individual's instructional services in most communities. If teachers are unionized, there may be even less of a market. Not only can one purchaser, a school district, dominate a local teacher market (a condition known technically as ''monopsony"), but one labor supplier, a teachers' union, can influence the price of teachers as well (a condition of monopoly). In the absence of a market, therefore, it is not possible to assume that actual costs of education inputs are equivalent to what school districts actually spend for those inputs. Expenditures may be too high, or too low, for adequacy.
Teacher compensation typically comprises at least 50 percent of a school district's expenditures, and other professional compensation (for school administrators, counselors, librarians, etc.), tightly linked to that of teachers, comprises another 10 percent (Protheroe, 1997). When so major a portion of an industry's technology falls outside the boundary of an active market, assigning costs can be especially difficult. It is for this reason that, in translating "adequacy" to revenues, it is necessary to rely upon inferences from expenditures, not on expenditures themselves.
As discussed earlier in this chapter, there is an increasingly important body of literature arguing that the number of statistical observations available in education is sufficiently large and varied that economic costs of school inputs generally and teacher salaries in particular can be calculated using statistical regression techniques (Chambers, 1995). However, as we suggested above, we are not persuaded that this approach can overcome the monopolistic and monopsonistic conditions common to almost all of the observations, that the number of observations is, in fact, sufficiently large in a state like Wyoming, or that statistical techniques can control for other characteristics (e.g., teacher quality) with sufficient precision (Rothstein and Smith, 1997; Mishel and Rothstein, 1997). Consequently, other methods of inferring real costs from observed expenditures must be utilized.
In Wyoming, Guthrie et al. assigned costs to prototypical school instructional components through two principal strategies. One, where there was no market immediately in operation, they inferentially linked market cost to a prototypical component. Second, where a supply and demand situation could reasonably be determined to be operating, they took school-incurred prices for such items to be an accurate expression of market costs.
Salaries of education professionals were the principal prototypical instructional components to which market costs could not easily be assigned. This condition was resolved by disaggregating teacher and other professional educator remuneration into four component categories: entry-level salaries, payments for college courses in excess of a bachelors degree, payments for seniority status, and fringe benefits.
Guthrie et al. identified an appropriate level for entry-level teacher salaries by examining teacher salaries in the metropolitan communities where it was presumed conditions approaching labor market competition for college-educated