Many large and successful firms sell both goods and services. Apple sells software and assistance with using its computers and mobile phones, Toyota provides both cars and loans to consumers buying these cars, Technip supplies fertilisers as well as technical and financial solutions for their use. This pattern of firms associating services with their products is particularly strong among manufacturing firms (Breinlich and Criscuolo 2011, Kelle 2013, Ariu 2016). Yet, economists and policymakers generally consider goods and services as two distinct sectors subject to their own market adjustments, calling for specific policies.
In a recent paper (Ariu et al. 2020), we challenge the view that goods and services are two independent items in the consumer portfolio, supplied by firms in separate industries. Thanks to a unique dataset on Belgian firms’ exports, we show that the most successful firms are actually those that sell both goods and services together.
Moreover, we show both empirically and theoretically that part of their success in international markets can be attributed to the provision of services. Specifically, the presence of a service increases the demand for goods supplied by the same firm, leading to higher sales.
Who are the firms that sell both goods and services? How do they compare with other manufacturing firms?
We exploit uniquely detailed trade data from the National Bank of Belgium that record exports of both goods and services at the firm, product (service), and destination levels. These detailed data allow us to study the interaction between services and goods at the firm level. We first provide a portrait of firms that export both goods and services – which we call ‘bi-exporters’ – and see how they compare to firms that sell only goods or services.
Bi-exporters represent only 8.8% of the firms that export;1 few exporters can actually provide both goods and services. However, they account for about 44.8% of overall goods exports and 42.3% of services exports, showing that these few firms are very successful in international markets and account for the lion’s share of Belgian exports.
Trade is not the only margin along which bi-exporters prove to outperform other firms that export only goods or services. They are larger in terms of sales, employees, and product and destination portfolios, and they are also often more productive and more likely to be multinationals.
We also find that bi-exporters almost never export services in destinations in which they do not provide goods, but the reverse does not hold: bi-exporters frequently sell goods in destinations where they do not provide services. Moreover, when exported to the same destination, services represent only a minor fraction of the total export flow (39.3% on average). Within the same destination-country and year, bi-exporters sell more goods than pure good exporters but less services than pure service exporters.
These results reveal an asymmetry in the relationship between goods and services within the same firm: the production of goods is the essential activity of bi-exporters, while services are typically an optional item in their export activities. Bi-exporters do ‘better’ than pure good or pure service exporters only in the provision of goods and not in the provision of services.
The question then arises: why do exporters in the manufacturing sector that also provide services perform better?
Do services increase goods exports?
To quantify the export premium of bi-exporters, we perform two exercises. Comparing firms that export the same product in the same market, we find that those that provide services together with the good export, on average, 58% more than those that do not. Exploiting variation across markets within the same firm, we find that the same firm sells 27% more of a given good in destinations where it also provides services, as compared to destinations where it does not. These premiums hold when we control for firm and market characteristics that could affect both the fact that firms provide services in a destination and their export performance for manufacturing goods in that destination.
However, it could still be the case that unobserved factors affect both firms’ decision to export services in a given destination and their manufacturing sales in that market. To correct for this potential bias and achieve causal inference, we use exogenous variations across firms in terms of ‘bundleability’ of their goods portfolio – i.e. how much the products in the firm's portfolio can be associated with services – as well as variations across destinations in the freeness of services trade.
The results confirm the causal positive effect of services provision on firm-level goods export performance in a destination. By breaking down exports into quantities and prices (unit values), we show that the positive effect is a combination of both: when firms provide services together with their goods, they can raise the price of their good without reducing the quantity they sell.
Note that this is the price of the good alone; the service is subject to a transaction in its own right in the data. This means that services are a determinant of quality, i.e. the same product produced by the same firm has a different quality evaluation depending on the presence (or not) of services. Therefore, services act as a demand shifter for the goods by making the product more appealing to consumers. Intuitively, the extra-warranty or the maintenance services offered together with a computer or an engine can effectively make them look better than without.
Overall, the effect of services provision on manufacturing sales is quite large. Based on our empirical results, it appears that up to 11.8% of overall Belgian manufacturing exports and up to 22.4% of the manufacturing exports of bi-exporters are triggered by the provision of services.
Mechanisms: Complementarity, market power and ‘quality’
Our results can hardly be obtained through standard trade models without positing ex ante that services are only exported with high-quality goods. We rather rationalise our empirical results in a theoretical framework featuring imperfect competition with asymmetric complementarity between goods and services. Using the terminology of Chen and Nalebuff (2006), we consider goods and services as one-way complements. Goods are essential to consumers’ consumption while services are optional.
In our framework, a good alone and the same good with the service are perceived as two different products, so that at a given price, the demand for the good is higher when a complementary – and costly – service is exported along with the good. This is equivalent to increasing the perceived quality of the good. Through this mechanism, large firms that are able to supply such complementary services may increase their market share. This translates unambiguously in an increase in bi-exporters’ mark-ups, eventually raising profitability.
Goods and services have been generally analysed as two independent outputs provided by separate firms and different industries. This also shows up both in the way trade agreements are negotiated and designed and in the way economists model and quantify their effects.
Considering goods and services separately in trade agreement negotiations is likely to miss part of the business and welfare gains and losses related to these treaties. Indeed, our results complement Ariu et al. (2019) by suggesting that the liberalisation of trade in services, which was and is at stake in many bilateral negotiations (such as those between the EU and the UK for Brexit), might also have important consequences for trade in goods in general and, in particular, for the biggest firms that are bi-exporters.
In the case of Brexit, ignoring such complementarities between goods and services will lead to an underestimation of the negative economic consequences. The identification of firms and sectors that will be mostly harmed by a hard Brexit, as well as its aggregate impact on the British economy, can hardly miss this channel, given the size of the service sector in the UK.
While the increasing participation of manufacturing firms in service activities is often seen as going hand in hand with deindustrialisation, our work provides a different perspective by showing that the two are not necessarily antagonistic. Both policymakers and economists are concerned about the decreasing share of the manufacturing sector, especially in developed countries. Our work shows that, paradoxically, complementarities between goods and services might in reality have slowed down the secular decline of the manufacturing sector.
Ariu, A (2016), “Services versus goods trade: A firm level comparison”, Review of World Economics 152(1): 19–41.
Ariu, A, H Breinlich, G Corcos and G Mion (2019), “The interconnections between services and goods trade at the firm-level”, Journal of International Economics 116: 173–88.
Ariu, A, F J Mayneris and M Parenti (2020), “One way to the top: How services boost the demand for goods”, Journal of International Economics 123.
Breinlich, H, and C Criscuolo (2011), “International trade in services: A portrait of importers and exporters,” Journal of International Economics 84(2): 188–206.
Chen, M K, and B J Nalebuff (2006), “One-way essential complements”, Cowles Foundation Discussion Papers 1588, Cowles Foundation for Research in Economics, Yale University.
Markus, K (2013), “Crossing industry borders: German manufacturers as services exporters”, The World Economy 36(12): 1494–515.
1 60.1% of firms export only goods and 31.1% export only services.