This article is published in collaboration with Vox EU.
Industrial classifications tend to depict the economy as a collection of separate sectors, and draw arbitrary lines between sectors. This presentation ignores the complexity of production processes and management strategies and creates a divide between ‘manufacturing’ and ‘services’ which is stronger than it should be.1 Manufacturing firms increasingly use services in their production process, and the distinction between manufacturing and service activities has become increasingly blurry. Of course manufacturing firms produce and use services for their own account, but an increasing number of them also produce and sell services to third parties. This practice, often called ‘servitisation’ (Vandermerwe and Rada 1988) is observed in both developed and developing countries (Neely et al. 2011). The fact that manufacturing firms are no longer sole goods producers but also supply services sheds new light on the dynamics behind the deindustrialisation of most advanced economies, and on the public policies designed to ease it off (Lodefalk 2015). Economic policies that fail to take into account the dual aspect of the activities of manufacturing firms may prove less adequate.
The shift towards services within manufacturing firms is seen as a way to gain (or restore) competitiveness in both local and global markets (Wise and Baumgartner 1999). Starting to sell services is often considered a big step for industrial firms. At least it implies adding a service to the portfolio of products offered by the firm. However, engaging in the provision of services often implies deep changes in the way production and management is organised within the firm. It also entails a rethinking of the way firms approach and interact with their consumers. This is what we observe, for instance, when firms from the electronic industry provide integrated solutions. The good is only part of a much larger offer which includes complementary services (software, installation, and training), financing facilities, or insurance. The expected benefits from servitisation depend greatly on the strategy adopted by firms. Servitisation can be a way for firms to acquire additional revenues which are perceived as more stable over time and less prone to business cycle fluctuations (Ariu 2015). It can also be a way for firms to differentiate their product from their competitors’. By offering a good-service bundle, a firm’s offer is harder to imitate and perceived as more differentiated by consumers. Whatever the expected benefits, engaging in servitisation entails investments in specific assets and a deep change of organisation. Upon engaging in the provision of services, industrial firms need to acquire new sets of skills and manage very different activities. These are complicated processes whose costs are often underestimated by firms. The expected benefits from servitisation can thus be dwarfed by the costly adjustments made within the firm. Many studies have shown that servitisation is not always a winning strategy (Gebauer et al. 2005).2 However, most of the literature is based on case studies, or on limited samples of large industrial firms, and there is virtually no evidence on the benefits of servitisation based on a comprehensive dataset.3
Our recent contribution (Crozet and Milet 2015) aims to fill this gap in the literature. We use individual data on more than 50,000 French manufacturing firms over the period 1997-2007 and document the extent and evolution of the servitisation of the French manufacturing industry, and assess its consequences on firm performance.
The extent and evolution of servitisation
A simple look at the data reveals two important facts. First, a vast majority of French manufacturing firms produces and sells services to third parties. Second, this trend is steadily (although slowly) growing over time and reflects a deep transformation of the manufacturing sector, which affects firms’ performance and ultimately their competitiveness.
Figure 1 shows two measures of the extent and evolution of servitisation. In panel (a), we look at the share of ‘servitised’ firms, i.e. firms reporting selling services to third parties, in the manufacturing sector. In 2007, 77% of manufacturing firms were servitised. This share varies by sector, ranging from 60% in the food, beverage and tobacco industry to 90% in transport equipment industry. Servitised firms tend to be larger on average than firms exclusively specialized in the production of goods. In 2007, servitised firms accounted for 92% and 91% of total manufacturing value added and employment respectively. We also observe that the share of servitised firms grew slowly over the period, increasing by one percentage point over the decade for the whole manufacturing sector. In panel (b), we consider the share of services of total industry output. In 2007, services accounted for 11% of the total manufacturing sales. There is much more variation across industries in the importance of services for total sales. In 2007, 4% of the total sales of the food, beverage and tobacco industry were made up by services, while they make up to 17% of the total sales of the machinery, electrical and optical equipment industry.
Overall, the share of services of total production grew by 1.6 percentage points over the decade (1.5% annual growth rate). It actually declined in the wood, paper and printing industry (at an average rate of -4% per year) while it grew substantially in the food, beverage and tobacco industry (+4.3% per year), or in the chemical and plastics industry (+2.9% per year).
