Do firms learn by exporting or learn to export? Evidence from small and medium-sized enterprises

Abstract

Using a matching approach, we compare the productivity trajectories of future export-entrants and matched nonentrants. Future exporters have higher productivity than do nonentrants before entry into international markets, which indicates self-selection into exports. More interestingly, we also observe a productivity increase among export-entrants relative to nonentrants before export entry. This might be explained by higher investments in physical capital prior to export entry. We find no evidence that the productivity gap between export-entrants and nonentrants continues to grow after export entry. Our results suggest that learning to export occurs but that learning by exporting does not. In contrast to previous studies on Swedish manufacturing, we focus particularly on small and medium-sized enterprises (SMEs).

1 Introduction

Numerous studies have documented that exporters enjoy higher productivity than do nonexporters within the same industry, controlling for observed factors that may affect productivity.¹ In the literature, two nonexclusive explanations have been put forward to explain such export productivity premia: self-selection and learning by exporting.

Self-selection means that only the more productive firms can afford the higher cost of exporting. This implies that future exporters have significantly higher productivity than do nonexporters before they start exporting; productivity for future exporters is higher ex ante. Most previous empirical studies have found support for self-selection.

Learning by exporting, on the other hand, should result in superior postentry productivity performance in new export-entrants relative to nonentrants. The reason might be that exporters are exposed to knowledge flows from international buyers and competitors and to more intense competition in international markets, which lead to larger opportunities and incentives to improve productivity than experienced by firms that sell only on the domestic market. Moreover, the exploitation of economies of scale and improved capacity utilization in connection with export entry could also be manifested in better postentry productivity performance in new export-entrants than in nonexporters.² However, in contrast to self-selection, the empirical evidence for any positive postentry effects of exports and for learning by exporting are mixed.³

An interesting possible explanation for the self-selection pattern identified by most previous empirical studies has been proposed by Alvarez and Lopez (2005). They argue that firms consciously increase their productivity by investing in physical and human capital and new technology with the explicit purpose of becoming exporters. The investments involve pre-entry improvements in productivity among future export-entrants; they learn to export rather than learning by exporting, and those learning effects are neither inevitable nor automatic.

Distinguishing between learning by exporting and learning to export among new export-entrants is an important aim of this paper. Toward that end, we exploit a large-scale panel dataset including all Swedish manufacturing firms with one or more employee during the period between 1997 and 2006. Access to detailed longitudinal firm-level data allows us to use modern econometric matching techniques, which means that we can solve potential endogeneity problems and evaluate the causal effect of export activities on firm performance.

According to the learning-by-exporting hypothesis, one would expect that the effect of exporting on productivity should occur at the time when firms enter international markets and should then give rise to a widening productivity gap between export-entrants and continuing nonexporters. In a standard matching approach, like the one we carry out at first, the postentry productivity of export-entrants and that of nonexporters with similar pre-export productivity histories and similar values for other pre-export covariates are compared. Such an approach does not allow for learning to export, which implies that, preceding entry into the export market, productivity increases for new export-entrants relative to nonexporters. To test the learning-to-export hypothesis requires a different matching strategy where the baseline for similar pre-export productivity (and other covariates) instead is set several years before the period of export entry, thus permitting the effect on productivity of exporting to appear even before the new export-entrants enter international markets.

Matching methods have been employed with Swedish data before. Greenaway et al. (2005) use a panel of manufacturing firms spanning almost 20 years from 1980 to 1997. However, their data include only firms with 50 or more employees.⁴ Export participation among such firms is quite high in Swedish manufacturing (more than 80%). Therefore, it is not surprising that they found that “in Sweden productivity growth of exporters on entry does not appear to differ significantly from non-exporters either in the periods leading up to or after entry” (Greenaway et al. 2005, p. 578). We obtain a similar result for this group of firms for a more recent period. However, the outcome appears to differ considerably for smaller—and from a policy perspective perhaps more interesting—firms.⁵ The fact that the export participation rate is significantly lower in smaller firms and that productivity is higher in exporting firms than in nonexporting firms⁶ is occasionally presented as a motive for intensified export promotion, particularly in small and medium-sized enterprises (SMEs).

To preview our findings, we observe an instantaneous productivity increase at export entry among the entering firms relative to nonentering firms and that thereafter, in subsequent periods, the productivity gap is constant. If we allow for different productivity trajectories before export entry for future export-entrants and for firms not entering the export market, we notice a significant productivity differential between them even before export entry. We also find that increased investments in physical capital precede entry into export markets, and this may explain the higher productivity among future export-entrants. Our results are largely driven by the smaller firms and are consistent with the learning-to-export hypothesis but to a lesser extent with the learning-by-exporting hypothesis.

The remainder of the paper is organized as follows. Section 2 presents our dataset and gives some descriptive facts and preliminary evidence regarding exports and productivity of Swedish firms. Section 3 describes our econometric strategy. Section 4 reports the results of the analysis. Section 5 concludes.

2 Data and description

2.1 Exporting and exporters in Swedish manufacturing

The data for this study come from Statistics Sweden’s firm-level data on export of goods and financial accounts.⁷ The data on firms’ exports of goods provide information on which types of products, and to which countries, a given firm was exporting during the period 1997 to 2006. For exports to EU countries, there is a threshold value for the registration of exports, while all transactions are registered by Swedish Customs for exports to countries outside the EU. The threshold value has risen over the studied period; before 1998, the yearly value of exports to EU countries had to be larger than 0.9 million SEK; between 1998 and 2004, the requirement was 1.5 million SEK or more; and after 2004, it was 4.5 million SEK or more. Due to this threshold for the registration of goods exported to the EU, and to avoid considering firms with very limited sales outside the EU during a single year as exporters, we define a firm as an exporter if it has an export value larger than 1.5 million SEK.

