VoxEU Column Institutions and economics International trade

Services trade restrictiveness, economic governance, and manufacturing productivity

There is some evidence suggesting a positive effect of services trade liberalisation on the productivity of manufacturing. This column argues that such an effect is conditional on the institutions within a country. Countries with better economic governance benefit more from open services trade policies. Higher quality institutions attract more productive services providers and support higher levels of services performance.

Services trade restrictions and productivity

Services play a crucial role as intermediate inputs for modern manufacturing. Trade is an important channel through which firms can improve their access to services, either in the form of lower prices or greater choice (more variety). Recent research has identified a positive effect of services trade liberalisation on the performance of manufacturing sectors that are downstream in the relevant supply chain.1 Given that services trade restrictions in many countries and sectors are often high, liberalisation is one mechanism to foster productivity improvements and economic growth.

An important question is whether reducing services trade restrictions is a sufficient condition to trigger higher productivity in downstream manufacturing, independent of initial conditions that are reflected in levels of economic development (per capita incomes) and the institutional environment that prevails in a country. A key feature of most services is that they are intangible and not storable. This has implications for both services production and exchange. Non-storability gives rise to a proximity burden – the agent providing a service must be in the same location as the buyer or consumer. Accordingly, exporters of services often must perform some stages of their economic activity in the importing country, and thus will be affected by local regulations and the prevailing business environment, i.e. the quality of economic governance and related institutions.2 These factors can shape the downstream impacts of services trade policy. Good economic governance – proxied by variables such as control of corruption, the quality of regulation, and strong rule of law – is likely to be a necessary condition for an economy to fully benefit from services trade liberalisation.3

The role of institutions on services trade: New empirical findings

In a new paper (Beverelli et al. 2015), we analyse the conditioning role of institutions using industry-level data for 58 countries spanning all stages of economic development. Information on services trade policies comes from the World Bank's Services Trade Restrictiveness Database, which contains policy measures for 5 major services sectors, distinguishing between policies that affect inward investment and cross-border supply of services. Following the literature, we build a composite indicator of services trade restrictiveness for each country that takes into account the intensity of input-output linkages between the five services activities and downstream manufacturing sectors.

A simple plot of manufacturing labour productivity against our composite policy indicator suggests that the quality of economic governance institutions conditions the effects of services trade restrictions on downstream industries. In Figure 1, light dots are manufacturing sectors in countries lying above the sample median of institutional performance (measured as control of corruption); dark dots are manufacturing sectors in countries lying below this threshold. In the case of countries with better economic governance, the (solid) regression line is negatively sloped, with a statistically significant coefficient of -0.112. This first cut at the evidence suggests that, ceteris paribus, less restricted services trade policies are associated with higher productivity in manufacturing industries (at the average input intensity level) in countries with better institutional quality. Conversely, for countries with low institutional performance the slope of the (dashed) regression line is not statistically different from zero.

Figure 1. Services trade restrictiveness and labour productivity in manufacturing

Econometric analysis confirms this conclusion.

  • The impact of lower services trade restrictions is highly conditional on the quality of economic institutions, a relationship that is statistically significant and robust.

What matters is the quality of economic governance; our results indicate that the relationship is not simply a function of levels of economic development as reflected in differences in per capita incomes. Our analysis allows for a quantification of the causal linkage from upstream services policies to downstream manufacturing productivity across countries with different quality of institutions. Taking two high-income countries with a similar level of services trade restrictiveness, Austria and Italy, and applying to them the more open services trade policy regime of the UK, the estimation results suggest that the average productivity of Austrian manufacturing firms would be some 18% higher, while in Italy this effect would amount to only a 7% increase.4

These results reveal that the significant differences in economic governance quality that prevails across countries with similar levels of per capita income matter for the likely benefits of services trade liberalisation. Austria ranks 6th in our estimation sample in terms of the institutional quality variable used in this quantification (control of corruption), while Italy ranks 25th. For countries with particularly weak institutional settings (the four lowest deciles in the distribution of institutional quality) the effect of an analogous shift in services trade policy is not statistically different from zero. Very similar results are obtained with a number of commonly used indicators of economic governance.

Why the sensitivity to institutional quality?

What are the micro-level mechanisms that make institutions such a relevant conditioning factor for the effect of services trade policy? We propose two broad mechanisms, which are further detailed in a simple theoretical model in our paper. The key feature underpinning both mechanisms is the need for foreign firms to be physically present in the importing market in order to sell services.

  • First, for a given level of trade restrictiveness implied by policy, a country’s institutional environment may affect entry decisions of potential foreign suppliers, giving rise to a selection effect of institutions.

