Advances in the Methodological Approach to Friction Period Estimation: A European Perspective

The friction cost approach (FCA) estimates the productivity costs of disease from an employer's perspective but the lack of estimates of friction periods in different countries limits its use. Our aim was to use labour market aggregates to generate two alternative estimates of the friction period for European countries and to apply the FCA to illustrate the impact on cancer-related lost productivity costs. We included thirty countries (EU 27 + the United Kingdom, Switzerland and Norway). Base-case Method 1 used annual Dutch vacancy stock and flow data (2001-2019) to estimate friction periods for this country. A regression model was employed using Dutch data and country-specific vacancy and unemployment rates to generate country-specific friction period estimates for the other 29 countries. Alternative Method 2 used country-specific newly occupied jobs as a proxy vacancy flow variable and vacancy stock data to generate friction period estimates. These were applied, within the FCA, to premature cancer mortality data (from GLOBOCAN2018) for all cancers combined for Western European countries. Costs are in €2018. Method 1 estimated friction periods in 2018 ranged from 70.8 days for Greece to 145.9 days for the Czech Republic, with a mean duration of 95.3 days. Method 2 produced a mean friction period of 80.0 days. On average, across countries, Method 2 friction periods were 15.4 days (-18.5%) shorter than Method 1 estimates. Friction period estimates over the last decade were shorter than those for 2018 reflecting lower vacancy rates. Total cancer premature mortality costs according to FCA Method 1 amounted to €1.0 billion in 2018 for Western Europe compared to €0.99 billion for Method 2. We developed two alternative - and viable - methods to estimate country-specific friction periods. These approaches will enable researchers to apply the FCA to estimate the productivity cost of diseases across Europe from an employer's perspective.