From Statistics Sweden’s compilation of figures from the financial accounts of enterprises, we obtain balance sheet information such as sales, value added, and employment. We link the data on the export of goods at firm level to the balance sheet information for firms with at least one employee operating in the Swedish manufacturing sector (NACE Rev. 1 15–36). This gives an unbalanced panel of firms that contains information on the included firms’ export status at every point in time. This means that we can identify whether a firm is a domestic producer, an export-entrant, a continuing exporter, or a firm that has quit exporting. Capital stocks are book values from the balance sheets. Value added is deflated with Swedish producer price indices (PPI) at industry level. As our productivity indicator we use labor productivity.

Sweden is a small export-dependent economy. The aggregate export intensity (the share of exports in sales) for manufacturing was 64% in 2006. Nevertheless, there are large variations in export participation rates and export intensities between firms of different sizes. Table 1 shows that the share of exporters is considerably larger among medium-sized and large firms (those with 50 or more employees, among which more than 80% of firms are exporters) than among small and micro firms. This is one reason why we focus our analysis of export entry on firms that have fewer than 50 employees. A similar pattern appears for export intensity, the number of export destination countries, and the number of export products; larger firms tend to have higher export intensity and to export more products to more destination countries.

Table 1 Share of exporters, export intensity, and number of export destination countries and export products among micro, small, and medium-sized and large manufacturing firms in 2006

How important are firms with fewer than 50 employees in terms of employment and value added in Swedish manufacturing, and what is their contribution to goods exports? From Table 2, it appears that firms with fewer than 50 employees represent a quarter of the employment in the Swedish manufacturing sector and less than a fifth of the value added, while their share of goods exports is significantly lower, not even 7%. Micro and small firms employ a fair share of those working in manufacturing, while their share of exports is quite low.

Table 2 Share of employment, value added, and exports for firms of different sizes in 2006

As pointed out in Sect. 1, a very robust result from most of the previous analyses of the relationship between exports and productivity at firm level is that exporters are more productive than nonexporters. It is evident from Table 3 that our study is no exception. Including industry dummies and firm controls, as in specification (3), substantially reduces the exporter productivity premium in comparison with specifications (1) and (2). However, the premium is still larger than 10% and strongly significant.⁸ If, as in specifications (4) to (6), we estimate the premium for different firm size classes, the value is highest for firms with fewer than 10 employees (micro firms) and lowest, and actually insignificant, for firms with 50 or more employees. In addition, we find that, except in the case of micro firms, the larger the firm’s export intensity, the higher its productivity.⁹

Table 3 Exporter productivity premia, 1997–2006

2.2 The dataset for analysis and descriptive statistics

An important aim of the study is to investigate the productivity trajectories of firms that start exporting, before and after they enter the export market, and compare them with the trajectories of firms not entering the export market. Toward this end, we use the unbalanced panel of manufacturing sector firms with at least one employee to construct a balanced panel of export-entrant and non-export-entrant firms observed for every year during a 7-year time window. The 7-year time window is used because we want to be able to examine all firms 3 years before and 3 years after potential export entry. We define export-entrants as firms that exported in year t but did not export in the years t − 3 to t − 1, whereas nonentrants are defined as firms that did not export in any of the years t − 3 to t. Given that our data cover the period from 1997 to 2006, the first year of potential export entry is 2000 (where export data for the period 1997–2000 are used to classify firms). The last year of potential export entry is 2003 (which allows for a 3-year follow-up period during 2004–2006).

With these conditions, we end up with a balanced panel of firms made up of four cross-sections with potential export entry in 2000, 2001, 2002, and 2003 and with time windows of 7 years for each cross-section. In the analysis, we compare firms entering the export market (treated firms) in a given year with firms not entering the export market (untreated firms) in the same year, and we follow the firms during the 7-year time window. In our panel, the total number of observations of export-entrants is 724, and the total number of observations of nonentrants is 44,120. The 724 observations of export-entrants represent unique firms. With the 7-year time window and the conditions applied, there is no possibility that a firm classified as an export-entrant in, for example, 2000 will subsequently reappear as an export-entrant. Only 14,483 of the 44,120 observations of nonentrants represent unique firms. The reason is that, if a firm is identified as a nonentrant in 2000, it might once again be classified as a nonentrant in 2001, and so on. In Sect. 4.4, we refine the classification of export-entrants and nonentrants depending on the firms’ export status not only in the years t − 3 to t but also in the years t + 1 to t + 3. This will enable us to study the importance of whether export-entrants continue to export or later on leave the export market and, similarly, whether nonentrants eventually enter international markets or continue not to export. Table 4 presents some descriptive statistics for our dataset, where we divide the firms into different size classes and classify them as either export-entrants or nonentrants.

Table 4 Sample means for export-entrants and nonentrants in different firm size classes

Table 4 shows that export-entrants have higher capital intensity (physical as well as human capital intensity) than do nonentrants the year before potential export entry. This holds true for micro and small firms, i.e., firms with fewer than 50 employees, but not for medium-sized and large firms. Furthermore, export-entrants are larger, have more employees, and are more often part of a multinational enterprise (MNE).

