AdobeStock_182073050.jpeg
VoxEU Column Industrial organisation Productivity and Innovation

Technology adoption via machine imports: Identifying who learns from peers

In less developed countries, upgrading production technologies by importing machinery is an important source of growth. Using new firm-level data from Hungary for the period 1992-2003, this column finds that firms are more likely to import a particular piece of sector-specific machinery when other local firms previously imported the same machine. A similar pattern holds regarding the choice of the machine’s source country. These benefits are concentrated in large and foreign-owned companies, while small and domestically owned firms may actually be adversely affected.

A key aspect of improving economy-wide performance in less developed countries is upgrading production technologies by importing machinery, the latter being an essential source. As trade liberalisation progresses and the economy opens up, many firms will be set to upgrade their production technologies. To decide which machine to purchase, many will look to other firms to learn from their experience. We report evidence on the importance of knowledge spillovers and discuss which types of firms are more likely to benefit from available information on the market. Understanding what kind of firms may benefit from spillovers is important for policies aimed at promoting learning from peers.

Capital goods, machines and manufacturing technologies are produced only in a few economies. Countries which do take part in developing these technologies can benefit from them via knowledge spillovers as suggested by endogenous growth theories. For developing countries that do not produce manufacturing technology themselves, a key vehicle for spillovers and growth are imports. Indeed, Coe and Helpman (1995) and Acharya and Keller (2009) found large spillover effects from imports from foreign, R&D-abundant countries on domestic productivity at the aggregate and sector levels. Importing technology embedded in machines and materials leads to increased productivity also at the level of the firm, as shown by Halpern et al. (2015). 

One channel of learning how to upgrade technology for a firm is to learn from others. Bisztray et al. (2018a, 2018b)  looked at very localised importing networks in Hungary and showed how firms may learn about which country to import from, showing evidence for spillovers. Similar evidence was shown for exporting decisions by several recent papers such as Fernandes and Tang (2014) or Koenig et al. (2010) for China and France, respectively.

In a recent paper (Békés and Harasztosi 2019), we looked at whether firms’ decision to import a specific machine is influenced by the local accumulation of experience regarding that same imported type of machine. As for the machines, we considered over 200 different kinds of production equipment. One example for a specific type of machine could be knitting machines and stitch-bonding machines (Standard International Trade Classification 72452) for the manufacturing of textile products. In particular, we investigated how investment in a particular machine may be encouraged by earlier imports of the same machine carried out by local firms. As more and more local firms import a particular machine, it becomes easier for another firm to be informed about the advantages and the specifics of the technology. In the absence of peers, a firm would be less inclined to import a given machine or it would import it much later.

To study possible spillover effects, we compiled a dataset that matches machine-level import observations to Hungarian manufacturing firms for 1992-2003. The period provides several advantages. It starts with Hungary’s early transition years, prior to which foreign machines were not generally available to domestic firms. Possibly, every machine imported in the early 1990’s can be regarded as technologically more modern and more advanced than the previously installed machines. In addition, the transition invited waves of foreign direct investment, which introduced new imported machines and technology to many sectors. This is not only true for greenfield investment, but also for a portion of the privatised companies as well, where firms upgraded their production facilities through imports. In the examined period, foreign machines indeed play an important role in manufacturing investments. 

Our analysis is based on a narrow geographical level – with over 3,000 local units defined as postal districts. Figure 1 shows the distribution of the total number of machines imported in each location over the sample period. In over 40 districts, more than 50 machines get imported. These are predominantly located in larger townships in Hungary. In about 100 districts, we see imports between more than 25 but less than 50 machines, and over 670 districts have firms importing less than 25 but more than 5 machines. In the remaining districts, numbering a bit more than 1,500, local firms import 5 machines or less.

Figure 1 Number of imported machines per postal district

We modelled the effect of peer presence on the probability that a firm at a given location in time t chooses a new imported machine from the set of machines it has not imported before t as a linear hazard. The unit of observation is a firm-machine-time triple. Spillover effects were modelled as a set of count variables indicating the number of other firms importing a particular machine before time t within 1 km to our firm. To study the role of distance, we added additional variables counting firms within 5, 15 and 30 km. 

To control for the average propensity of a firm to import and that of a machine to be imported, we added firm and machine specific interactions with time as there could be many potential confounding factors, such as local policies, local time variant business cycles, and machine-, sector- and year-specific shocks such as new transportation links. To capture these potential confounders, we estimated linear probability models with fixed effects – firm*year, machine*year, sector*year, location*year, machine*sector*year and location*sector*year. 

