1 Innovation for Growth – i4g Policy Brief N°21 Innovative firms in Global Value Chains´ Reinhilde Veugelers Findings Global Value Chains (GVC) challenge prevailing policy thinking about competitiveness. GVCs drive countries interdependence: - Country's exports increasingly embody technology, labour and capital of other countries by imports of intermediate goods; - Imports increasingly reflect tasks which complement, rather than substitute for, domestic production. GVC-involved-firms simultaneously organize part of their production abroad, import components for local production and export their produced goods. The EFIGE database analysis shows that relatively few GVC firms exist (5%). However, these firms are the bigger ones, substantially driving value added creation and employment. These firms matter critically for the innovation performance of countries and sectors. They are more heavily engaged in R&D activities, have a sophisticated human capital base, hire more workers with a university degree and are consequently being able to support high unit labour costs. These firms introduce more product innovations and display the highest productivity premia. Recommendation The growing interconnectedness limits the effectiveness of national policies and requires more international (or EU-level) coordination of policies. In this context, government intervention should be sufficiently generic, and avoid picking particular sectors to support. However, it should - Provide the framework conditions to support a GVC innovation-based growth path for enterprises, whatever their size, age or sector affiliation; - Aim at capabilities that support of higher value-added GVC activities; - Support productive internationalised firms with unique innovative capabilities, having a GVC strategy which is more than simply exporting.
16
Embed
Innovative firms in Global Value Chains´ · mostly missing is a robust analysis of the trends and impact of global value chains across sectors and countries at the level of individual
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
Innovation for Growth – i4g
Policy Brief N°21
Innovative firms in Global Value Chains´
Reinhilde Veugelers
Findings
Global Value Chains (GVC) challenge prevailing policy thinking about competitiveness.
GVCs drive countries interdependence:
- Country's exports increasingly embody technology, labour and capital of other countries by
imports of intermediate goods;
- Imports increasingly reflect tasks which complement, rather than substitute for, domestic
production.
GVC-involved-firms simultaneously organize part of their production abroad, import components for local
production and export their produced goods.
The EFIGE database analysis shows that relatively few GVC firms exist (5%). However, these firms are the bigger
ones, substantially driving value added creation and employment.
These firms matter critically for the innovation performance of countries and sectors. They are more heavily
engaged in R&D activities, have a sophisticated human capital base, hire more workers with a university degree
and are consequently being able to support high unit labour costs. These firms introduce more product
innovations and display the highest productivity premia.
Recommendation
The growing interconnectedness limits the effectiveness of national policies and requires more international (or
EU-level) coordination of policies.
In this context, government intervention should be sufficiently generic, and avoid picking particular sectors to
support. However, it should
- Provide the framework conditions to support a GVC innovation-based growth path for enterprises,
whatever their size, age or sector affiliation;
- Aim at capabilities that support of higher value-added GVC activities;
- Support productive internationalised firms with unique innovative capabilities, having a GVC
strategy which is more than simply exporting.
2
REINHILDE VEUGELERS
PROFESSOR AT KULEUVEN, SENIOR FELLOW AT BRUEGEL, MEMBER OF THE I4G EXPERT GROUP
OCTOBER 2013
1. Introduction and motivation
By now a substantial piece of research and evidence has been developed on trends in global value
chains at the aggregate level of economies and sectors (see eg OECD (2013)). What is however still
mostly missing is a robust analysis of the trends and impact of global value chains across sectors and
countries at the level of individual firms. In this contribution, we will exploit the EFIGE database to
look at innovation and global value chains at the firm level1. With its detailed information on
internationalization strategies of firms together with information on their innovative profile and
performance information, EFIGE provides a unique opportunity to analyse GVC involvement, its
impact and relationship with innovation at the firm level.
We first describe how we measure a firm’s involvement in global value chains in the EFIGE dataset.
Section 2 presents the firms which we identify as involved in global value chains. We will show
how concentrated overall value added and employment are in few GVC involved firms. It thus
matters to get to know these few but pivotal firms better. GVC involved firms are also likely to
matter for competitiveness because of their productivity and innovation profile. Section 3 looks at
the performance profile of multiple mode firms, analyzing whether higher GVC involvement is
associated with higher productivity. Section 4 looks at the R&D, human capital investment and
innovative performance of GVC involved firms. Section 5 looks at how vulnerable GVC firms are to
the crisis and how innovative performance helps to shield GVC firms from the crisis. Section 6
summarizes the main findings. Section 7 provides some policy implications.
