A CENTRAL THESIS OF THE ANALYSIS IS THAT THE EXTENT OF PUBLIC PROVISION OF A SERVICE DEPENDS IN PART ON THE RELATIVE COSTS OF PUBLICITY AND PRIVATELY PROVIDED INPUTS. SUCH COSTS MAY BE AFFECTED BY INSTITUTIONS OF PUBLIC DECISIONMAKING, THE RELATIVE STRENGTH OF UNIONS AND EFFICIENCY OF ADMINISTRATION IN THE PUBLIC AND PRIVATE SECTORS, AND THE TECHNOLOGY OF PRODUCING INPUTS IN BOTH SECTORS. THE GENERAL ANALYSIS OF THE CHOICE BETWEEN PRIVATE AND PUBLIC SERVICES IS APPLIED TO THE CASE OF PUBLIC VERSUS PRIVATE POLICE. THE ANALYSIS ENCOMPASSES PUBLIC STATE AND LOCAL POLICE, AND PRIVATE PROTECTIVE AND DETECTIVE AGENCIES. THE ANALYSIS DOES NOT YIELD A DEFINITE ESTIMATE OF THE ELASTICITY OF SUBSTITUTION BETWEEN PUBLIC AND PRIVATE PROTECTIVE EMPLOYMENT. IT IS CLEAR, HOWEVER, THAT THE COSTS OF PUBLIC INPUTS FOR PROTECTION AGAINST CRIME HAVE INCREASED RELATIVE TO THE COSTS FOR PRIVATE INPUTS. ONE IMPLICATION OF THE ANALYSIS IS THAT SUCCESSFUL ATTEMPTS BY PUBLIC EMPLOYEE UNIONS TO RAISE WAGES WITHOUT INCREASING PRODUCTIVITY COULD RESULT IN RELATIVE OR ABSOLUTE DECREASES IN PUBLIC EMPLOYMENT OR AT LEAST IN A MODERATION IN THE RATE OF GROWTH IN LOCAL PUBLIC SPENDING. WITH REGARD TO PROTECTION AGAINST CRIME, SUCH OPPORTUNITIES FOR SUBSTITUTION MAY EXPLAIN THE GROWTH OF PRIVATE PROTECTION IN THE UNITED STATES. MATHEMATICAL FORMULATIONS, EXPLANATIONS OF THE DATA USED IN THE ANALYSIS, AND A LIST OF REFERENCES ARE INCLUDED. (AUTHOR ABSTRACT MODIFIED--LKM)