How much do technological gap, firm size, and regional characteristics matter for the absorptive capacity of ltalian enterprises?
The absorptive capacity represents the ability of enterprises to efficiently absorb and internalise knowledge from outside sources through the adaptation and application of external knowledge sources (Cohen and Levinthal 1989. 1990). Therefore, it represents the link between the firms' capabilities to implement new products and the external stock of technological opportunities. Starting from the seminal works of Cohen and Levinthal (1989, 1990), many scholars (Crespo and Fontoura 2007; Jordaan 2008; Keller 2009) have stressed the importance of the absotptive capacity of domestically-owned firms (DOFs) as one of the main factors for the existence, sign, and magnitude of Foreign Direct Investment (FD1) spillovers in the host countries. Following this strand of literature, the present paper explores whether the effect of FDI on the productivity of Italian DOFs is dependent on their absoiptive capacity. In particular, the main contribution of our analysis is twofold.
First, we present further evidence of inward spillover effects on the Italian manufacturing sector by investigating spillovers at both the intra-industry and inter-industry level, the latter through the empirical analysis of backward and forward linkages between Multinational Enteiprises (MNEs) and domestically-owned suppliers and customers, respectively. Secondly, given the peculiar characteristics of the Italian productive system, we test the capacity of DOFs to absorb external knowledge and technology from MNEs by focusing on three different dimensions, i.e., (i) the size of technological differences, or technology gap, between foreign-owned firms (FOFs) and DOFs, (ii) the size of DOFs, and (iii) the regional distribution of Italian companies. We believe that these three dimensions are very relevant for the Italian productive system since it is characterized by low levels of R&D investment and innovations, the presence of a large number of micro and small firms, and the coexistence of different models of production across the peninsula (alongside the historical dualism between the more advanced North and the less industrialized South).
The reminder of the paper is as follows. The authors first analyse the channels through which DOFs may absorb external knowledge from MNEs. Next, the paper depicts some stylized facts about the Italian economy, stressing the structural characteristics of the productive system. Following that, the paper discusses the estimation strategy and data used and then presents the empirical results obtained. The paper concludes with a summary of our findings.
Absorptive Capacity and FDI Spillovers
The concept of absorptive capacity is based upon the idea that some technical knowledge is generally freely available to all firms, i.e., it can be exploited without paying any fee for its use. The literature identifies two different kinds of absoiptive capacity (Arundel et al. 1998). The first type, which is linked to the process of diffusion or technological transfer across different organisations, concerns skills and expertise required to adopt technologies already developed by other firms. The second type concerns the ability of firms to develop new or improved products and processes by benefiting from discoveries made by other firms or universities. In this framework, it may be of great value to investigate the capacity of firms to adopt and/or develop knowledge and technologies stemming specifically from cross-border interactions among enterprises. In particular, it is worth exploring the ability of DOFs to adapt and/or apply external knowledge and technology spilled-out from MNEs through inward FDI.
According to the literature, knowledge may spill over from FOFs to DOFs through several channels. The first relates to demonstration effects, since the presence of MNEs can provide DOFs with the opportunity to learn by watching that indirectly contributes to the rising intensity of domestic R&D: DOFs can observe FOFs' technologies and management practices and imitate them in their own operations, thus increasing their productivity (Blomstrom and Kokko 1998). The second channel occurs through domestic linkages both at a backward level, i.e., by subcontracting activities between MNEs and suppliers, and at forward level, i.e., between MNEs and domestic buyers. When MNEs build these backward and forward linkages with domestic suppliers and distributors, knowledge from FOFs is transmitted to the suppliers, distributors, and ultimately to DOFs that use the same suppliers and distributors (Spencer 2008). The third channel is the workers' mobility and training arising from skills of workers, managers, engineers, etc. which is acquired from FOFs and then transferred to DOFs (Blomstrom and Kokko 1998).(1)
However, the capacity of firms to absorb (adapt and/or develop) external knowledge from MNEs may vary according to the different dimensions of the companies. In particular, the magnitude of inward FDI spillovers may be strongly affected by the internal capabilities of domestic-owned enterprises, above all in terms of the size of technological differences (or technological gap) between FOFs and DOFs, the size of the domestic firms, and the regional distribution of Italian companies.
The degree of the technological gap between FOFs and DOFs, i.e., the extent to which FOFs in an industry are technologically advanced by comparison with DOFs in the same industry, represents an important spillover determinant both at a horizontal and vertical level. More specifically, at the horizontal level, the extent of spillovers is likely to depend on the technological sophistication of DOFs; similarly, at the vertical level, the extent of backward (forward) linkages between MNEs and domestic suppliers (buyers) of intermediate goods is likely to depend upon the stock of technological capabilities of DOFs in supplying (buying) sectors.
