Agencies that provide support to exporting firms were in place in at least 103 countries as of 2005 (Lederman et al. 2010). In this column, I survey recent academic findings on the effects of export promotion policies using firm-level observational data.1 Export promotion policies for the purpose of discussion here refer to those implemented by public export promotion agencies (EPAs) in the form of direct support to exporting firms, excluding subsidisation of production and investment, which could be export promotion measures in a broader sense. Specific types of support include offering information on destination countries and export procedures, assisting firms in participating in overseas market survey missions and trade fairs, and supporting firms with their negotiations with importing firms.
I focus here particularly on the problem of identifying the causal effects of export promotion measures and the approaches used to address it. For instance, a positive correlation between a firm's use of an export promotion measure and the value of its exports may be reflecting the actual causal effect of the export promotion policy on exports, or it may simply be the consequence of reverse causality, i.e. firms that are willing to export are apt to take advantage of export promotion support (self-selection bias). Typical approaches to addressing the identification problem in empirical analysis of export promotion policies include an instrumental variables estimation, a difference-in-differences estimation, propensity score matching, estimation using a sub-sample in which self-selection bias is limited, and controlling for fixed effects or numerous variables. For a more comprehensive survey, see Van Biesebroeck et al. (2016).
Recent literature analysing the effects of export promotion policies includes a multitude of studies that have found a positive effect of export promotion policies on firms' export behaviour using firm-level data in each country.2 Let's take a specific look at some of the studies. Munch and Schaur (2018) analyse the effects of export promotion policies on firms' export behaviour, export values, profits, labour productivity, and so on, using data on export support services provided by the Denmark Trade Council. In order to determine the causal effect of export promotion measures on firms' behaviour, they select a control group of firms comparable to those that receive support services by using propensity score matching and then analyse the differences between the two groups in performance changes over time in a difference-in-differences approach. In addition, they examine the effect of export support services on those firms that are approached by the Trade Council, a subgroup of the treatment group. This analysis is premised on the following assumption: Trade Council officials select firms to approach based on observable firm characteristics (which are also observable to econometricians) – for example, selecting firms belonging to industries that are showing significant growth in local markets overseas – and thus, those firms' participation in the support programme is considered random once such observable variables are controlled for. As a result of this analysis, they find that 1) export support services increase the probability of firms being exporters in the year they receive such services by an average 3.9 percentage points; and 2) the probability increases by 5.9 percentage points two years after receiving such services.
Broocks and Van Biesebroeck (2017) analyse the effects of export support services on Belgian firms' exports to countries outside the EU, using data provided by Flanders Investment and Trade (FIT), a Belgian EPA. In analysing the causality between export support services and firms' export behaviour, they control for observable characteristics of firms. Furthermore, in order to verify the robustness of the results, they perform the same analysis by restricting the sample to firms with 20 or more employees. This additional analysis is based on the presumption that self-selection bias is relatively small in this subsample of firms because, in a small economy like Belgium, advancing into foreign markets is vital for firms above a certain size. As a result, they find that export support services increase the probability of exporting by an average of 8.5 percentage points. As such, empirical studies using firm-level data generally find a positive effect of export promotion measures.
Effects by firm characteristics
What has been discussed above is about the average effects of export promotion policies. However, those measures should have different effects on individual firms depending on their characteristics. Indeed, many studies have actually shown some differences in the pattern of effects depending on firm characteristics. For instance, regarding the effects of export promotion measures by firm size, some studies show that export promotion measures have a particularly strong positive effect on small firms (Munch and Schaur 2018; Broocks and van Biesebroeck 2017, Volpe Martincus and Carballo 2010b, Volpe Martincus et al. 2012), whereas some other studies find a particularly positive effect on medium-sized firms (Olarreage et al. 2015, Kim et al. 2016). There are also studies that examine the relationship between product types and the effectiveness of export promotion policies, with the conclusion that the measures are particularly effective on firms exporting complex and differentiated products (Volpe Martincus and Carballo 2010a; Volpe Martincus and Carballo, 2010b). It has been also shown that export promotion measures have a greater impact on firms entering the export market for the first time or attempting to expand into new markets, rather than those that are already established as regular exporters (Munch and Schaur 2018, Volpe Martincus and Carballo 2008, 2010a, 2010c Volpe Martincus et al. 2011). These findings are consistent with hypotheses represented by various theoretical models, particularly one that considers barriers to information access as an obstacle to entering the export market.
Effects by type of support
As described above, there exist many studies that have examined the effects of export promotion policies, but very few have gone so far as to analyse the effects characterised by the types of support given to firms. Volpe Martincus and Carballo (2010d), Broocks and Van Biesebroeck (2017), and Munch and Schaur (2018) are among the very few examples of such studies. Volpe Marticus and Carballo (2010d) analyse the effects of different forms of export promotion, using panel data provided by Proexport Colombia. Specifically, they group export promotion measures into 1) "counselling", which includes providing information on export destinations and training in export procedures; 2) "trade agendas", which refer to assisting with arrangements for business meetings and negotiations; 3) "trade fairs, shows and missions", which provide support for firms participating in trade fairs, shows, and missions; and 4) combinations of these types, and compared their effects. The problem here is distinguishing between two possible relationships. That is, when a certain firm succeeds in exporting goods, is it because the firm has taken advantage of type A support (e.g. trade fairs, shows and missions), or is it simply due to the fact that firms likely to succeed in exporting goods are apt to choose to take advantage of type A support? In order to address this identification problem, they select a control group of firms similar to those in the treatment group by using propensity score matching, and then analyse the differences between the two groups' changes in performance over time in a difference-in-differences approach. They find that firms supported by all types of export promotion programs in a bundle show more pronounced growth in terms of the total value of exports and the number of export markets than those supported by individual programs.
