Are geographic variations in secondary hospital expenditure caused by supply and demand factors? Evidence from migration in England

Abstract

Regional variations in hospital expenditure are a global concern because they may reflect unmet healthcare needs or inefficient resource use. Several recent studies have used internal migration to differentiate demand and supply causes of these regional variations. Most of these studies have focused on specific populations in competitive insurance settings and have not fully considered the timing of individual migrations. Using individual level data on 55 million patients from national Hospital Episode Statistics between 2010 and 2018, we examine geographical variations in overall hospital expenditure and in admissions (emergency and planned), outpatient visits and emergency department attendances. We estimate models similar to the need-based formula used by the English NHS to allocate resources to local commissioners, but we include individual fixed effects to capture demand factors. We find that 10–18 % of the regional variation in hospital expenditure stems from supply-side factors. This contribution is higher for ambulatory care (67.70 % for emergency department attendances and 56.89 % for specialist visits) than for hospital admissions, emergency (12.99 %) and planned (17.91 %). We also find that, after adjusting for need indicators, hospital expenditure is on average higher when individuals live in northern England and in regions with lower prices and higher than expected budget allocations. Despite this significant variation, the share attributable to supply in the English NHS is lower than other countries. Our results have implications for debates about regional variations in health care expenditure and for the importance and design of needs-based resource allocation.