Measures of school achievement over a 4 year lag period in Virginia's 95 school districts were used to investigate the relationship of investment in education to student achievement since variations in spending among counties would have four years to affect student achievement levels. Because educational output is a function of selected inputs--some of which can be modified by local decision makers--analysis of the production relationship of public education is of particular importance to rural communities faced with limited tax revenues. Analysis of the lagged 1974 cross-sectional achievement scores (grade four) and 1978 achievement scores (grade eight) showed that increasing local inputs for public education had a positive influence on educational output (standardized achievement test scores). The population measure was also significant, providing some support for the hypothesis that the complex community interactions available in more populous areas have a positive influence on educational production. However, using the 1974 data as both input and output measures suggested completely different conclusions; per capita income became the primary determinant of educational achievement and increasing expenditures for education would not necessarily improve school performance. Policy decisions based on research that does not incorporate the proper lag structure could provide results far different from those intended by decision makers. (BRR)