Regarding our outcome variable, labor productivity, Table 4 indicates that export-entrants have higher productivity than do nonentrants even 3 years prior to potential export entry, which implies that more productive firms appear to become exporters (self-selection). Moreover, the productivity gap tends to widen during the 7-year time window. In other words, export-entrants are inclined to improve their performance relative to nonentrants in connection with their export entry. However, if we divide the firms into different size classes, these patterns are valid only for micro and small firms, not for medium-sized and large firms. Hence, the descriptive statistics in Table 4 produce some interesting distinctions in terms of productivity differentials and productivity trajectories between export-entrants and nonentrants, especially for firms with fewer than 50 employees. Nevertheless, to obtain more direct and reliable evidence regarding the relationship between export entry and firm productivity requires a careful econometric analysis.

3 Econometric strategy

One main purpose of this paper is to estimate the causal effect on firm productivity of starting to export. The majority of studies focusing on this question have been dominated by different types of regression-based methods.¹⁰ Recently, some papers have been published that employ matching methods.¹¹ While regression and matching approaches are both based on conditional independence for drawing causal inference, there are a few differences between the approaches that are more than cosmetic. First, matching does not rely on the type of functional form assumptions that regression typically relies upon. Second, matching is more explicit in assessing whether or not comparable untreated observations are available for each treated observation. Current econometric research suggests that avoiding functional form assumptions and imposing a common support condition can be important for reducing selection bias in studies based on observational data.¹² In this section, we give a brief sketch of how matching solves the evaluation problem and discuss some specific features when implementing matching in our particular context.¹³

To begin, let t ⁻ and t ⁺ indicate time periods before and after a period of potential export entry t ⁰. Furthermore, let $D_{{t^{0} }} = 1$ denote that a firm starts to export in period t ⁰ and $D_{{t^{0} }} = 0$ indicate that a firm does not start to export in period t ⁰ (starting to export is equivalent to receiving “treatment” in the typical evaluation terminology). Moreover, let ${\text{LP}}_{{1t^{ + } }}$ be the potential labor productivity in period t ⁺ for firms that start to export in period t ⁰ and ${\text{LP}}_{{0t^{ + } }}$ be the potential labor productivity in period t ⁺ for firms that do not start to export in period t ⁰. Finally, let $X_{{t^{ - } }}$ denote a set of observed covariates affecting both export entry and productivity.

The main parameter of interest is the average treatment effect on the treated, ATT, which can be defined as

$$ {\text{ATT}} = E\left( {{\text{LP}}_{{1t^{ + } }} - {\text{LP}}_{{0t^{ + } }} } \right.\left| {\left. {D_{{t^{0} }} = 1} \right) = E\left( {{\text{LP}}_{{1t^{ + } }} } \right.\left| {\left. {D_{{t^{0} }} = 1} \right) - E\left( {{\text{LP}}_{{0t^{ + } }} } \right.\left| {\left. {D_{{t^{0} }} = 1} \right)} \right.} \right.} \right. $$

(1)

In this specific context, ATT corresponds to the average effect on labor productivity of export entry for firms that actually start to export. The fundamental evaluation problem is that we only observe ${\text{LP}}_{{1t^{ + } }}$ or ${\text{LP}}_{{0t^{ + } }}$ for each firm, but never both. $E\left( {{\text{LP}}_{{1t^{ + } }} } \right.\left| {\left. {D_{{t^{0} }} = 1} \right)} \right.$ can be estimated directly from the observed data. Missing is the information required to estimate $E\left( {{\text{LP}}_{{0t^{ + } }} } \right.\left| {\left. {D_{{t^{0} }} = 1} \right)} \right.$, referred to as the counterfactual outcome. If export entry is nonrandom and we substitute the unobservable $E\left( {{\text{LP}}_{{0t^{ + } }} } \right.\left| {\left. {D_{{t^{0} }} = 1} \right)} \right.$ for the observable $E\left( {{\text{LP}}_{{0t^{ + } }} } \right.\left| {\left. {D_{{t^{0} }} = 0} \right)} \right.$ when estimating ATT, we end up with selection bias equal to $E\left( {{\text{LP}}_{{0t^{ + } }} } \right.\left| {\left. {D_{{t^{0} }} = 1} \right)} \right. - E\left( {{\text{LP}}_{{0t^{ + } }} } \right.\left| {\left. {D_{{t^{0} }} = 0} \right)} \right.$.

In experimental studies, randomization in a sense makes the counterfactual a factual. In observational studies, some assumptions must be made to eliminate the selection bias. The method of matching solves the evaluation problem by assuming that, conditional on $X_{{t^{ - } }}$, ${\text{LP}}_{{0t^{ + } }}$ is independent of $D_{{t^{0} }}$:

$$ {\text{LP}}_{{0t^{ + } }} \,\bot\, D_{{t^{0} }} \left| {X_{{t^{ - } }} } \right. $$

(2)