Broocks and Van Biesebroeck (2017) also conduct an additional analysis in which they compare the effectiveness of different types of export promotion services, restricting the sample to those firms supported by such services. They classify services into "questions" in which firms make inquiries about matters requiring information analysis and so on; "actions", which refers to events and seminars; "subsidies" that assist firms in making business trips, participating in trade fairs, arranging meetings with distributors (by subsidizing costs incurred); and "communication". which includes those services that do not fall into any of the above. According to their findings, firms that receive "subsidy" support have a 4.6 to 8.4 percentage point higher probability of exporting than those that receive "question" and "communication" support. Lastly, Munch and Schaur (2018) classify support services used by firms into two types – "partner search and matchmaking" and "intelligence and analysis" – to examine respective effects of each type of support services, as an additional analysis to the above-discussed average effects. According to their findings, small firms' probability of exporting increases by 9.4 percentage points two years after receiving the former type of support services, and by 6.7 percentage points in the case of the latter.
These findings suggest that bundled services combining multiple types of support – and the more direct measures, such as helping firms find distributors – are more effective.
Broocks, A and J Van Biesebroeck (2017), "The Impact of Export Promotion on Export Market Entry”, Journal of International Economics, 107: 19-33.
Lederman, D, M Olarreaga, and L Payton (2010), "Export Promotion Agencies: Do They Work?" Journal of Development Economics, 91: 257-265.
Makioka, R (2019), “Are Export Promotion Measures Effective? A Survey”, RIETI Policy Update.
Munch, J and G Schaur (2018), "The Effect of Export Promotion on Firm-Level Performance”, American Economic Journal: Economic Policy, 10(1): 357-387.
Olarreaga, M, A Sperlich, and V Trachsel (2015), "Export Promotion: What Works?" Working Paper.
Van Biesebroeck, J, E Yu, and S Chen (2015), "The Impact of Trade Promotion Services on Canadian Exporter Performance”, Canadian Journal of Economics, 48(4): 1481-1512.
Van Biesebroeck, J, J Konings, and C Volpe Marticus (2016), "Did Export Promotion Help Firms Weather the Crisis?" Economic Policy, 31(88): 653-702.
Volpe Martincus, C (2010), Odyssey in International Markets: An Assessment of the Effectiveness of Export Promotion in Latin America and the Caribbean, Special Report on Integration and Trade, Inter-American Development Bank, Washington, DC.
Volpe Martincus, C and J Carballo (2008), "Is Export Promotion Effective in Developing Countries? Firm-Level Evidence on the Intensive and the Extensive Margins of Exports”, Journal of International Economics, 76: 89-106.
Volpe Martincus, C and J Carballo (2010a), "Entering New Country and Product Markets: Does Export Promotion Help?" Review of World Economy, 146: 437-467.
Volpe Martincus, C and J Carballo (2010b), "Beyond the Average Effects: The Distributional Impacts of Export Promotion Programs in Developing Countries”, Journal of Development Economics, 92: 201-214.
Volpe Martincus, C and J Carballo (2010c), "Export Promotion Activities in Developing Countries: What Kind of Trade Do They Promote?" IDB Working Paper Series No. IDB-WP-202.
Volpe Martincus, C and J Carballo (2010d), "Export Promotion: Bundled Services Work Better”, The World Economy, 33(12): 1718-1756.
Volpe Martincus, C, J Carballo, and A Gallo (2011), "The Impact of Export Promotion Institutions on Trade: Is It the Intensive or the Extensive Margin?" Applied Economics Letters, 18: 127-132.
Volpe Martincus, C, J Carballo, and P M Garcia (2012), "Public Programmes to Promote Firms' Exports in Developing Countries: Are There Heterogeneous Effects by Size Categories?" Applied Economics, 44: 471-491.
 The original RIETI article on which this column is based (Makioka 2019) also reviews some theoretical predictions, empirical studies using country- and regional-level data, and those using randomized controlled trials (RCTs).
 The list of these studies includes Munch and Schaur (2018) for Denmark; Broocks and Van Biesebroeck (2017) for Belgium; Van Biesebroeck et al. (2015) for Canada; Volpe Martincus and Carballo (2010a) for Uruguay; Volpe Martincus et al. (2012) for Argentina; Volpe Martincus and Carballo (2008) for Peru; Volpe Martincus and Carballo (2010b) for Chile; and Volpe Martincus and Carballo (2010c, 2010d) for Costa Rica.