1 This contribution draws heavily on Veugelers et al (2013). For more detailed information on the EFIGE database, see the various Bruegel Blueprints (The Happy Few and Triggers of Competitiveness) and the EFIGE website (http://www.bruegel.org/datasets/efigedataset/). EFIGE was an FP7 funded project.
Metal, Machinery and Equipment 29.3% 32.2% 34.1% 4.4%
Electrical and Optical equipment 18.7% 29.7% 42.0% 9.6%
Transport Equipment 22.8% 25.0% 43.4% 8.7%
Source: Bruegel calculations on the basis of EFIGE; Veugelers (2013)
The sectoral profile reported in Table 2 remains robust in multivariate analysis, identifying which
firm characteristics are significantly associated with multiple mode use, controlling for the presence
of other determining characteristics. This multivariate (econometric) analysis is reported in
Veugelers et al (2013). With respect to country affiliation, there are no significant differences
between UK, German and French firms being triple mode internationalisers. Spanish firms are less
5
likely to have dual and triple moders. Austrian firms are significantly more likely to be triple
moders, c.p.
The multivariate results also show that bigger firms are significantly more likely to deploy more
complex GVC strategies, being dual and triple mode internationalisers. The following table
illustrates that multiple mode firms are indeed the bigger ones. They are even more sizable in terms
of value added compared to employment, reflecting a substantial labour productivity premium for
multiple moders, especially triple moders. This translates into triple moders being able to sustain
higher unit labour costs.
TABLE 3: EMPLOYMENT AND VALUE ADDED AND MULTIPLE MODE INTERNATIONALIZATION
Zero Single Dual Triple
Employment 33 44 97 276
Value Added
(thousands of euros)
1246 2377 7930 20005
Labour Productivity 734 1042 2467 5471
Unit Labour Cost 0.79 0.73 0.73 0.93
Source: Bruegel calculations on the basis of EFIGE; Veugelers et al (2013)
Multiple moders, being the bigger firms, matter substantially for aggregate performance of sectors
and countries, as Figure 1 illustrates. The 5% of firms which are triple moders account for about 27%
of total sample VA and 24% of total sample employment. The group of dual moders (i.e. mostly firms
who simultaneously substantially import and export) representing 34% of total sample firms,
account for 53% of total sample VA and 45% of total sample employment.
6
FIGURE 1: SHARE IN NUMBER OF FIRMS, TOTAL TRADE, VA AND EMPLOYMENT OF MULTIPLE MODERS
Source: Bruegel calculations on the basis of EFIGE; Veugelers et al (2013)
3. GVC involvement and Productivity
Are firms who are involved in GVCs the higher productivity firms? We know from the empirical trade
literature that the more productive firms are more likely to sort into exporting and offshoring, as
they can overcome the costs associated with internationalization (Melitz (2003), Helpman,
Melitz&Yeaple (2004)). One can expect this to a fortiori hold for multiple mode firms being
simultaneously involved in exports, imports and location of activities abroad. Only the most
performing firms will be able to overcome the costs and risks associated with multiple mode
internationalisation. At the same time GVC involvement can be expected to increase firm
performance, allowing, if only, for more efficient sourcing and larger market access. It does however
also make GVC firms more vulnerable for external shocks.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Distribution ofFirms
Total Trade Value Added Employment
Triple Mode
Dual Mode
Single Mode
Zero Mode
7
Our measure for productivity in this section will be Total Factor Productivity. Total Factor
Productivity (TFP) refers to the estimated productivity of all inputs taken together and it is a
measure of the global efficiency of a firm3. As a 'residual', TFP basically accounts for effects in total
output growth not caused by capital and labour. TFP is commonly interpreted as a measure of the
technology of production and its rate of growth as a measure of technical progress.
Unfortunately, since in the EFIGE dataset TFP is measured simultaneously with the identification of
the internationalization strategy, we can only investigate a possible correlation between productivity
and GVC involvement. We cannot identify any causal relationship, i.e. whether multiple mode
internationalization strategies are making firms more productive or whether more productive firms
select into multiple mode internationalization.
As the distribution of firms along TFP levels is highly skewed, it makes more sense to look at the full
distribution of productivity levels, rather than averages (see also Altomonte et al (2012)).
FIGURE 2. TFP DENSITY AND MULTIPLE MODE INTERNATIONALISERS
Source: Bruegel calculations on the basis of EFIGE; Kernel-density functions; Veugelers et al (2013)
3 For a more elaborate discussion of our productivity measure, see Veugelers et al (2013).
8
Figure 2 shows how firms with higher GVC involvements are on average characterized by higher
levels of TFP, particularly the triple mode firms who combine importing intermediates, exporting and
international location of production. The superior performance of these triple moders holds
especially in the upper tail of the distribution, i.e. triple moders are not just more likely to have on
average a higher TFP, they are especially more likely to be the (very) high performers.
