Exporting creative and cultural products: Birthplace diversity matters!

3.1 Endogeneity

A potential issue that needs to be addressed is endogeneity. Omitted variables concerns are reduced in our case since we include country-sector fixed effects which capture all unobserved destination-specific factors affecting the settlement of immigrants in the host countries (i.e., average income level, immigration policy and any country-specific economic factor attracting immigrants). However, an unobserved country-year-specific shock might contemporaneously affect the exports of creative goods and the settlement of immigrants across destinations. Moreover, reverse causality could produce biased ordinary least squares (OLS) estimations if changes in the export of creative goods affect the demand for labour in the country and, in particular, demand for high-skilled (immigrant) workers. We address these endogeneity concerns by adopting an instrumental variable approach (two-stage least squares—2SLS).

Much of the migration literature adopts the shift-share instrument à la Card (2001) to solve the problem of endogeneity in immigrant settlement.10 Unfortunately, in our framework, the total bilateral flows of immigrants used to build the shift-share instrument would be unlikely to be exogenous (since they can be assumed to be affected by destination-year-specific export performance and labour demand). So, we introduce an alternative instrumental variable, in line with Card (2001), but based on the predicted supply-driven number of migrants in each destination (exporting) country i. Basically, we run a structural gravity model that predicts the bilateral stocks of migrants (taken from IAB)11 using: (i) destination-year fixed effects; (ii) origin-year fixed effects; (iii) a set of country-pair-specific geographic variables (distance, common border, language and colonial past); (iv) the share of immigrants coming from a specific origin in 1980; and (v) its interaction with the total stock of migrants from a given country – ln(Immi)_jt.12 We use a PPML estimator to account for the many zeros in bilateral migration stocks and to estimate the following equation:

(3)

where the destination-year fixed effects (γ_it) capture the pull factors of bilateral migration, the origin-year fixed effects (γ_jt) account for the origin-year-specific push-factors affecting the outflow of migrants, and the set of gravity type covariates control for the country-pair-specific migration cost. The inclusion of the share of immigrants in i coming from j in 1980 (ShareImmi_ij,1980) and its interaction with ln(Immi)_jt mimics the idea in Card (2001) that new immigrants tend to settle in destinations where previous immigrants from the same origin already reside. The main advantage of this specification relies on the fact that Geography_ij will contribute to predicting the bilateral stock of migrants also for those origin-destination pairs ij with zero migrants in the base year of ShareImmi_ij,1980. From Equation 3, we take the predicted value (fit of the regression) from which we exclude the destination-year fixed effect component:

(4)

Thus, our instrumental variable does not include the demand-driven component of bilateral migration, which is the origin of the endogeneity concern in our framework (and the basis for the main critiques of the shift-share instrument). Finally, we build the birthplace concentration index as in (Equation 1) using :

(5)

Notice that:

is simply the share in the total population of the supply-driven predicted number of migrants in country i originating from country j. Thus, the BD index is built purely on the supply-driven migration component and can be used safely as an instrumental variable, IV_it. The exclusion restriction assumption here is that this BD index based on supply-driven migration stocks (i.e., IV_it) affects the export of creative goods only through the BD index based on total migration stocks. This is tested in Appendix Table A2 which shows that IV_it has no direct effect on the export of creative goods.

As a robustness check, we use the 5-year lag of the BD index as an instrumental variable, which allows us to test an overidentified model and to apply a Sargan test for validity of the instrumental variables. The idea is that the composition of the migrant population is persistent over time, so the lagged value of the diversity index is a good proxy for the current diversity index value (relevant IV). Moreover, any contemporaneous shock to creative products exports cannot be related to previous values of the diversity index (IV validity).

3.2 Data and Descriptive evidence

Our calculation of the BD index relies on IAB bilateral migration stocks data. IAB data include information on migrants’ country of origin (nationality) and education level for 20 OECD destination countries, during the period 1980–2010 (5-year intervals). Data on the education level of immigrants (primary, secondary, tertiary) allow us to test the effect of birthplace diversity by skill level.13 To calculate the concentration index, we consider migrants as foreign-born individuals aged at least 25 years. A potential drawback of using IAB data is that they report only the nationalities. Naturalised foreign-born individuals and second-generation immigrants are not classified as “immigrant” in our data. So, our measure of birthplace concentration underestimates the real degree of cultural diversity.