This is referred to as the conditional independence assumption (CIA). The intuition behind this crucial assumption is that it makes treatment assignment random conditional on $X_{{t^{ - } }}$, which in a sense ex post reproduces the essential feature of a randomized experiment. When the CIA holds, we can therefore use the productivity of firms not making export entry as an approximation of the counterfactual outcome (the productivity that firms making export entry would have experienced had they not started to export). Heckman et al. (1998) show that, for an unbiased estimation of ATT, it is only necessary to assume mean conditional independence:

$$ E\left( {{\text{LP}}_{{0t^{ + } }} } \right.\left| {\left. {X_{{t^{ - } }} ,D_{{t^{0} }} = 1} \right) = E\left( {{\text{LP}}_{{0t^{ + } }} } \right.\left| {\left. {X_{{t^{ - } }} ,D_{{t^{0} }} = 0} \right)} \right.} \right. . $$

(3)

The type of cross-sectional matching estimator described above assumes that conditioning on the set of observed covariates $X_{{t^{ - } }}$ is sufficient to remove selection bias. However, if there are unobserved characteristics affecting treatment assignment and outcomes, this will violate the identification conditions that justify cross-sectional matching. It has been shown that, under these circumstances, the time-invariant portion of the remaining selection bias can still be eliminated by using a conditional difference-in-differences (DID) matching estimator.¹⁴ The conditional DID matching strategy requires that

$$ E\left( {{\text{LP}}_{{0t^{ + } }} - {\text{LP}}_{{0t^{ - } }} } \right.\left| {\left. {X_{{t^{ - } }} ,D_{{t^{0} }} = 1} \right) = E\left( {{\text{LP}}_{{0t^{ + } }} - {\text{LP}}_{{0t^{ - } }} } \right.\left| {\left. {X_{{t^{ - } }} ,D_{{t^{0} }} = 0} \right)} \right.} \right. . $$

(4)

Whereas the cross-sectional matching estimator assumes that conditioning on the observed covariates is sufficient to remove bias in the posttreatment period, the conditional DID matching estimator assumes the same cross-sectional bias in the pre- and posttreatment period, so that by differencing the before–after differences for export-entrants and non-export-entrants, the time-invariant bias will be removed. The conditional DID matching strategy extends the conventional matching method because it does not require that selection bias is eliminated by conditioning on the observed covariates, only that the bias is the same in the pre- and posttreatment period.¹⁵

Furthermore, both the conventional and the DID matching method rely on a common support or overlap condition that for ATT can be formally stated as:¹⁶

$$ \Pr \left( {D_{{t^{0} }} = 1} \right.\left| {\left. {X_{{t^{ - } }} } \right) < 1} \right. .$$

(5)

This condition prevents $X_{{t^{ - } }}$ from being a perfect predictor of treatment status. In our context, this ensures that, for every $X_{{t^{ - } }}$, there are firms choosing to start to export and firms choosing not to start to export, which means that, for every $X_{{t^{ - } }}$, we will be able to construct the counterfactual outcome. When $X_{{t^{ - } }}$ has high dimension (i.e., includes continuous variables or discrete variables with many values), it becomes difficult to find comparable observations along all dimensions of $X_{{t^{ - } }}$. Rosenbaum and Rubin (1983) have shown that, if matching on $X_{{t^{ - } }}$ is valid, so is matching on the conditional probability of receiving treatment, referred to as the propensity score. The propensity score reduces the dimensionality of the matching problem by allowing us to match on a scalar function of the covariates rather than the entire covariate space.

All matching estimators are weighting estimators in the sense that they take a weighted average of the outcomes of the untreated observations to construct an estimate of the unobserved counterfactual for each treated observation. For ATT, the cross-sectional (CS) and the DID version can be written in the form

$$ {\text{ATT}}_{\text{CS}} = {\frac{1}{{n_{1} }}}\sum\limits_{{i \in \{ D_{{t^{0} i}} = 1\} }} {\left[ {{\text{LP}}_{{1t^{ + } i}} - \sum\limits_{{j \in \{ D_{{t^{0} j}} = 0\} }} {w\left( {i,j} \right){\text{LP}}_{{0t^{ + } j}} } } \right]} , $$

(6)

$$ {\text{ATT}}_{\text{DID}} = {\frac{1}{{n_{1} }}}\sum\limits_{{i \in \{ D_{{t^{0} i}} = 1\} }} {\left[ {\left( {{\text{LP}}_{{1t^{ + } i}} - {\text{LP}}_{{1t^{ - } i}} } \right) - \sum\limits_{{j \in \{ D_{{t^{0} j}} = 0\} }} {w\left( {i,j} \right)\left( {{\text{LP}}_{{0t^{ + } j}} - {\text{LP}}_{{0t^{ - } j}} } \right)} } \right]}, $$

(7)

where $n_{1}$ is the number of treated observations and $w(i,j)$ is the weight placed on the jth comparison observation in constructing the counterfactual for the ith treated observation. The primary difference between alternative matching estimators is how they construct the weight, which typically involves a trade-off between bias and variance. For instance, in single nearest-neighbor matching, each treated observation i is matched to the nearest (in terms of propensity score) comparison observation j, with the weight given by $w\left( {i,j} \right) \in \left\{ {1,0} \right\}$. Single nearest-neighbor matching trades reduced bias for increased variance (using additional neighbors would raise bias due to increasingly poorer matches but decrease variance because more information would be used to construct the counterfactual for each treated observation). In the empirical work, we will consider two alternative weighting regimes: single nearest-neighbor matching and kernel matching based on the Epanechnikov kernel. For the latter, we will employ different bandwidths covering a fairly wide interval. Increasing the bandwidth will generally increase bias and reduce variance, because heavier weight will be put on more distant observations when constructing the counterfactual for each treated observation (i.e., the effect of increasing the bandwidth is similar to that of using additional neighbors in nearest-neighbor matching).

