Productivity, Wages and Training in Ageing Societies

The foreseen shrinking and ageing of the labour force in most industrialised countries will put a significant burden on the social security system. To preserve the sustainability of the welfare system, productivity growth needs to be sustained. Besides technological progress, human capital investment, economic structure, etc., the age structure of the workforce has an important impact on economic productivity. The question arises whether an ageing workforce might hamper economic productivity and therefore increase the "demographic burden" (the relation between retired and working persons) even further. To disentangle the relationship between economic productivity and the age structure of the workforce, we follow mainstream literature that currently argues that the firm level constitutes one of the most promising levels of analysis since it controls for the structural constraints the workforce is confronted with. We base our study on a recently created panel data set of matched employer-employee data for Austria that became available in September 2008. Currently the dataset covers the years 2002-2005 and during the ongoing project the data will be updated continuously. The first objective of the project is to verify whether a higher share of older workers negatively effects firm-level productivity for Austria. By applying advanced panel regression methods we can control for unobserved (time- invariant) heterogeneity and the endogeneity of independent variables. Another objective is to test for the sensitivity of the results with respect to the representation of the workforce age structure. Obviously it is the age composition of the workforce-as characterised by the mean as well as the variance of the age distribution-that matters for firm-level productivity. The third objective of our proposal is to contrast the age-productivity profile with the age-wage profile at the firm level. Based on the theoretical and empirical literature so far, we expect that these measures diverge, with wages being below productivity at lower ages and above productivity at older ages. The fourth objective of the proposed project is to also control for training at the firm level. We will test whether the age-productivity relation changes when we control for training at the firm level. Moreover we shall also investigate who gains most from training by investigating the effect of training on firm-level productivity versus its effect on the wage sum. Finally we will also conduct our analysis on the sector level (as opposed to the firm level) to test whether external productivity effects emerge between firms within the same sector. Summing up, our study constitutes the first comprehensive analysis of the age structure impact of a firm`s workforce on labour productivity using panel data for Austria. Through our project we also aim to promote the establishment of a continuously updated employer-employee dataset that will be of great use for any further research projects dealing with topics in labour and business economics.

In most industrialized countries the labour force will be shrinking and ageing in the coming years. Whether social security systems of the pay-as-you-go type will be under pressure, given these demographic developments, will essentially depend on productivity growth. The question arises whether an ageing workforce might negatively affect labour productivity. To disentangle the relationship between economic productivity and the age structure of the workforce, we consider the firm level as the appropriate unit of analysis. Based on our initiative, Statistics Austria set up a matched employer-employee dataset for Austria. This dataset is mainly built on the structural business statistics as well as on data from the Main Association of Austrian Social Security Institutions (Hauptverband der Sozialversicherungsträger). Covering the years 2002-2005 it comprises 19,633 firms in Austria.In a first step we investigated the correlation between the age structure of the employees and the labour productivity of firms (measured by value added per worker). Results of this descriptive analysis already indicated that the relation between productivity and the age structure at the firm level differs across economic sectors.We then applied modern panel data methods and conducted three different studies that differ by the level of aggregation at which the data were analysed. These are the firm level, the sector level and the firm level combined with the sector and the regional level.Our results indicate that the commonly expected negative relation between labour productivity of a firm and its share of elderly employees cannot be verified. On the contrary, our findings yield a negative association between the share of young employees within a firm and labour productivity. Overall we could not verify an overpayment of elderly workers either, i.e. the average wage is not significantly higher as compared to the average productivity if the share of elderly workers is higher within a firm. In summary, our results cannot confirm the hypothesis of decreasing productivity with higher ages of the workforce. Labour productivity within a firm rather seems to be determined by firm-specific characteristics as well as the regional and sectoral context in which firms operate.