It is worth stressing that from both a theoretical and empirical point of view, it is not clear what the relation between the level of the technological gap (whether small or large) and spillovers should be, since the absorptive capacity literature suggests two opposing arguments. The first argument, proposed originally by Findlay (1978) and confirmed by several other works like Wang and Blomstrom (1992), Blomstrom and Wolff (1994), and more recently, by Jabbour and Mucehielli (2007) and Jordaan (2008), argues that the potential for positive spillovers is higher when there is a large technological gap between DOFs and MNEs. This assumption is based on the idea that firms with lower stocks of technology have a greater scope for technological accumulation in that they have a larger backlog of established knowledge to assimilate.
The second argument states that when the technological gap is too large, the diversity of MNEs may have a weak impact on the productivity of the DOFs. since MNE affiliates may be too advanced to leave any mark on host country firms. Cantwell (1989), for instance, states that a firm's ability to follow and adapt the technological developments of other firms depends largely on its existing technological capability. The basic idea is that, for a large technological gap, DOFs have no internal knowledge resources that enable them to recognize the value and content of a variety of knowledge elements brought by MNEs, thus making spillovers unlikely to occur. From an empirical point of view, such an argument is supported by Kokko (1994), and more recently by Dimelis (2005), Takii (2005), and Hamida and Gugler (2009).
Firm size may also influence inward FDI spillovers insofar as large firms are more able than small ones (i) to recognize, understand, and learn technologies and management practices brought by MNEs. (ii) to spread the fixed costs of R&D over larger sales, and (iii) to exploit economies of scale and scope in R&D activities. Moreover, large firms possess larger stocks of internal resources and knowledge that can be employed as a complementary asset to the technology transferred from MNEs. Thus, large DOFs have more internal capabilities that can be used to exploit FDI spillovers (Zhang et al. 2010). On the other hand, small medium enterprises (SMEs) could be hampered in their ability to absorb new technology from MNEs because of a lack of scientific and technical staff or experience.
Finally, the capacity of DOFs to exploit external knowledge may be affected by the structural characteristics of the domestic productive system, for example in teams of regional development, sectoral innovation system, institutional context, etc. In some cases, it is possible to observe the coexistence of different productive models within the same country. For instance, organizational and production models based on SMEs and specialized in traditional industries can coexist with science-based industries characterized by an intensive use of technical and scientific knowledge inputs. On these grounds, it would be worth investigating to what extent the specific characteristics of local productive systems can affect the absorptive capacity of DOFs.
The Italian Case: Some Stylized Facts
In the empirical analysis, we test the effects of inward FDI spillovers on the absorptive capacity of Italian-owned firms. The Italian case is relevant for a number of reasons. First of all, in the last 10 years, Italy has received increasing flows of inward FDI, whose value passed from 6.911 million dollars in 1999 to 40.202 million dollars in 2007 (ICE 2010). In 2007, the number of foreign-controlled firms amounted to 14,401, with 1,246,794 workers employed. Specifically, the number of FOFs in the manufacturing sector was 3,301 (with 466,698 workers employed). Moreover, FOFs performed better than their domestic counterparts, since they were more productive, had more workers, and were more profitable (ISTAT 2010a). In this framework, it is worth exploring if DOFs were able to exploit the indirect effects arising from the presence of MNEs in terms of positive externalities.
Secondly, indicators of R&D effort are not favourable to Italy. R&D expenditures (in both the private and public sector) was slightly above 1 % of GDP in 2008, compared with the OECD average of 2.3 % (OECD 2009), although under-recording of R&D activity in SMEs, where it is often performed informally, may bias these figures somewhat downwards. There arc a number of reasons why R&D activity and innovation arc low in Italy. In particular, the small size of most Italian firms implies difficulty in meeting the up-front cost of R&D with only limited access to external capital. The last Italian Innovation Survey (ISTAT 2010b) shows that, in 2008, large enterprises were the most innovative (65.1 %), against SMEs that were innovative at respectively 49.8 % and 28.2 % of the total. This low propensity to innovate, which is typical of Italian firms, may suggest the presence of a relevant technological gap between DOFs and FOFs that may affect the capacity of Italian DOFs to exploit technological spillovers from MNEs.