Our results indicate that the presence of a previous importer of a specific machine in the close vicinity increases the probability of a firm importing the same machine. The presence of one additional peer within 1 km of the firm increases import probability by 0.27 percentage points. In terms of magnitude, this difference corresponds to about one quarter of the baseline probability of a machine import. As expected, distance plays a key mediating role – there is a decaying spillover effect as learning takes place for firms mostly within 15 km. Firms, especially in small cities, learned mostly from neighbouring (i.e., within 1 km) peers. We also found that even for different types of imported machines, the source country of the machine matters a great deal – confirming results for Budapest firms by Bisztray et al. (2018).

To better understand the nature and extent of the spillovers, we looked at whether they vary according to key features of the importing firm as well as the composition of peers.

First, we divided our sample into different groups by size, according to the attributes of the importing firm. For the largest firms, each additional peer company increases import probability by 1.12 percentage points, while the results for medium-sized firms indicate an increase of 0.40 percentage points. In contrast, the smallest firms are unaffected by the presence of an additional peer importer. We repeated this exercise by ownership, by export status and by age of the firm. For each group, we estimated our model. Figure 2 shows regression coefficients and 95% confidence intervals.

Figure 2 Spillover coefficient estimates by attributes of the importer

Note: Regression coefficients from a linear probability model. Peer variables are measured with count variables. Coefficients are multiplied by 100 to express percentage points.

Our results show a strong variation in the spillover coefficients. Some types of firms will not benefit, on average, from local peers – small firms and firms serving only domestic markets will be less likely to import a machine when their peers import a machine in their close proximity. At the same time, young and internationalised firms are the ones that really reap the benefits of proximity to importing peers. These results are consistent with the existing literature on spillover heterogeneity (for FDI projects, such heterogeneity was shown by Békés et al. 2009, while Cardoso-Vargas (2017) looked at variations for exports). 

Some firms may emit a stronger signal than others making our firms more likely to start importing a machine. To investigate, we look into how the heterogeneity of peers affects the strength of spillover effects. Figure 3 shows a similar exercise, where we examine heterogeneity in terms of importing peers. We see a similar pattern, but estimates are closer to each other. 

Figure 3 Spillover coefficient estimates by attributes of peers

Note: Regression coefficients from a linear probability model. Peer variables are measured with count variables. Coefficients are multiplied by 100 to express percentage points.

Overall, we found that the extent of spillover effects varies a great deal both with respect to the importing firm and, to a lesser degree, with respect to the composition of peers. Larger or foreign-owned and internationalised firms are the ones that benefit from having importing firms in their vicinity, while small and domestic market-oriented firms could actually be adversely affected by peer effects.

Our results could be indicative for policymakers interested in the indirect impact of technology upgrade subsidy programs. We found that such indirect effects do exist. However, they are centred on large-to-large firm interactions. As smaller sized firms producing for the domestic market do not benefit much from import spillovers, policies aimed at helping such firms may not be able to rely on these indirect effects. 

References

Acharya, R C and W Keller (2009), “Technology transfer through imports”,  Canadian Journal of Economics 42(4): 1411-1448.

Békés, G, J Kleinert and F Toubal (2009), “Spillovers from multinationals to heterogeneous domestic firms: Evidence from Hungary”, The World Economy 32(10): 1408-1433.

Békés, G. and P Harasztosi (2019), “Machine imports, technology adoption and local spillovers”, Forthcoming in the Review of World Economics and CEPR discussion paper 13623.  

Bisztray, M, M Koren and A Szeidl (2018a), “Learning to import from your peers”, Journal of International Economics 115(C): 242-258. 

Bisztray, M, M Koren and A Szeidl (2018b), “Learning to import from your peers”, VoxEU.org, 18 November. 

Cardoso-Vargas, C-E (2017), “Does the type of neighbor matter? Heterogeneous export spillovers on domestic companies in Mexico”, Estudios Económicos 32(2): 255-292.

Coe, D T and E Helpman (1995), “International R&D spillovers”, European Economic Review 39(5): 859-887.

Fernandes, A P and H Tang (2014), “Learning to export from neighbors”, Journal of International Economics 94(1): 67-84.

Halpern, L, M Koren and A Szeidl (2015), “Imported inputs and productivity”, American Economic Review 105(12): 3660-3703.

Koenig, P, F  Mayneris and S Poncet (2010), “Local export spillovers in France”, European Economic Review 54(4): 622-641.

1,680 Reads