Multivariate analysis reported in Veugelers et al (2013), which corrects for other firm and industry
characteristics, confirms a significant and sizeable productivity premium for dual mode and
particularly for triple mode internationalisers. This confirms that GVC firms are more productive not
only because they are typically bigger and older or that they belong to high productivity sectors, but
that their productivity premium is intrinsically connected to their more complex international GVC
involvement, combining several internationalization modes.
4. GVC involvement and Innovation
Is the superior productivity of GVC involved firms, as shown by their TFP premium associated with
being more innovative as driver for their superior productivity? To investigate this in more detail
we look at direct evidence in the EFIGE dataset on the innovative profile of the sample firms. On the
input side, we know the firms’ expenditures on R&D, their employment in R&D activities and the
quality of their human capital base, as measured by the share of university graduates in their
workforce. On the innovation output side, we know whether they introduced any new product
and/or process innovations in the time period considered. This information allows to test more
directly whether GVC involved are more innovative, driving their higher productivity performance.
TABLE 4: INNOVATION INPUTS AND MULTIPLE MODE INTERNATIONALIZATION
Zero Single Dual Triple
% Firms involved in R&D 42% 59% 72% 86%
Share of employees involved in R&D 7.0% 8.1% 8.9% 10.4%
Share of University Graduates in Employment 6.7% 8.6% 11.0% 14.6%
Source: Bruegel calculations on the basis of EFIGE
9
Table 4 shows that dual and triple moders are more likely to be involved in R&D activities, having a
higher share of their employers engaged in R&D activities. Furthermore, their human capital base
is more developed, as they hire a larger share of employees with university degrees.
Because of their higher R&D profile and being the bigger firms, triple mode firms account for a
substantial part of overall sample R&D activities. With only representing 5% of the total number of
firms, triple mode firms account for 38% of total sample R&D expenditures, 29% of R&D employees
and 33% of graduates employment. They therefore matter substantially for the overall R&D
performance of sectors and countries.
FIGURE 3: SHARE IN NUMBER OF FIRMS, R&D AND GRAUDATE EMPLOYMENT OF MULTIPLE MODERS
Source: Bruegel calculations on the basis of EFIGE;
Based on their higher deployment of R&D resources and more developed human capital base, triple
mode firms are more likely to introduce product innovations (either sole but particularly in
combination with process innovations) (Table 4). The higher likelihood to introduce process
innovations solely, often associated with introducing cost cutting, labour saving technologies, is
much less pronounced for multiple moders, particularly for the triple moders. All this is consistent
with the story that higher levels of GVC involvement, as evidenced through using multiple
internationalization modes, are associated with higher innovative performance. It does not support
the often claimed adagio that GVC firms are the cost cutting process innovators. This is also
confirmed in multivariate analysis, correcting for other determining factors of innovative
performance (such as R&D and Human Capital inputs).
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Distribution offirms
R&Dexpenditure
R&D employees Graduateemployees
triple
dual
single
zero
10
TABLE 4: INNOVATION AND MULTIPLE MODE INTERNATIONALISERS SHARE OF FIRMS HAVING INTRODUCED PRODUCT, PROCESS, BOTH OR NO INNOVATIONS, ROW%
No Innovation Product Process Both
Zero 50% 15% 18% 16%
Single 36% 21% 16% 26%
Dual 24% 24% 14% 37%
Triple 17% 31% 9% 43%
Source: Bruegel calculations on the basis of EFIGE;
BOX 1: Econometric analysis of complexity of international involvement and innovation We look at which firm characteristics significantly correlate with innovative performance in a multivariate fashion. Using a multinomial logit specification, we look at which firm characteristics determine the likelihood for firms to be introducing product innovation, process innovations or the combination of product and process innovations. The multivariate analysis of innovative performance controls for several inputs into the innovation process. It finds that R&D investments increase the probability of innovative performance, for product as well as process innovations. While ICT investments are significant drivers, enabling process innovations, human capital is most associated with product innovations. Our main focus is on the use of a single, dual or triple internationalization mode, the base comparison being firms that have no substantial international activities. Table B.1 displays the results.
