No single overall measure of the performance of a nonprofit organization is analogous to the profit measure in a for-profit company. The goals of nonprofit organizations are usually complex and often intangible, and their outputs are difficult or impossible to measure.

In general, output information is needed for two purposes: (1) to measure efficiency, which is the ratio of outputs to inputs (i.e., expenses) and (2) to measure effectiveness, which is the extent to which actual output corresponds to the organization's goals and objectives. In a for-profit organization, gross margin or net income are useful measures for both these purposes. In a nonprofit organization, no such monetary measure exists because revenues do not approximate true output as they may in a for-profit company. This note looks at alternative ways of measuring output in nonprofit organizations.

In the absence of a profit measure, analyses of efficiency and effectiveness require adequate substitute measures of output. Despite the importance of devising such alternatives, current nonprofit management control systems tend to be deficient in this respect. Indeed, few nonprofit organizations have specific, written criteria for measuring their effectiveness.

The problem of measuring output in non-monetary terms is not unique to nonprofit organizations. The same problem exists in for-profit organizations in which discretionary costs predominate (e.g., research, law, personnel). Conversely, the output of many individual activities in nonprofit organizations can be measured as readily as can that of corresponding activities in for-profit ones (e.g., food service, vehicle maintenance, clerical work).

Example. A library estimated that it should take two minutes to re-shelve a book (including an allowance for personal time). One hundred books were replaced by a staff person who took four hours to complete the task. The output (100 books replaced) multiplied by the standard time per book (2 minutes) gave a total expected time of 200 minutes, or 3.3 hours (200 ÷ 60 minutes per hour). This can be compared with the actual total of 4 hours to measure productivity. This sort of analysis could be performed in a library of any sort, whether it is part of a nonprofit organization or a for-profit company.

BASIC MEASUREMENT CATEGORIES

Many different terms are used to classify output measures according to what they purport to measure. For our purposes, there are three: (1) social indicators, (2) results measures, and (3) process measures.

Social Indicators

A social indicator is a broad measure of output that reflects the impact of an organization’s work on society at large. Unfortunately, few social indicators can be related to the work of a single organization because in almost all cases they are affected by external forces; that is, forces other than those of the organization being measured. The crime rate in a city may reflect the activities of the police department and the court system, but it is also affected by unemployment, housing conditions, and other factors unrelated to the effectiveness of these organizations. Similarly, life expectancy (or its converse, mortality) is partly influenced by the quality of health care, but it is also affected by nutrition, environment, heredity, and other factors.

Example. The Peace Corps sponsored an attempt to measure the effectiveness of its program in Peru, using measures that purported to show the change in Peruvians’ well-being. Since there was no plausible way of relating the measures of well-being to the efforts of Peace Corps workers, the effort to measure output was probably a waste of time and money.

Valid social indicators are difficult to collect. Indeed, those that can be collected fairly easily are likely to be of dubious validity. Social indicators also are difficult to use properly because there ordinarily is no demonstrable cause-and-effect relationship between what an organization does and the change in a social indicator. Likewise, proxy indicators for intangible factors, such as percentage of registered citizens voting as an indicator of citizenship, or crime and disturbance statistics as an indicator of social unrest, may be collected fairly easily, but frequently are of limited reliability.

Thus, social indicators are nebulous, difficult to obtain on a current basis, little affected by an organization’s current programmatic efforts, and much affected by external forces. As a result, they are only a rough indication of what an organization has accomplished, and therefore are of limited usefulness for management control purposes.