There are a few specific circumstances to consider when implementing matching in our particular context. The first is related to the two aforementioned principal explanations for why export firms enjoy higher productivity. According to the learning-by-exporting hypothesis, the effect of exporting on productivity should occur once the firms enter international markets, not before. To test this hypothesis, we either compare postexport productivity for export-entrants and nonentrants with similar pre-export productivity histories and similar values for other pre-export covariates (the cross-sectional case) or compare the before–after differences in productivity for export-entrants and nonentrants conditional on other pre-export covariates (the conditional DID case).

Under the alternative learning-to-export hypothesis the effect of exporting on productivity may occur before firms actually enter international markets. The argument is that firms make a deliberate effort to increase their productivity by investing, for instance, in human and physical capital and new production technologies and products with the explicit intention of becoming exporters. In this case, pre-export productivity is not treated as exogenous (as in the typical self-selection hypothesis); instead, it is regarded as endogenous with respect to the decision to enter international markets. Here, the “treatment” in focus should be a firm’s conscious decision to try to become an exporter rather than the actual realization of export sales (as in the case of learning by exporting). However, the former is unobservable, so for practical reasons we are left with the observable cases where the intent to become an exporter is successfully accomplished.¹⁷ We test the learning-to-export hypothesis by examining the evolution of productivity and capital accumulation before entry into international markets. In this case, we use a matching approach where we only require that future export-entrants and nonentrants have similar productivity and other attributes in a specific initial pre-export reference period. This approach allows export-entrants and nonentrants to experience divergent development in terms of productivity and other firm attributes after the initial period.

Consequently, in our empirical work, we will consider model specifications where (i) export is allowed to affect productivity at the time of firms’ export entry and thereafter, and (ii) export is permitted to influence capital accumulation and productivity even before firms enter international markets.

A second circumstance that warrants special attention has to do with dynamics in firms’ export status. Some of the firms that enter the export market will continue to export (entrant-stayers), while others will cease to export (entrant-stoppers). Similarly, some of the nonentrants will continue not to export (never-entrants), while others will eventually enter international markets (not-yet-entrants). In Sect. 4, we will examine how robust our results are with regard to changes in firm export status.

Although the analysis of different types of subgroups is uncomplicated as such, it is important to recognize how the construction of the various samples may change the interpretation of the results from an econometric perspective. For instance, if we anticipate a positive effect of export entry and choose to narrow the treatment group to entrant-stayers (instead of using all export-entrants, including entrant-stoppers) this will induce an upward bias in the estimated treatment effect. All firms that for one reason or another fail to endure as exporters will be disregarded, even though export failure should be viewed as part of the overall causal effect of export entry rather than being considered as exogenous with regard to the treatment. Similarly, if we refine the comparison group to consist of never-entrants (instead of using all nonentrants, including not-yet-entrants) and continue to expect a positive effect of entering the export market, we will once again end up with upward bias in the estimated treatment effect. The problem here is that we try to transform what is actually a process of dynamic treatment assignment (where some firms choose to enter the export market early, others decide to go in later, and some prefer to never enter) into a static one (where firms once and for all decide whether or not to enter).

In both cases above, the definition of the treatment and comparison group involves conditioning on the future and therefore produces samples that are selective in terms of the outcome of interest. It is beyond the scope of this paper to present any formal methodological solutions to these problems.¹⁸ We merely want to emphasize that the conditioning in the subsample analysis introduces bias with regard to the typical treatment parameter in question and actually leaves us with a set of different treatment parameters with slightly different interpretations.

4 Empirical results

We begin the presentation of our results in Sect. 4.1 and discuss the estimates of the propensity scores used in the following matching analyses. Among other things, these estimates indicate whether firms self-select into the export market. Then, in Sects. 4.2 and 4.3, we report estimates of the causal effects of export entry on firm labor productivity. In Sect. 4.2, we use specifications that restrict productivity to be affected at the time of export entry and thereafter. In Sect. 4.3, we employ specifications that allow productivity and also capital accumulation to be influenced even before export entry takes place. The latter gives us an indication of whether the temporal pattern of investments can help explain pre-entry development of productivity. Finally, in Sect. 4.4, we show the outcome of some robustness checks where we refine the export-entrant and nonentrant groups in different ways.

4.1 Propensity scores and self-selection

In this section, we present estimates of the propensity scores (i.e., the probability of starting to export) that will be used in the matching analyses to follow. The covariates included in the propensity scores are standard variables suggested by theory and previous empirical literature to affect both export entry and future productivity. These include physical capital per employee (K/L), share of employees with postsecondary education (H/L), size in terms of employment (EMP), a dummy variable indicating whether a firm is part of a multinational enterprise (MNE), two-digit NACE industry dummies (21 industries), and dummies for the year of potential export entry. In addition, the propensity scores for the cross-sectional specifications include pre-export labor productivity (LP). For the conditional DID specifications, labor productivity prior to potential export entry is not included as a covariate in the propensity scores but is instead used to construct the before–after potential export entry differences.