The multivariate analysis controlling for other firm characteristics, confirms that firms which are
more GVC involved are significantly more likely to introduce new products, new processes solely or
11
in combination. This suggests that innovation can indeed be considered as a likely source for the
higher productivity of GVC involved firms. This holds particularly for dual and triple mode
internationalisers and for product innovations. Firms that combine importing, exporting and foreign
production are particularly more likely to be introducing product innovations, but not sole process
innovations.
5. GVC involvement, Innovation and crisis sensitivity
With the positive association between higher GVC involvement and productivity and innovation,
documented supra, one could expect that their higher productivity profile shields firms with higher
GVC involvement from the crisis, making them less sensitive to the crisis and consequently less likely
to shed jobs. On the other hand, there is also evidence that GVCs act as a channel for the rapid
transmission of real and financial shocks, thus amplifying the national fluctuations of demand for
final goods, making GVC involved firms more crisis-sensitive (cf Altomonte et al., 2012;. OECD 2013)).
With the EFIGE survey it is possible to investigate whether or not firms witnessed a decrease in
turnover between 2008 and 2009 and by how much. Tables 5 show these responses for the different
types of GVC involvement. The descriptive analysis shows that GVC involved firms are somewhat
more likely to have reductions in turnover, but the differences are small.
TABLE 5: MULTIPLE MODERS AND TURNOVER REDUCTION DURING THE CRISIS
Reduction of Turnover in 2009
No
reduction
<10% 10%-
30%
>30%
Zero 31% 20% 31% 18%
Single 28% 19% 35% 18%
Dual 27% 18% 36% 19%
Triple 28% 19% 39% 14%
Source: Bruegel calculations on the basis of EFIGE
The multivariate analysis confirms these statistics. GVC involvement does not significantly correlate
with crisis sensitivity. At the same time and all else equal, the results interestingly show that firms
which invest in innovative strategies are less likely to have seen cuts in turnover or have smaller cuts
12
in turnover. This evidence corroborates the assertion that innovation strategies are important for
firm resilience to external shocks.
Box 2: Econometric analysis of the international involvement and sensitivity to the crisis We analyse the firm characteristics conducive to a higher crisis sensitivity in a multivariate fashion using on ordered probit with the categories being: no change in turnover, <10%, 10-30%, > 30% loss in turnover. The use of multiple internationalization modes are our key firm characteristics of interest. Simultaneously we correct for firm size, age, foreign ownership, competition from abroad, intermediate status, as well as sector and country affiliation. We also correct for the presence of inputs for innovation (R&D, Human Capital, ICT). Table B.2 presents the results. Correcting for other firm characteristics, there is no significant difference in crisis sensitivity for single, dual or triple mode internationalisers. A significant driver of crisis resilience is the firms’ strategy in terms of innovation (R&D as a proxy) and human capital: those firms with a higher share of university graduates, a higher share of investment in R&D and ICT are less likely to experience a reduction in turnover or only smaller reductions in turnover.
Table B.2: Ordered Probit on Turnover change during the crisis
dummies, foreign group, foreign competition included as well
Obs 11152
Pseudo R-square 0.059
Chi-square 1741.147
* p<0.10, ** p<0.05, *** p<0.01
Note: Small firms: 20-49 employees; Medium-sized firms: 50-249 employees; Large
firms: above 250 employees.
13
6. Main findings
In this contribution we exploit the EFIGE database to study the prevalence and impact of GVCs at the
firm level, thus further complementing the macro-evidence on GVC involvement, presented in the
literature (OECD (2013). We characterize firms involved in GVCs through the complexity of their
internationalization strategy: GVC involved firms are characterized as “multiple moders” with the
highest involvement firms those that simultaneously organize (part of their) production abroad,
import components for local production and exports their produced goods, which we label as “triple
moders”.
The firm level analysis not only shows the heterogeneity within sectors and countries on GVC
involvement. It also shows the skewedness of performance of sectors and economies towards GVC
involved firms. There are relatively few multiple mode firms, combining different international
activities, particularly triple mode firms that combine importing of components and exporting of
produced goods with organizing part of their production activities abroad. But these few firms are
the bigger firms, substantially driving total value added creation and employment in sectors and
economies. It therefore matters to get to know them better.
Multiple moders, particularly the triple moders, also matter critically for the innovation performance
on countries and sectors. They are more heavily engaged in R&D activities, have a more
sophisticated human capital base, hire relatively more workers with a university degree and are
consequently being able to support higher unit labour costs. They are significantly more likely to
introduce new product innovations, displaying the highest productivity premia. And thanks to their
superior innovative profile, multiple mode firms are not more likely to be crisis sensitive.
