A Comparative Analysis of Employee Turnover in the British and Greek Banking Industry. Issues and Recommendations

Vivi Maltezou - Doctorate gained 2011

Vivi’s doctoral research was co-supervised by Professor Geraint Johnes of the Department of Economics and Professor Paul Sparrow of the Centre for Performance-Led HR. It investigated the issue of employee turnover in both the British and Greek banking sector by comparing original personnel record and event history data in two financial institutions. The study enabled a range of economic and psychological factors to be assessed as predictors of turnover. Reasons for leaving could be linked to a range of other variables such as tenure, current and previous performance, demographic factors and so forth.This dissertation examined the likelihood of worker-firm separations in the context of Great Britain and Greece by using high quality longitudinal databases upon the personnel records from two firms in the banking industry, one British and one Greek.

Duration models were estimated to examine employee turnover in relation to a number of predictor variables, well established in the literature, at a comparative level between the national settings. The available data for the British bank cover the period of 1989-2001, with data on over 20,000 exits, while comparative data were gathered for the Greek organisation over the same period. These longitudinal data provided information relevant to each firm’s workforce in accordance to demographics, and workers’ occupational, organisational and external environment characteristics. The analysis of the data was undertaken through both the formulation of parametric (Weibull) and non-parametric (Cox) duration (survival) models, exploring the effect of particular determinants on employee turnover over time. Vivi also estimated competing risk models by identifying the different outcomes of/reasons for worker-firm separations over time. The analysis also captured the effect of unobserved heterogeneity across workers as approached by the Heckman-Singer model.

The project allowed for an examination of the critical impact of employees’ separation behaviour on organisational effectiveness and the modelling of turnover to the specific industry. The findings on workers’ performance and turnover indicated that it was sustained, rather than instantaneous, performance that was linked to separations. In common with some earlier studies, there was qualified support for a u-shaped relationship between performance and separations, but only in the case of the British data. The negative effect of age on turnover behaviour was also examined. Both of the banks under investigation experienced substantial reorganisation activity over the time period considered. The year following this was characterised by increased separation propensities.

While some findings were consistent across the firms in the two countries studied, there were also some notable differences across the two national settings. Single men were more likely than their female counterparts to quit in the bank based in Britain, but less likely to quit in that based in Greece. Although age appeared to be related non-linearly to turnover, the propensities of British workers to be separated from the bank started to increase at an earlier point than their Greek counterparts.