As a robustness check, we use bilateral migration flows rather than stocks to compute the concentration index. In this case, we rely on the data in Abel and Sander (2014), which provide information on bilateral migration flows for a 178×178 matrix of origin–destination combinations, for the years 1995, 2000, 2005 and 2010.14 Bilateral migration stocks (and flows) are used to compute a country-year-specific birthplace diversity measure (see Equation 1) that can be merged with data on creative and cultural goods exports.

To calculate the volume of creative and cultural goods exports, we combine trade data from BACI (CEPII) with three recent HS-based classifications of creative and cultural goods. The first classification was released by UNESCO and includes only the core creative industries (used here as baseline). The second was also released by UNESCO and includes both core and non-core creative industries (robustness check). The third classification used in this paper was released by UNCTAD (robustness check). Trade data from BACI provide information on the value of export flows (in US$) for a complete set of exporting and importing countries in the period 1989–2015, by product HS 6-digit level. Then, the UNESCO (and UNCTAD) classification is used to identify whether a specific HS 6-digit code belongs to the category “creative and cultural exports.” This allows us to compute the total amounts of creative and cultural goods exports by country-sector(HS2)-year, which we use as the main dependent variable in our empirical exercise. Notice that, in our regression sample, we include only the HS 2-digit sectors containing at least one HS 6-digit product classified as “creative or cultural.”

Merging trade data on cultural products (UNESCO core classification) with the BD index based on the stock of migrants (baseline estimations) produces a sample of 1,425 observations: 19 exporting countries,15 15 HS 2-digit sectors, and 5 years (1990, 1995, 2000, 2005 and 2010). See Table 1 panel A for in-sample descriptive statistics. When we use the UNESCO (core plus non-core) classification, the number of observations increases slightly (1,615 observations) because two additional HS 2-digit sectors include non-core creative products (see Table 1 panel B). When we use the UNCTAD classification for creative goods, the number of observations increases because it includes 29 HS 2-digit chapters with creative products. See Table 1 panel C for in-sample descriptive statistics using the UNCTAD classification to define creative products. Importantly, our results are robust to the three classifications used to define creative exports. In contrast, if we merge trade data with the BD index based on flows of migrants (robustness check), the number of observations increases since the sample of exporting countries is larger (178 exporting countries).

Figure 3 depicts the simple correlations between countries’ exports of creative products (country averages across sectors), conditioned on the size of the country and the diversity of immigrants’ birthplaces (HH index). For the years 1995, 2000, 2005 and 2010, the relation is always negative, suggesting a positive correlation between the export of creative products and diversity in immigrants’ birthplaces.

4.1 Baseline results

Ordinary least squares results for Equation 2 are presented in Table 2. Across all specifications, the concentration of immigrant birthplaces has a negative and significant effect on the export of creative and cultural goods. This means that immigrant diversity has a positive impact on the creative goods exports. In particular, using our preferred specification (column (3)), a 10% decrease in the birthplace concentration index implies a 4% boost in the export of creative and cultural goods. Notice that our results hold for both the UNESCO and the UNCTAD classifications of creative and cultural goods (see columns (3)–(5)). As expected, the share of natives in the total population has no effect on the export of creative products, confirming the idea that it is the diversity in immigrants’ origins that matters for the export of creative goods, not the size of the foreign-born population. Our baseline results are robust to the inclusion of per capita GDP among the covariates. See Appendix Table A1.

