Incorporating model selection and uncertainty into forecasts of economic conditions in companion animal clinical veterinarian labor markets.

Abstract

Following the onset of COVID-19, demand for US veterinary services rapidly increased, similarly driving up demand in upstream veterinary labor markets. This increasing demand starkly contrasted with the "excess capacity" period following the Great Recession, marked by fewer opportunities and low wages for veterinarians. Better models and forecasts of veterinary labor market conditions improve information for decision-making by veterinarians, relevant business owners, professional associations, and educational institutions. Well-performing approaches will adapt to new data as they become available and avoid the assumption that short-term market dynamics will persist permanently. In this paper, we modify an approach often used to identify the forecasting model best suited to the data. Then, using the best forecasting model, we generate forecasts and prediction intervals around them to meet the available data and realities of veterinary labor markets to 2035. This approach removes subjectivity, accommodates model refinements, and acknowledges that the future is unknown. We find that the supply and demand indicators for veterinary services both indicate continued growth. We do not find statistically significant or consistent differences in either supply or demand growth, which would be consistent with a competitive market for veterinary services and labor. The data do not support an expectation of a continued shortage in the US veterinarian labor markets, as suggested in previous studies. We identify several areas where improved data resources and further methodological advancements could shed more light on the economic conditions facing veterinarians.