CEOs may not boost public sector productivity1 min read . Updated: 29 May 2019, 03:57 PM IST
More than waiting for CEOs to lead a turnaround, public sector firms should focus on skilling middle-managers, suggests study
Across the world, a commonly proposed solution to public sector management woes is to bring in better leadership. Research though suggests that new leadership at the top may have little effect on public sector performance - at least in the health sector. In a new National Bureau of Economic Research working paper, Katharina Janke and others measure the impact of CEOs on the performance of the National Health Service (NHS), the United Kingdom’s public health provider and one of the largest public sector organizations in the world.
Beginning the 1980s, the NHS decentralized its hospital management with individual hospitals responsible for hiring and paying for their own CEOs. This allowed the authors to measure how effective different CEOs were by using data on hospital productivity measures and CEO remuneration data.
They find that, between 2000-13, CEO pay grew faster than other NHS employees, such as nurses and middle managers, and there was significant variation in pay among different CEOs. However, these pay differentials did not translate to better performance: the authors find little impact of higher-paid CEOs on hospital productivity indicators.
One reason for this could be that CEOs were in charge of hospitals for only a short tenure which made it difficult to affect change in a complex organizational setup with significant organizational inertia. In addition, the public sector nature of NHS hospitals also meant that CEO’s tended to prioritize political goals, such as ensuring their hospital receives no negative publicity, over long-term performance.
The authors suggest that expecting ‘turnaround’ CEOs to improve hospitals may not be realistic. Instead, public sector hospitals should invest in improving the skills of middle managers.
Snap Fact features new and interesting reads from the world of research