Overall, the analysis of global value chain involvement and its impact on performance at the firm
level provides consistent evidence supporting that firms taking on the opportunities of global market
access and global resources sourcing, by combining multiple modes of internationalization,
simultaneously importing components, organizing production abroad and selling their products to
international markets, although perhaps small in number, but with their superior productivity and
innovation capacity profile, are well placed to be engines of Europe’s innovation-based growth,
driving its external competitiveness with globally sustainable comparative advantages. Given their
highly specific position for overall performance, it matters for policy makers to better understand
who they are, what they do and what challenges they face.
14
In order to understand them better, more firm-level analysis is needed, particularly tracing GVC
firms over time, to better identify the causal relationship between internationalization strategy and
innovative performance: do firms need to be strong innovators before they can benefit from the
opportunities offered from engaging in global value chains or does engagement in global value
chains turn firms into strong innovative players, being able to wither the gales of fierce global
competition?
7. Some Policy Implications
Global value chains challenge prevailing policy thinking about competitiveness. The growing
upstream and downstream interconnections within GVCs make countries more interdependent: one
country's exports increasingly embody the technology, labour and capital of other countries from
which intermediate goods are imported; imports increasingly reflect tasks which complement, rather
than substitute for, domestic production; the offshoring of a production stage which can be
performed more efficiently abroad makes domestic activities more competitive. This growing
interconnectedness limits the effectiveness of national policies, requiring more international
coordination of policies. For European countries this means first and foremost deeper coordination
at EU level.
For countries to benefit from GVCs, GVCs must become effectively linked to domestic capabilities in
certain tasks. Here a host of national and European economic policies largely determine which
position countries occupy in GVCs: which activities, jobs and what value they are able to create. The
activities to focus on are the higher value-added activities that build on unique and innovative
capabilities. The presence of these activities in Europe will secure productivity growth and external
competitiveness, and retaining them will require having in place the framework conditions that are
most pivotal for these innovative, higher value-added, growth contributing, activities.
Although sectoral idiosyncrasies exist, the challenges and opportunities identified here apply to most
if not all sectors. The shift towards innovative activities is not confined to high-tech sectors and
manufacturing only. Any type of government intervention should be sufficiently generic, and avoid
picking particular sectors to support. What matters most is providing the framework conditions for
viable activities to continue to prosper, and for new activities to develop and grow into leading
world market status, irrespective of in which sector they are classified. What matters more is to
provide the framework conditions to support a GVC innovation-based growth path for enterprises,
15
whatever their size, age or sector affiliation. The target should be productive firms with unique
innovative capabilities, which are able to develop an internationalisation strategy that involves more
complex GVC strategies than simply exporting.
The offshoring of routine jobs and the structural shift from classic routine production jobs towards
higher value-added types of jobs that have some of the characteristics of services jobs, has major
implications for employment policy. Although innovative capacity helps firms to shield against
negative shocks, adjustment difficulties are likely to result, because the skill requirements for the
newly created jobs tend to be higher than and different to those for the jobs lost. Effective domestic
policies are therefore needed to reduce the adjustment costs borne by displaced workers. Policies
to pursue, primarily at EU member-state level, include improving the functioning of labour markets
and strengthening education and training.
16
References
Altomonte, C., Aquilante, T. and Ottaviano, G. (2012) ‘The triggers of competitiveness: the EFIGE cross country report’, Blueprint no 17, Bruegel, Brussels.
Altomonte, C., Di Mauro, F., Ottaviano, G., Rungi, A. and Vicard, V. (2012), ‘Global Value Chains During the Great Trade Collapse. A Bullwhip Effect?’, European Central Bank Working Paper no 1412.
Mayer T., Ottaviano, G. (2007), ‘The happy few: the internationalisation of European firms,’ Blueprint 3, Bruegel.
Melitz M. (2003), ‘The Impact of Trade on Intra-industry Reallocations and Aggregate Industry Productivity’, Econometrica, Vol. 71, No. 6, 2003, pp. 1695-1725.
Helpman E., Melitz M. and Yeaple S.R. (2004), ‘Export versus FDI with Heterogeneous Firms’, American Economic Review vol 94(1), 300-316.
OECD (2013), ‘Interconnected Economies: Benefiting from Global Value Chains’, OECD Publishing. doi: 10.1787/9789264189560-en
Veugelers, R., F. Barbiero, M. Blanga-Gubbay, (2013), Meeting the manufacturing firms involved in GVCs, in Veugelers, R. (2013) (Ed.) Manufacturing Europe’s Future, Blueprint no 21, Bruegel, Brussels.