The specification of the propensity scores further differs for the matching models focusing on the learning-by-exporting and learning-to-export hypotheses. In the former case, we are seeking to find export-entrants and nonentrants that are as similar as possible all the way up to the period of potential export entry. These sets of propensity scores therefore include labor productivity for a 3-year period prior to potential export entry ($t - 3$ to $t - 1$), while the other covariates refer to the year prior to potential export entry ($t - 1$). In the specifications focusing on the learning-to-export hypothesis, all covariates refer to the third year prior to potential export entry ($t - 3$).¹⁹ The latter specifications thus allow for export-entrants and nonentrants to experience divergent development in terms of labor productivity and other firm attributes during the years up to potential export entry (i.e., during $t - 2$ and $t - 1$).²⁰

In all cases, we use a probit model to estimate the propensity scores. To the extent that interactions and higher orders of the covariates improved the balancing between export-entrants and nonentrants, they were included. For brevity, we will focus on the linear terms for the most important variables and further restrict the presentation to the cross-sectional specifications.²¹

Table 5 presents estimates of the propensity scores pertaining to the cross-sectional learning-by-exporting specification. Beginning with the first column, which gives the results for all firms irrespective of size, we find that the probability of becoming an export-entrant seems to increase with pre-export labor productivity. However, this result only holds in $t - 1$. Due to high correlation between productivity in the different years, it is difficult to obtain precise estimates for each year. To avoid the problem of multicollinearity, we have experimented with a specification that instead includes average labor productivity over the years $t - 3$ to $t - 1$. The result (not reported in the table) indicates a highly significant and positive effect of pre-export labor productivity on the probability of export entry. These results are thus in line with the self-selection hypothesis that more productive firms enter international markets. Furthermore, the results show that more capital-intensive firms (in terms of physical capital as well as human capital) tend to become exporters, and that the same applies to larger firms and firms that are part of multinational enterprises.

Table 5 Estimated propensity scores for the cross-sectional learning-by-exporting specification

However, if we look at the results for firms of different sizes, the positive effect of pre-export labor productivity on the probability of becoming an exporter appears to be valid only for micro firms (firms with fewer than 10 employees).

Turning to Table 6, which shows estimates of the propensity scores for the cross-sectional learning-to-export specification, we find more or less similar results. One notable difference is that the positive effect of pre-export labor productivity on the probability of export entry also seems to hold for small firms (firms with 10–49 employees).

Table 6 Estimated propensity scores for the cross-sectional learning-to-export specification

In sum, our estimates of the propensity scores reveal some interesting patterns in terms of self-selection of firms into international markets. Our results indicate that the self-selection hypothesis—that more productive firms enter the export market—primarily applies to micro firms and to some extent to small firms, but not to medium-sized and large firms.

4.2 Learning by exporting

In this section, we continue by presenting the propensity score matching estimates of the causal effect of export entry on labor productivity. The estimates are obtained using both cross-sectional (Eq. 6) and conditional difference-in-differences (DID) matching (Eq. 7). In both cases, we have applied two different weighting regimes: single nearest-neighbor matching and kernel matching based on the Epanechnikov kernel. For the latter, we have used bandwidths in the interval [0.001, 0.01].²² For brevity, we will only report results based on the Epanechnikov kernel using a bandwidth of 0.005.²³ Details regarding the specification of the propensity scores are provided in the previous section.

To begin, we focus on model specifications pertaining to the learning-by-exporting hypothesis. The estimates are thus based on export-entrants and nonentrants with similar pre-export entry firm attributes up to year $t - 1$, and which we follow during the years $t$ to $t + 3$. Table 7 presents the differences in log labor productivity between export-entrants and matched nonentrants. These estimates can be interpreted as the approximate percentage effects of export entry on labor productivity. Approximate standard errors are reported in parentheses below each estimate.²⁴

Table 7 Matching estimates of the effect of export entry on labor productivity (learning-by-exporting specification)

In Table 7, we can see that firms that become exporters increase their productivity by the time of export entry t relative to matched firms that do not enter as exporters at t. The percentage effect on labor productivity of export entry is 5.4% or 4.2% depending on the estimator (CS or DID). Interestingly, the effect is fairly stable over time and is about the same at year $t + 3$. When we look at the results for different firm sizes, it becomes apparent that the productivity effect of export entry is larger and statistically more significant for smaller firms. Furthermore, the estimates based on cross-sectional matching tend to be larger than those based on DID matching.

Table 7 also reports some aggregate balancing indicators that give a sense of how successful the matching has been in terms of balancing differences in the covariates between export-entrants and nonentrants. The first is the mean standardized bias over all covariates used in the propensity scores, which is between 12% and 18% before matching and between 1% and 5% after matching.²⁵ On average, the matching generates a reduction in mean bias by roughly a factor of 10. The other indicator is the pseudo R ² before and after matching. This statistic indicates how well the variables in the propensity score explain the probability of receiving treatment. After matching, the pseudo R ² should be fairly low because there should be no systematic differences in the distribution of covariates between the treatment and the comparison group. Before matching, this statistic is between 0.08 and 0.18. After matching, it drops to virtually zero. In sum, the balancing indicators suggest that the matching has been fairly successful in terms of balancing differences in the covariates between the treatment and the comparison group. In fact, after matching, there remain no statistically significant differences in the means for the pre-export firm attributes of export-entrants and nonentrants.