Table 3 presents the results of the 2SLS using the instrumental variables described in Section 3.1. Column (1) presents the estimation results using the IV based on the predicted supply-driven migration stocks and, column (2) presents the results using a 5-year lag of the BD index as an instrumental variable. In columns ((3)–(5)), we employ an overidentified model and put both IVs in the first stage. The relevance of the two IVs is suggested by the first-stage coefficients at the bottom of the table. In all the specifications, the instruments are good proxies for the (potential) endogenous BD index. The first-stage F-stat are all considerably above the rule of thumb value of 10 and remove any potential problem of weak instruments. The Sargan tests presented in columns ((3)–(5)) prove the validity of our IVs (orthogonality). Finally, to support the exclusion restriction of our IVs (as suggested by Conley, Hansen, & Rossi, 2012), we estimate Equation 2 by adding the two IVs (separately, and then together) as additional covariates to test whether they have a direct effect on the export of creative goods. The exclusion restriction assumption is satisfied if the IVs affect the export of creative goods only through their influence on the endogenous regressor (and, thus, when they have no direct effect on the export of creative goods). As expected, none of the IVs has a significant direct effect on the export of creative goods, supporting the validity of the exclusion restriction (see Table A2). The second-stage results of the 2SLS approach are reported at the top of Table 3 and confirm what discussed above. Diversity in the birthplace of immigrants has a positive and significant effect on the export of creative goods (no matter which classification is used to identify creative products). In particular, a 10% decrease in the BD index (i.e., an increase in birthplace diversity) implies a 4.4% increase in the export of creative products—see column (3) in Table 3.

Table 4 presents our first robustness check using a PPML estimator to address the zero trade flow problem. Our results perfectly hold when we use the UNESCO classification (both core and core plus non-core classifications)—see columns (1) and (2). The sign of the coefficient is in line with expectations for the UNCTAD classification also, but is estimated with errors (coefficient not significantly different from zero).

Table 5 presents another robustness check using the alternative definition of birthplace diversity proposed by Alesina et al. (2016) —one minus the HH index applied to the population of immigrants. The results are qualitatively similar to those obtained using our BD index, but have the opposite signs (the measure proposed by Alesina et al. (2016) is a direct measure of diversity). Appendix Table A3 presents the 2SLS results using the birthplace diversity index, as in Alesina et al. (2016). As in our baseline estimations, we use two instrumental variables: (i) the diversity index (1-HH) based on the supply-driven predicted migration stocks, and (ii) a 5-year lag of the diversity index (1-HH). In all the specifications (for the three classifications of creative goods), an increase in birthplace diversity index boosts the export of creative goods.16

4.2 Robustness check using migration flows

The main drawback to IAB data on bilateral migration stocks is the small sample of destination countries (19 OECD countries), which implies the validity of the diversity-creative export nexus for the subsample of rich countries only. In this section, we account for this limitation by increasing the set of destination (exporting) countries using Abel and Sander (2014) data on bilateral migration flows to/from 178 countries. The construction of the BD index does not change; we simply apply Equation 1 to an alternative migration data set including a bigger sample of destination countries.

The results of this robustness check are reported in Table 6 and show that, even if we consider a bigger set of exporting countries (both developed and developing), the positive effect of birthplace diversity on the export of creative goods holds, but the coefficient is smaller, indicating that the inclusion of poor countries weakens the nexus. This suggests that the effect of birthplace diversity for boosting exports of creative goods is particularly important for developed countries.

4.3 Results by skill level

As discussed in the introduction, the diversity and export of creative goods nexus is supposed to be particularly strong for the most talented workers (see Florida, 2002). The biggest contribution to the “creative” process in the creative and cultural goods industry is supposed to be generated by tertiary educated individuals. So, we expect diversity among high-skilled immigrants to have the strongest effect on the export of creative goods. We exploit information on immigrants’ education level from IAB and compute the BD index (as in Equation 1) by education level. The baseline specification results by education level are reported in Table 7. We find negative and significant coefficient for the concentration of tertiary and secondary educated migrants, confirming the intuition that what matters for the export of creative goods is the diversity in the group of the most talented workers. Moreover, in line with the intuition, the coefficient of tertiary educated is bigger than the coefficient of secondary educated immigrants.

4.4 Results by macrosector

Finally, Table 8 reports the baseline estimation results by macrosector. Diversity in immigrants’ birthplaces may be of particular interest for some sectors and less relevant for others. While the estimated specification does not change (see Equation 2), we needed to adapt the set of fixed effects included in the regression to achieve sufficient variability in the data to identify the coefficient of interest. Therefore, we include only country and year fixed effects in the sector-specific regressions. Interestingly, the strongest contribution of birthplace diversity in boosting the export of creative goods is in: (i) plastic and rubbers, (ii) wood and wood products; and (iii) miscellaneous. Notice that the miscellaneous macrosector includes HS chapters 90–97, that is, optical, photographic and cinematographic products, clocks and watches, musical instruments, toys and games, works of art.