenditures, is often prohibitively costly. Therefore researchers often use an "average" profile for a representative firm from the relevant industry. They, then apply to this profile some measure of the scale of operations for the firm as a proxy for total expenditure on inputs. Total revenue is one such proxy, but if the firm is profitable, revenue is typically larger than total input expenditures. Wages are another potential proxy, but using wages will understate total input expenditures because wages do not include non-labor expenses (e.g., the annualized cost of fixed capital).

We conservatively estimate economic impacts in our analysis. At every turn, we have taken steps to ensure that we underestimate the aggregate wage, employment, revenue, and income impacts of cooperative business. For example, we used wages and benefit as a proxy for input expenditure, rather than revenue. This is apparent in our impact estimates where induced impacts are always larger than indirect impacts. We have applied this rule uniformly across each of the 17 sectors, fully recognizing that we may sometimes underestimate indirect economic impacts. This approach is particularly likely to underestimate the full economic impact of lenders in our Financial Services sector. Banks lend to consumers and businesses that in turn invest in various projects ranging from home repair to the launch of an entirely new business venture. In principle, some portion of the value of these projects could be attributed to banks in assessing their economic impact. We do not attempt to do this, as that method would require significant additional data collection and a methodological approach for separating the impact of banks per se from the projects they fund.

We report results on four measures of impact defined below:

1. Revenue: Value of sales
2. Wages: Value of compensation (wages and benefits) paid to employees
3. Income: Value of payments to owners (dividends and patronage refunds) and employees (wages and benefits)
4. Employment. Number of jobs.

For each measure, we estimate direct, indirect, and induced economic impacts across each subsector in our analysis. Aggregate sector reports are compiled by summing impacts across the subsectors in a given aggregate sector.

In some sectors, our data covers all the firms in the given sector. The Credit Union sector, for example, has a trade association and a national regulatory body that collect detailed data on all credit unions in the U.S. However, in some sectors we surveyed individual firms to request data for our analysis, because it was prohibitively costly to survey (and obtain responses) from all firms. In these cases, we imputed values for a representative firm in the relevant sector using the average value for each impact across the firms for which we had data. We then applied the impact from a representative firm to the entire sector by multiplying impacts by the number of firms in the sector. For example, if a given sector included 1,000 consumer cooperatives and we had data on 300 to measure the direct impact for the entire sector, we multiplied the average value from those 300 firms by 1,000. Our aggregate sector tables (see the Commercial Sales and Marketing section, for example) report data only for the cooperatives for which we have direct (not imputed) data, while "direct impacts" in the individual sectoral impact tables (see Agricultural and Marketing, for example) report total imputed values. The Appendix IMPLAN Methodology provides further details about our methodology.