Figure 1 illustrates the results of the cross-sectional matching estimates for different firm sizes in Table 7. Here, we notice the instantaneous productivity increase at export entry t for export-entrants with fewer than 10 employees and the constant, 10–15% productivity gap in the subsequent years relative to the matched firms that do not enter international markets at t. For the larger firms, the productivity increase is much smaller and, in most cases, statistically insignificant. Figure 1 (and Table 7) indicates that there is a positive impact on productivity at the time of entry among smaller firms entering the export market. However, with the reservation that the postexport period is rather short (3 years), there does not seem to be any evidence of continuous learning through exporting. For this, we would have expected to see a widening productivity differential over time. The fairly stable gap might instead indicate more of a static productivity effect due to increased potential for economies of scale following export entry. Finally, looking at the pre-export productivity differentials, they tend to be close to zero and are statistically insignificant for all firm sizes. This can be regarded as providing additional support that we are actually comparing comparable export-entrant and nonentrant firms.

Fig. 1

Cross-sectional matching estimates of the effect of export entry on labor productivity (learning-by-exporting specification). Notes: Based on the cross-sectional estimates in Table 7. Filled data marker indicates effect significant at the 10% level or lower

4.3 Learning to export

So far we have presented results that compare export-entrants at t with nonentrants at t with similar pre-export entry firm attributes up to $t - 1$. As we pointed out before, this approach is primarily designed to test the hypothesis of learning by exporting. By definition, such a strategy precludes any impact of exporting on productivity taking place before firms enter international markets; any productivity differences prior to export entry between future exporters and firms not entering the export market are balanced in the matching. Exporting may only affect productivity at the time of export entry or after it has taken place. To test the hypothesis of learning to export, we have to allow for export-entrants and nonentrants to experience divergent development in terms of labor productivity and other firm attributes even before the time of potential export entry.

In this section, we present estimates based on export-entrants and nonentrants at t that have similar labor productivity and other firm attributes at $t - 3$ but for which the trajectories of these attributes may differ thereafter. In addition, we report results from a similar analysis but where the focus is on capital accumulation instead of productivity. This will provide some insights on whether the development of investments possibly can help explain the pre-entry path of productivity.

Table 8 reveals that there is a significant productivity differential already at $t - 1$ between export-entrants and nonentrants at t with similar productivity and other firm attributes at $t - 3$. Moreover, the productivity gap continues to grow to 8.8% (CS) or 5.7% (DID) at $t + 1$, and thereafter, the gap is basically constant. When we focus on the results for different firm sizes, we again find that the rising productivity differential is driven by the smallest firms, i.e., those with fewer than 10 employees. Looking at the balancing indicators, we also find that the matching has been quite successful in terms of balancing differences in the covariates between export-entrants and nonentrants.

Table 8 Matching estimates of the effect of export entry on labor productivity (learning-to-export specification)

Figure 2 visualizes the estimates from the cross-sectional matching for different firm sizes in Table 8. In the initial period $t - 3$, the productivity differentials are close to zero and statistically insignificant for all firm sizes. This indicates successful matching. After the initial period, but still prior to actual export entry, we observe a growing productivity gap between small export-entrants and small nonentrants. This development continues until export entry takes place. In the postentry period, the gap remains essentially constant. We are not able to observe any similar pattern among the firms in the other size classes.

Fig. 2

Cross-sectional matching estimates of the effect of export entry on labor productivity (learning-to-export specification). Notes: Based on the cross-sectional estimates in Table 8. Filled data marker indicates effect significant at the 10% level or lower

According to the learning-to-export hypothesis, firms prepare to become exporters by making different types of investments. If this hypothesis is correct, we should be able to observe that firms who enter international markets accumulate more capital before entry than do firms that produce exclusively for the domestic market. Figure 3 present cross-sectional matching estimates of the effect of export entry on physical capital accumulation.²⁶ In the baseline period $t - 3$, future export-entrants and nonentrants have almost identical amounts of physical capital per employee. Immediately after the initial period, we observe a rapidly increasing gap in the capital intensity between future export-entrants and nonentrants. This especially holds for micro firms. After a short halt around the time of export entry, the gap continues to widen throughout the period. Also, for small firms, we find a steadily growing difference in capital intensity between export-entrants and nonentrants, although the gap is only statistically significant during the postentry phase. For the larger firms, we find no differences in capital intensity.

Fig. 3

Cross-sectional matching estimates of the effect of export entry on physical capital per employee (learning-to-export specification). Notes: Based on the cross-sectional estimates in Table 10. Filled data marker indicates effect significant at the 10% level or lower

Our interpretation of the pattern shown in Figs. 2 and 3 (and of the findings in Table 8 and Table 10 in the Appendix) is that smaller firms, at least, appear to prepare themselves for entering the export market. The results suggest that increased investments in physical capital precede entry into export markets, and these investments may explain the higher productivity among future exporters before they enter international markets. In other words, firms seem to learn to export.²⁷ However, one caveat is that the fairly high threshold value for the registration of exports (see Sect. 2.1) means that some of the smaller entering firms in particular actually might have been exporters already in $t - 2$ and $t - 1$. Since the timing of export entry plays such an important role in our analysis, we have to acknowledge that this might be a potential problem.