Figure 1. Extent and evolution of servitisation
- Share of servitised firms by industry
- Share of services in industry output
The effect of servitisation on firm performance
In Figure 2, we compare the profit rate (percent) and the number of employees of servitised firms with those of ‘pure goods producers’, i.e. firms that do not sell services to third parties. The figure shows the difference in terms of performance according to the share accounted for by services in total production.4 The difference in performance is quite obvious. Compared to pure goods producers, servitised firms are 3.5 percentage points more profitable (which corresponds to a 7% difference given the average profit rate in our sample). The difference is even larger in terms of employment. Manufacturing firms in which services account for less than 10% of their production sales are twice the size of pure goods producers.
Figure 2. Performance gap between servitized firms and pure goods producers
- Profit rate
- Number of employees
These results are simple correlations however, and remain silent on any causal impact of servitisation on firm performance. The decision to sell services may not be independent of firm performance, and we can expect a strong selection bias. It is likely that large firms, which are already profitable have the financial and human resources to move towards services. However, a meticulous econometric analysis that controls for the selection biases reveals a positive and significant impact of servitisation on firm performance. On average, starting to sell services increases the profit rate by between 4% and 5%, employment by 30%, and total sales by 4%. Moreover, this shift towards services does not substitute for the sales of goods, which increase by 3.5% on average. These results may seem large, especially the impact on employment. However, most of our sample of firms is made of small and medium-sized enterprises, which drive most of our results. Thus, with a median number of nine employees in our sample, an increase by 30% represents 2.7 additional jobs.
Note that we do not aim to draw any macroeconomic implication of these results, as they do not take into account general equilibrium effects (loss of competitiveness of pure goods producers, greater competition in the service market). Notwithstanding this warning, selling services as a complement to goods appears to be a way of enhancing (or restoring) competitiveness for manufacturing firms. Part of the future of manufacturing firms lies in services.
Ariu, A (2015), “Crisis-Proof Services: Why Trade in Services Did Not Suffer During the 2008-2009 Collapse”, Journal of International Economics, forthcoming.
Baumol, W J (1967), “Macroeconomics of Unbalanced Growth: The Anatomy of Urban Crisis”,American Economic Review 57(3): 415-426.
Crozet, M and E Milet (2015), “Should Everybody Be in Services? The Effect of Servitization on Manufacturing Firm Performance”, GSEM working paper series 15-10-2.
Cusumano, M A, S J Kahl, and F F Suarez (2015), “Services, Industry Evolution, and the Competitive Strategies of Product Firms”, Strategic Management Journal 36(4): 559-575.
European Commission (2014), “For a European Industrial Renaissance”, Communication from the Commission to the European Parliament, The Council, The European Economic and Social Committee and the Committee of the Regions, Brussels.
Fang, E, Palmatier, W R, J-B E Steenkamp (2008), “Effect of Service Transition on Firm Value”,Journal of Marketing 72(5): 1-14.
Gebauer, H, E Fleisch, and T Friedli (2005), “Overcoming the Service Paradox in Manufacturing Companies”, European Management Journal 23(1): 14-26.
Lodefalk, M (2015), “Servicification of Manufacturing Firms Makes Divides in Trade Policy-making Antiquated”, Orebro University, Working Paper 2015:1.
Neely, A, O Benedetinni, and I Visnjic (2011), “The Servitization of Manufacturing: Further Evidence”, paper presented at the 18th European Operations Management Association Conference, Cambridge.
Stigler J (1956), Trends in Employment in the Service Industries, National Bureau of Economic Research, General Series 59.
Vandermerwe, S and J Rada (1988), “Servitization of Business: Adding Value by Adding Services”,European Management Journal 6(4): 314-324.
Wise, R and P Baumgartner (1999), “Go Downstream in Manufacturing”, Harvard Business Review135.
1 This was underlined early on by Stigler (1956) – “There exists no authoritative consensus on either the boundaries or the classification of the service industries”.
2 The authors called this phenomenon the “service paradox”.
3 Most of the quantitative studies rely on a very specific sample of firms often restricted to one industry, or to very large firms (Cusumano et al. 2015, Fang et al. 2008).
4 We also control for annual trends and sectoral characteristics.
Publication does not imply endorsement of views by the World Economic Forum.
To keep up with the Agenda subscribe to our weekly newsletter.
Authors: Matthieu Crozet is a professor at the University of Paris / Paris Saclay. Emmanuel Milet is a post-doc fellow at the University of Geneva.
Image: Motor production worker Andrea Brooks assembles batteries for Ford electric and hybrid vehicles at the Ford Rawsonville Assembly Plant in Ypsilanti Twsp, Michigan. REUTERS/Rebecca Cook.