Purpose The purpose of this paper is to use stochastic frontier analysis (SFA) to estimate the efficiency of public investments and their impact on economic growth in the USA using panel data. Results of the study show highly significant and positive relationships between gross state product (GSP) and expenditures on education, transportation, health, welfare, and public safety (police and fire), and negative but significant relationships between output and employment in health care and public safety services. Inefficiencies in the study are measured using per capita tax revenue and time. Tax revenue has a very minimal positive and significant effect on efficiency, while time inversely relates to efficiency. Design/methodology/approach The present study uses SFA to investigate the efficiency of government expenditures in five service sectors – education, transportation, health, welfare, and public safety (police and fire), using recent data and economic trends. The study hypothesizes that changes in the current levels of expenditures in the public sector have a significant impact on the aggregate economy, as measured by GSP. The study uses GSP as the dependent (output) variable, and government expenditure on the five service sectors as the independent (input) variables. Findings Analysis of efficiency for individual states for all 21 years produced interesting results. Overall, the technical efficiency of the public sector was quite high. The average TE score across all years and all states was 0.878. This suggests that public sector operates at a relatively high efficiency level. Originality/value The current SFA model followed Battese and Coelli approach of estimating efficiency of public sectors in each state of the USA. It allowed estimation of policy impact on the overall efficiency. It was applied to macroeconomic panel data.

Objective: Childhood obesity is on the rise in South Africa (SA) and child-directed marketing (CDM) is one of the contributing factors to children’s unhealthy food choices. This study assessed CDM on packaged breakfast cereals available in SA supermarkets and their nutritional quality. Design: Photographic images were examined in a descriptive quantitative study. A codebook of definitions of CDM was developed for this purpose. REDCap, an online research database, was used for data capturing and SPSS was used for data analyses including cross tabulations and one-way ANOVAs. Setting: The current study was set in the Western Cape province of SA. Subjects: Photographic images of all packaged breakfast cereals sold in major retailers in the Western Cape province of SA in 2019 were studied. Results: CDM strategies were classified as direct (to the child) or indirect (through the parent). A total of 222 breakfast cereals were studied, of which 96.9% had a nutritional or health claim, 95.0% had illustrations, 75.2% had product and consumption appeals, 10.8% had characters, 10.8% consisted of different appeals, 8.6% alluded to fantasy and 7.7% had role models. In breakfast cereals with direct CDM the protein and fibre content were significantly lower than in breakfast cereals without direct CDM. This study found a significantly higher total carbohydrate and total sugar content in breakfast cereals with direct CDM than those without direct CDM. Conclusion: CDM was highly prevalent in breakfast cereals sold in SA. Regulations to curb the marketing of packaged foods high in nutrients of concern is recommended.