4.4 Robustness to dynamics in export status

In this section, we will continue by examining how robust our results are with regard to changes in firms’ export status. Remember that we defined export-entrants (treated firms) as firms that exported in year $t$ but did not export in the years $t - 3$ to $t - 1$, whereas nonentrants (untreated firms) were defined as firms that did not export in any of the years $t - 3$ to $t$. With this approach, we are most likely mixing export-entrants that continue to export, often referred to as export successes, with those firms that cease to export, so-called export failures. Similarly, we are mixing nonentrants that continue not to export with those that eventually enter international markets. In this section, we proceed by estimating productivity effects of export entry using a more detailed classification of firms’ export status.²⁸

We divide our treated firms into two subgroups: export-entrants that continue to export throughout the period $t + 1$ to $t + 3$ (entrant-stayers) and export-entrants that leave the export market during at least one of the years $t + 1$ to $t + 3$ (entrant-stoppers). We also split our untreated firms into two subgroups: nonentrants that continue to stay out of the export market throughout the period $t + 1$ to $t + 3$ (never-entrants) and nonentrants that eventually enter the export market during the period $t + 1$ to $t + 3$ (not-yet-entrants).

Table 9 presents statistics on the export status types for the different firm size classes. The majority of our export-entrants exit the export market during at least one of the years following entry. The share of stoppers decreases with firm size. Two-thirds of the entrants in the micro firm category (1–9 employees) stop exporting, whereas four out of ten entrants stop in the medium-sized and large (50–∞ employees) firm category. Looking at the nonentrants group, there seems to be considerably less dynamics going on, in particular in the smaller firm size classes. Only 1.5% of the nonentrants in the micro firm class eventually enter the export market (98.5% belong to the never-entrants category) compared with 19% of the nonentrants in the medium-sized and large firm class (81% belong to the never-entrants category). Note that, due to the earlier mentioned threshold value for the registration of exports (see Sect. 2.1), we are not able to assess to what extent the changes in export status type are a result of major swings in firms’ export values or a consequence of smaller fluctuations around the threshold.

Table 9 Export status types by firm size

In this section, we are particularly interested in comparing the estimated effect of export entry on labor productivity for entrant-stayers relative to never-entrants versus that of entrant-stoppers relative to nonentrants. In a sense, the former comparison is the most distinct classification of treated/untreated firms, whereas the latter is less clear-cut.

Before turning to the results, we would like to briefly recapitulate that the conditioning on the future used when constructing the different subsamples implies that we are estimating a new set of treatment parameters that are actually biased in different respects. The entrant-stayers/never-entrants comparison excludes export failures and future entrants, which will result in an upward bias in the estimated treatment effect. Similarly, the entrant-stoppers/nonentrants comparison disregards export successes but includes future entrants, which will induce a downward bias in the estimated treatment effect. In both cases, the bias is a result of conditioning on future export status and therefore implicitly on future outcomes.

With these reservations in mind, Fig. 4 illustrates the results based on cross-sectional matching for the two combinations in question.²⁹ Note that the specifications of the propensity scores are the same as in Sect. 4.2 (i.e., the learning-by-exporting specification), and accordingly the results should be compared with those in Fig. 1 and Table 7.

Fig. 4

Cross-sectional matching estimates of the effect of export entry on labor productivity for different export status combinations (learning-by-exporting specification). Notes: Based on the cross-sectional estimates in Table 11. Filled data marker indicates effect significant at the 10% level or lower

In Fig. 4, we observe that the effect of export entry for entrant-stayers relative to never-entrants in the micro firm category is considerably larger than the effect for entrant-stoppers relative to nonentrants in the corresponding class. For the former, the effect on labor productivity of export entry is between 22% and 26% and has a slight tendency to increase over time. For the latter, the productivity effect is between 6% and 11% but is not consistently statistically significant. The corresponding results from Table 7, in which we compare export-entrants with nonentrants, are between 11% and 14%. For the larger firms, there seem to be no differences in the estimated effects depending on the applied definitions of export status.

In sum, we conclude that, when we refine the export-entrants into entrant-stayers and the nonentrants into never-entrants, the positive productivity effect of export entry among micro firms becomes larger. Furthermore, we may discern a small increase in the productivity gap between export-entrants and nonentrants subsequent to entry.

5 Conclusions

The exporter productivity premium in Swedish manufacturing is larger in smaller firms, and while the export participation rate in general is high, it is still fairly low among the smaller firms. This means that policymakers might be particularly interested in whether, above all, smaller firms that enter the export market tend to improve their productivity performance relative to nonentering firms, i.e., whether they learn by exporting.

Using propensity score matching techniques, we find that there is an instantaneous productivity increase at the time of entry, especially for smaller firms, but that the productivity gap between entrants and nonentrants appears to be constant in the periods subsequent to entry. If the firms had learnt by exporting, we would have expected to see a widening productivity gap. However, when we look exclusively at smaller successful exporters—i.e., smaller firms that enter the export market and, after entrance, continue to be exporters—and compare their productivity trajectory after entry with that of firms that never enter the export market, we may see a tendency toward an increase in the productivity gap.

Pre-entry labor productivity is significantly higher for smaller future exporters than for firms that do not enter the export market, which indicates that those firms self-select into export. Furthermore, if in our matching analysis we allow for different productivity trajectories before export entry, we observe that there is a significant productivity differential, at least for smaller firms, between export-entrants and matched nonentrants even before export entry. We also find that increased investments in physical capital precede entry into export markets, and this may explain the higher productivity prior to export entry among future export-entrants. We interpret these results as an indication that learning to export may indeed exist. Although we can provide evidence that both self-selection and learning to export may prevail, we are not able to say anything about their relative importance.

Appendix

See Tables 10 and 11.

Table 10 Cross-sectional matching estimates of the effect of export entry on physical capital per employee (learning-to-export specification)

Table 11 Cross-sectional matching estimates of the effect of export entry on labor productivity for different export status combinations (learning-by-exporting specification)