A study of the efficiency of food safety programs of local Public Health Departments in the state of Michigan

The importance of food safety cannot be overstated. "In 2011, the CDC estimation, was that 48 million people get sick, 128,000 are hospitalized, and 3,000 die each year as a result of a food-borne illness" (Kosa, Cates, Karns, Godwin, & Coppings, 2012; Scallan, 2011). The enormous economic encumbrance and personal anguish of foodborne illness can be prevented.;At the forefront of efforts to prevent foodborne illness and promote, monitor and enhance food safety are local health departments (LHDs). In Michigan, LHDs are charged with providing essential public health services, but these LHDs routinely struggle to increase efficiencies and operate with restricted budgets. One of the essential public health services in Michigan is the administration of the food safety program by LHDs. This study examines the efficiency of the food safety program in the 45 LHDs in the state. For purposes of this study, each LHD is a Decision Making Unit (DMU).;The technical efficiency scores for all 45 LHDs in Michigan were calculated using Data Envelopment Analysis (DEA) using secondary data of inputs and outputs from the Michigan Department of Agriculture and Rural Development (MDARD) Annual Reports for the fiscal years 2009, 2011 and 2013. In DEA, scores can range from 0.000 to 1. The minimum score was 0.031. The maximum score, indicating full efficiency, was 1. The average score was 0.707. The total number of DMUs analyzed was N=135. These scores were grouped into four bins using a k-means cluster analysis to avoid tautology. These scores were used as dependent (outcome) variables to complete the testing of the data utilizing multinomial logistic regression analysis. Independent (predictor) variables were inputs and outputs from the secondary data in MDARD Annual Reports for fiscal years 2009, 2011, and 2013. The independent (predicator variables include: Population per LHD, Total Program Revenue per LHD, Program Dollars per Citizen, Program Dollars per Full Time Eligible staff, Licensed Food Service Establishments per LHD.;In this study, the results of the multinomial logistic regression to test five hypotheses sets revealed that population size (p = .001, p = .009 and p = .006), program dollars per citizen (p < .000, p < .000 and p < .000), and number of food service establishments per local health department (p < .000) does significantly predict efficiency. Therefore, the null hypotheses were rejected (p < 0.05) for and the alternate hypotheses were accepted.;The multinomial regression analysis results revealed that program dollars per full time eligible staff (p = 0.739, p = .166 and p = .879) and program revenue per local health department (p = .102, p = .085 and p = .081) does not significantly predict efficiency. For these hypotheses sets the researcher failed to reject the null hypotheses (p > 0.05).;This study is the first to analyze the efficiency of food safety programs administered by local health departments in the United States. Much more research is needed to determine specific areas of improvement in all Environmental Health Programs. Quality of the work performed is of concern as there are some programs that have struggled to maintain minimum program requirements and contractual obligations yet had higher than expected DEA scores.;The DEA reveals a substantial variation of efficiencies among local health departments in Michigan. The main reasons for the variation include funding and staffing. There continues to be a concern regarding efficiency and funding for local health departments in Michigan.