Differences in productivity are an important explanatory factor for differences in living standards across countries (Restuccia and Rogerson 2017). In turn, differences in productivity might arise from a genuine divergence in overall conditions of production efficiency or due to different economic structures, not only in sectoral terms, but also in the face of heterogeneous productivity across production units (Bartelsmann and Doms 2000). The latter finds its theoretical foundation in the Melitz (2003) industry model and has been empirically documented in recent years owing to the increased availability of firm-level data (ECB 2017). In this column, I focus on the firm size dimension, exploiting the observation that, despite sectoral differences (Berlingieri et al. 2018), there is an overall positive relation between firm size and labour productivity.
Firm size concerns: Limited overall but significant in southern EU
In a recent report, my co-authors and I develop a decomposition analysis that splits labour productivity differences of members states relative to the benchmark of the EU aggregate into three factors (Bauer et al. 2020: Chapter 7):
- A sectoral composition effect, which accounts for differences in employment weights across countries and the fact that productivity is higher in certain sectors than others (e.g. manufacturing compared with trade services).
- Firm size distribution within a sector. Given that larger firms are associated on average with higher productivity, this effect would be positive if larger firms concentrate more employment than the EU aggregate.
- Intrinsic differences in productivity levels across similar businesses (i.e. same sector and firm size class) in different countries.
The analysis relies on data from Eurostat’s Structural Business Statistics (SBS) for five firm size classes (fewer than ten persons employed, 10-19 persons, 20-49, 50-249, and 250 or more) within eight NACE sections representing the bulk of the business economy (manufacturing, construction, trade, transport, accommodation, ICT services, professional activities, and administrative services).
In general terms, country differences in productivity levels within each firm size class play by far the most important role in explaining the divergence across member states (Figure 1), whereas both the sectoral composition effect and the firm size distribution effect play a more limited role.
Figure 1 Labour productivity relative to the EU28 aggregate, decomposition by factor (percentage points, 2016 or latest)
Note: labour productivity is defined as the ratio of value added over persons employed; value added has been adjusted by GDP-based purchasing power parity; NACE sections included are C (manufacturing), F (construction), G (trade), H (transport), I (accommodation), J (ICT services), M (professional activities) and N (administrative services); comparable data not available for Ireland, Luxembourg, and Malta
Source: own elaboration based on SBS data
However, for a few Southern European countries, having a firm distribution tilted to smaller firms seem to be significantly detrimental for productivity performance. This is particularly the case for Greece and Italy. In Greece, the firm size distribution effect accounts for a quarter of the productivity difference with respect to the EU benchmark; in Italy, it fully offsets the positive contribution from intrinsic higher productivity levels than EU peers. The firm size distribution effect is also significantly negative in Portugal and Spain. Together with the negative sectoral effect in Spain, which is itself explained by a lower proportion of manufacturing employment, the firm size distribution effect makes up the other half of the country’s productivity gap.
Despite the fact than in recent years this negative effect has somewhat softened in these countries, it remains to be seen whether these developments could be further extended to close the existing gap. It’s still unclear whether the declining share of employment in smaller firms is associated with the aftermath of the crises (i.e. being less resilient than bigger firms) or/and the result of structural reforms supporting a higher number of enterprises.
Government and judicial efficiency to unleash firm size growth
The identification of factors contributing to productivity divergence across countries could help fine-tuning policy recommendations.
For instance, in our analysis we find that intrinsic differences in productivity levels – the main factor behind dispersion across EU member states – show a very strong country component across sectors, suggesting economy-wide obstacles for further convergence within the EU, such as educational attainment or technological absorption capacity. There are nonetheless certain cases that deviate from the country performance, pointing to the presence of specific aspects affecting productivity developments in a particular sector or firm size class.1
On the other hand, keeping in mind those countries for which we have identified that firm size distribution plays a significant role, we are particularly interested in analysing the potential determinants of the dimension of firms.
For this purpose, and using a number of metrics related to firm size, we first estimate the country and sector effects for each indicator. Then we focus on explaining the country effects with a broad set of variables, ranging from country economic size and economic development to educational attainment and dimensions of the institutional framework – such as regulation, labour relations, or government and judicial efficiency.
The results are in line with findings from previous related literature (Kumar et al. 1999), confirming the significant role played by the institutional framework in shaping firm size distributions, with judicial and government efficiency playing a particularly supportive factor for increasing firm size. In this sense, Southern European countries, where firm distributions seem to be significantly detrimental to productivity, underperform in aspects such as regulatory quality or protection of property rights.
There are a number of interesting areas for future research. One way to extend the explanatory variable of firm size indicators would be to include the so-called correlated distortions – i.e. a positive relation between productivity and the degree of misallocation, which were found to have a detrimental effect on firm size by Bento and Restuccia (2018). We could also analyse the deviations from the expected value of firm size metrics (i.e. once country and sector effects have been removed). These deviations could stem from specific factors affecting a particular sector in one country (such as regulation, including size-dependent policies as studied by Guner et al. 2008), or from the interaction of country and sector determinants, as found by Kumar et al. (1999) for patent protection and R&D intensity or judicial efficiency and capital per worker.
Bartelsman, E J and M Doms (2000), “Understanding productivity: Lessons from longitudinal microdata”, Journal of Economic literature, 38(3): 569-594.
Bauer, P, I Fedotenkov, A Genty, I Hallak, P Harasztosi, D Martínez-Turégano, D Nguyen, N Preziosi, A Rincón-Aznar and M Sánchez-Martínez (2020), Productivity in Europe: Trends and drivers in a service-based economy, JRC Technical Report.
Bento, P and D Restuccia (2018), “On Average Establishment Size across Sectors and Countries”, National Bureau of Economic Research, Working Paper No. 24968.
Berlingieri, G, S Calligaris and C Criscuolo (2018), “The productivity-wage premium: Does size still matter in a service economy?”, OECD, Science, Technology and Industry Working Papers No. 2018/13.
ECB (2017), “Firm heterogeneity and competitiveness in the European Union”, Article 2, Economic Bulletin, Issue 2.
Guner, N, G Ventura and Y Xu (2008), “Macroeconomic implications of size-dependent policies”, Review of Economic Dynamics 11(4): 721-744.
Kumar, K B, R G Rajan, and L Zingales (1999), “What determines firm size? National Bureau of Economic Research”, Working Paper No. 7208.
Melitz, M J (2003), “The impact of trade on intra‐industry reallocations and aggregate industry productivity”, Econometrica 71(6): 1695-1725.
Restuccia, D and R Rogerson (2017), “The causes and costs of misallocation”, Journal of Economic Perspectives 31(3): 151-74.
Syverson, C (2011), “What determines productivity?”, Journal of Economic Literature 49(2): 326-65.
1 In this vein, we recommend Syverson (2011) for a comprehensive work on why businesses differ in their productivity levels, ranging from internal to external factors. More specifically, Restuccia and Rogerson (2017) and Bento and Restuccia (2018) explore the role of misallocation in accounting for productivity differences.