The role of import for KIBS intensity: A comparative analysis of European Union countries

The aim of this paper is to study the role of import for the growth of KIBS intensity in the economy. The study covers the EU countries, making it possible to carry out a comparative analysis between countries, in particular between the EU-15 and the EU-12. The study is based on the data from the World Input-Output Database (WIOD). Owing to the availability of relevant data, the period analysed covers the years 1995–2011. The study demonstrates that KIBS intensity is higher in the more developed EU countries, and KIBS are mainly delivered from domestic markets. Exceptions include Ireland and Luxembourg, distinguished by the highest and rapidly growing KIBS intensity (satisfied mainly by import), which is accompanied by their strongest KIBS export performance. In the analysed period, the importance of imported KIBS increased (mainly in the EU-15), but to a much lesser degree than with respect to manufacturing products. The role of imported KIBS was higher in the EU-12 than the EU-15 (with a trend to decrease these disparities), as well as in smaller economies. #0#


Introduction
The development of knowledge-based economy, the information and communications technology (ICT) revolution, and increased competition on markets have resulted in the need to reorganize production processes in enterprises so that they can effectively exploit the intellectual potential of more educated workers, supported by ICTs and automation (Bhagwati, 1984;Jones, Kierzkowski, 1990;Motohashi, 1997;Fixler, Siegel, 1999;

Global sourcing of KIBS
The dynamic development of global sourcing is linked to the ICT revolution. The rapidly declining costs of communication and coordination have made it possible to perform most manufacturing stages in different locations (Baldwin, 2006(Baldwin, , 2014. Most recently, the ICT revolution has increased significantly the tradability of many services, and then the process has spread from factories to offices (Baldwin, Lopez-Gonzalez, 2015). This, in turn, has stimulated the development of international trade in intermediates with respect to both manufacturing goods and services.
Over the recent years the range of business services that have been subject to global sourcing has not only grown, but has also changed its structure. The growing demand for specialized business services has reduced the importance of simple, routine and standardized tasks, while increasing the role of processes with higher value added (Lewin, Massini, Peeters, 2009). It should be noted, however, that the possibility of providing services from a distance is primarily applicable to highly standardized services (Rybiński, 2008). Tailor-made services, produced in interaction with clients -and a large part of KIBS can be considered as such -require direct contact between service suppliers and recipients. In this case, international production through foreign direct investment seems to be a better way of internationalisation. On the other hand, the technological progress in ICTs is constantly creating better and cheaper communication between service providers and their clients, which together with the ease of travel should have a positive impact on the possibility of distant deliveries, including with respect to KIBS. As a result, the issue becomes of the extent to which they are provided by companies operating on domestic markets, and the extent to which they are traded internationally.

Data and research methodology
The study is based on input-output tables from the WIOD database (National Input-Output..., 2013)  Two main indicators are used in the study. The first indicator examines the role of import in providing EU companies with KIBS input, and it takes the form of the share of imported KIBS input in total (domestic and imported) KIBS input (i.e.

∑ ∑
, where j is the industry, i the country, t time, Im imported, and T total).
The second indicator, in the form of the ratio of total or imported KIBS input and GDP (i.e. In order to characterize countries with different imported KIBS intensities, the indicators measuring the role of ( where T means total enterprises) are included in the analysis. Figure 1 shows the EU-27 average structure of KIBS input, taking into account the share of domestic and imported KIBS inputs, as compared with a similar structure for: a) services input; b) manufacturing input; and c) total intermediate inputs. Import plays much a more important role in delivering manufacturing input than services input to EU companies. In 1995, 34% of manufacturing input and only 8% of services input came from import, and in 2011 these disparities were further increased.