Consider for instance a global provider of telecommunication services, Vodafone. This firm has a direct presence in 21 markets (i.e. it has established a physical presence through investment), and an indirect presence in 55 markets through partnerships with local firms.5 Of these 76 markets, 19 (25%) are countries with relatively low institutional quality (measured by the control of corruption variable being less than the sample median), while the other 57 (75%) are countries with relatively high institutional quality. If we consider the markets where Vodafone is not present, either directly or in partnership with a local provider, 87 out of 142 (61%) are countries with relatively low institutional quality and 55 (38%) are countries with relatively high institutional quality.6 Regression analysis suggests that after controlling for country size (GDP) and for the degree of services trade restrictiveness in the telecommunications sector, institutional quality has a positive and statistically significant effect on the probability of Vodafone entering a market by establishing either a direct or indirect commercial presence.

  • The second mechanism is conditional on trade decisions. Once a provider has committed to supply services to a foreign market, the quality of the service provided may depend on the institutional environment that prevails in the country where demand is located and the service is performed.7

In this case institutions have a performance effect. Both the selection and the performance effects provide a theoretical underpinning for the conditioning role of institutions on the effect of services trade policy on downstream sectors’ productivity.


Services trade liberalisation may not be sufficient to trigger higher productivity in downstream manufacturing. Given that services often must be traded through establishment of a commercial presence in a market, the quality of economic governance that prevails in the importing country matters for (potential) foreign services providers. Our analysis suggests that the economic payoffs to liberalisation of services transactions will be higher the better is the economic governance that prevails. We know that institutions matter for growth. They seem to matter in particular for the benefits of services trade reform.

Author’s note: The opinions expressed in this column should be attributed to its authors. They are not meant to represent the positions or opinions of the WTO and its Members and are without prejudice to Members' rights and obligations under the WTO.


Acemoglu, D, S Johnson, and J A Robinson (2005), “Institutions as a fundamental cause of long-run growth,” in Philippe Aghion and Steven N. Durlauf (eds.), Handbook of economic growth, Elsevier: 385-472.

Arnold, J M, B Javorcik, and A Mattoo (2011), “Does Services Liberalization Benefit Manufacturing Firms? Evidence from the Czech Republic,” Journal of International Economics 85(1): 136-46.

Arnold, J M , B Javorcik, and A Mattoo (2015), “Services Reform and Manufacturing Performance. Evidence from India,” Economic Journal, At: DOI: 10.1111/ecoj.12206.

Barone, G, and F Cingano (2011), “Service Regulation and Growth: Evidence from OECD Countries,” Economic Journal 121(555): 931-57.

Beverelli, C, M Fiorini and B Hoekman (2015), “Services Trade Restrictiveness and Manufacturing Productivity: The Role of Institutions,” CEPR Discussion Paper No. 10834 and EUI Robert Schuman Centre for Advanced Studies Working Paper 2015/63.

Bourlès, R, G Cette, J Lopez, J Mairesse, and G Nicoletti (2013), “Do Product Market Regulations in Upstream Sectors Curb Productivity Growth? Panel Data Evidence for OECD Countries,” Review of Economics and Statistics 95(5): 1750-68.

Dollar, D, M Hallward-Driemeier, and T Mengistae (2005), “Investment Climate and Firm Performance in Developing Economies,” Economic Development and Cultural Change 54(1): 1-31.

Francois, J and B Hoekman, 2010, “Services Trade and Policy,” Journal of Economic Literature, 48(3): 642-92.

Rodrik, D, A Subramanian, and F Trebbi (2004), “Institutions Rule: The Primacy of Institutions Over Geography and Integration in Economic Development,” Journal of Economic Growth 9(2): 131-65.

Winters, L A, and A Masters (2013), “Openness and Growth: Still an Open Question?,” Journal of International Development 25(8): 1061-70.


1 See Arnold et al. (201, 2015) for country-specific studies of the effect of services trade liberalisation using firm-level data. Francois and Hoekman (2010) survey much of the relevant literature.

2 Barone and Cingano (2011) and Bourlès et al. (2013) show that the quality of regulation prevailing in a country impacts on the performance of affected industries.

3 The quality of economic institutions has been found to be a key determinant of economic development – see e.g., Acemoglu et al. (2005) and Rodrik et al. (2004); institutional quality has also been found to condition the benefits of economic reforms like trade liberalisation (e.g. Winters and Masters 2013).

4 The differences in services trade restrictiveness for EU members illustrates that the EU is not yet a customs union when it comes to services trade policy.

5 Data compiled from the Vodafone web page: http://www.vodafone.com/content/index/about/about-us/where.html.

6 A test of equality of means rejects the null hypothesis that the probability of Vodafone's commercial presence is the same in the two groups of countries with low and high institutional quality (106 countries each), in favour of the alternative hypothesis that such probability is higher in the group of countries with high institutional quality.

7 See for example Dollar et al. (2005).

1,260 Reads