Empirical study
While, the growing role of import is visible with respect to both manufacturing and services inputs (including KIBS), the growth was significantly stronger in the case of manufacturing (13 percentage points (pp) compared to 3 pp for services). As a result, in 2011 on average almost half of manufacturing input (ranging from 87.6% in Luxembourg to 31.8% in Italy) came from import, while only 11% of services input (ranging from 60.7% in Luxembourg to 6.5% in France) came from import. The importance of import was slightly higher with respect to KIBS input in comparison to overall services input, and also showed an upward tendency.
In Table 1     Source: own calculations based on data derived from: the source as in Figure 1; National accounts... (2017).
In 2011, the highest KIBS intensity was recorded in Ireland (15.1%) and Luxembourg (11%). KIBS intensities increased in almost all countries, and they generally reached higher values in countries with higher GDP per capita, i.e. in the EU-15 than the EU-12, as well as in the higher-income countries within both groups. A positive correlation between KIBS intensity and GDP per capita occurred with respect to both imported and domestic KIBS inputs, and it was stronger in the case of imported KIBS input (the Pearson's correlation coefficient achieved the values of 0.5 and 0.3, respectively). The analysis based on KIBS input per capita shows even larger disparities between the more and less developed EU countries. Generally, it is not evident that a relatively low level of domestic KIBS intensity was compensated for by a relatively higher level of imported KIBS intensity. Excluding the cases of Ireland and Luxembourg, this is apparent only in Finland and Hungary. Bulgaria, followed by Lithuania, Poland and Romania, recorded the lowest domestic and imported KIBS intensity.
While taking into account a relative importance of imported KIBS, it was also usually higher in countries with higher incomes and the resulting higher KIBS intensities. The situation was different in the largest, high-income EU-15 countries, where relatively high total KIBS intensities were accompanied by the lowest shares of imported KIBS input in total KIBS input. The average importance of KIBS imports was significantly higher for EU-12 than EU-15. One should note, however, that the EU-15's weighted average is underestimated because of very low shares in the largest EU countries, whereas in this group there are also a few smaller economies with the highest reliance on KIBS imports (Ireland, Luxembourg, Netherlands, Finland). Thus, in reality the disparities between these two groups are less significant. In the EU-12 the relationship between market size and the role of imported KIBS was less apparent than in the EU-15.
Most of the EU-12 countries experienced a decrease in the share of imported KIBS input in total KIBS input, with the most significant drop in Latvia and Slovakia. In the EU-15 the trend was generally reversed, and the strongest increase, much higher than in other countries, took place in Ireland (21 pp). A relatively high growth is also visible in Luxembourg (13.8 pp), Greece (7.7 pp) and Spain (6.6 pp), as well as in Poland (12.2 pp). In 2011 the share of imported KIBS input in total KIBS input ranged from 38% in the Netherlands to 6% in Germany. Ireland and Luxembourg clearly stood out among the EU countries, as imported KIBS input was higher than the domestic one, with a high upward trend -as a result in 2011 in Ireland 76% of KIBS input and in Luxembourg 61% came from import (in general, these two countries were distinguished by a relatively high importance of imported service input). Tables 1 and 2 show that the countries with a relatively high importance of imported KIBS usually also enjoyed a relatively high importance of KIBS exports (the Pearson's correlation coefficient for imported KIBS input per capita and KIBS exports per capita amounted to 0.8), and the highest values of KIBS exports per capita, much higher than in other countries, were achieved by Ireland and Luxembourg. KIBS exports played a more important role in the EU-15 than the EU-12, and the disparities between them tend to increase. One should note, however, that in most cases a large part of KIBS trade was carried out by FCEs, e.g. in Ireland they accounted for 61% of KIBS exports and 56% of KIBS imports. The importance of FCEs in KIBS trade was significantly higher in the EU-12 than the EU-15, with their highest share in Romania. EU-15 *** 6.9 2.7 1.2 0.9 26.0 **** 34.1 EU-12 *** 3.9 0.9 0.3 0.3 48.5 **** 46.4 * KIBS exports/KIBS imports by FCEs as % total KIBS exports/total KIBS imports. KIBS defined as comprising 'Professional, scientific and technical activities', and 'Information and communication'. ** Change in the period 1995-2011. *** Weighted averages, with weights assigned based on each country's share in the EU-15's; and EU-12's GDP respectively (all values in current prices, US dollars). **** EU-13 (without LUX and ESP), EU-9 (without CYP, CZE and SVN).

The values of indicators presented in
Source: own calculations based on data derived from: the source as in Table 1; and Trade... (2017).