Thirdly, the Italian productive system is characterized by a large presence of micro and small firms. In 2007, the number of firms with only one employee amounted to 2,593,079 (61.0 %), and firms with two to nine employees were 1,654,102 (38.0 %) (ISTAT 2008). On the contrary, large firms (with 250 employees or more) amounted to only 3,630 (less than 1 %). The vast presence of micro and small firms makes Italy an interesting case for analysing the hypothesis according to which small enterprises are hampered in their ability to absorb new technology from inward FDI spillovers because of the lack of scientific and technical staff and/or experience.
Finally, the Italian economy is historically characterized by a social-economic dualism between the more advanced North and the less industrialized South of the peninsula. In particular, the level of per capita income in the southern regions was 17,324 euro in 2009, a much lower value than that observed in the centre-north (29,399 euro). Such a substantial gap is also persistent over time, given that it has not significantly changed over the last 30 years (ISTAT 2010c). Moreover, in 2007 about 43 % of Italian firms were localised in the North (29.0 % in the Northwest, 22.0 % in the North-East), whereas only 28 % were located in the South (ISTAT 2008). According to EUROSTAT figures, Italy is the only European country having, at the same time, a per capita income in line with the continental average together with a huge percentage of the population (29 %) that lives in a province where per capita income is less than 75 % of the EU average, as well as 26 % of the population that resides in a province with a level of per capita income equal to 125 % of the average.
Along with such dualism between the North and South, Italy is characterized by the coexistence of different productive and innovative systems. In particular, the northwestern area is specialized in capital intensive and large-scale industries, and in the so-called science-based industries, characterized by an intensive use of technical and scientific knowledge inputs. Since in such a network enterprise model, firms arc already endowed with higher levels of technological capability, they can exploit the positive externalities arising from networking relationships with other enterprises. The northeastern area is characterized by the presence of industrial district models (Bccattini 1987; Brusco 1986), i.e., organisational and production models based on SMEs which are mainly specialized in traditional industries and that create a network of flexible relationships strictly linked to the local economic and social context. Firms situated in the industrial districts generally have a self-propelling capacity to achieve efficiency and to be competitive at an international level. Finally, the productive system in the central-southern areas is founded upon a backward model of production, characterized by an atavistic lower level of industrialization and by different basic social conditions. On these grounds, we investigate the existence of spillover effects in the Italian economy by also taking into account the geographical distribution of firms, along with the traditional dimensions of DOF size and the technological gap between DOFs and FOFs.
Estimation Strategy and Data Used
The traditional approach to the analysis of productivity consists in estimating a production function and men in using the residuals not explained by the input factors (capital, labour) as a proxy for the Total Factor Productivity (TFP) (Solow residuals). However, as Levinsohn and Petrin (2003) have shown, when estimating the production function, one must account for the correlation between input levels and productivity, since profit-maximizing firms respond to increasing productivity by an increased use of factor inputs. Thus, methods that ignore this endogeneity problem (such as OLS or the fixed-effects estimator) inevitably lead to inconsistent estimates of the parameters of the production function. For this reason, in line with the recent literature, we employ the semi-parametric approach suggested by Olley and Pakes (1996), and then modified by Levinsohn and Petrin (2003). This method allows for firm-specific productivity differences that exhibit idiosyncratic changes over time. In principle, the method estimates a traditional Cobb-Douglas production function, taking into account that the error term has two components, one of which is correlated with the choice of inputs by the firm, but is not observable by the econometrician. The authors develop an estimator that uses a free variable such as intermediate inputs (material costs or fuel or electricity) as a proxy for this unobservable productivity shock.
Following this technique, we use a two-step method. In the first step, we estimate a log-log transformation of a traditional Cobb-Douglas production function using the Levinsohn and Petrin semi-parametric approach. In the second step, we investigate FDI spillovers by estimating the following equation:
[TFP.sub.ijt] = [sigma] + [[theta].sub.1][HERFl.sub.jt] + [[theta].sub.2][SCALE.sub.jt] + [[theta].sub.3][FDISECTOR.sub.jt] + [[theta].sub.4][HSPILL.sub.jt]+ [[theta].sub.5][BACKSPILL.sub.jt] + [[theta].sub.6][FORSPILL.sub.jt] + [[theta].sub.7][D.sub.t] + [[mu].sub.i], + [[epsilon].sub.ijt] (1)
where we relate the total factor productivity for firm i (TFP), to the foreign presence variables (HSPILL, BACKSPILL and FORSPILL) and to other control variables, i.e., the level of competition within the sector (HERF), the economics of scale (SCALE), and the size of the sector (FDISECTOR). In the Eq. (1) [[mu].sub.i] is a firm-level fixed effect that allows us to consider the unobserved firm heterogeneity, [D.sub.t] are yearly dummies to control for unobserved aggregate shocks that are common to all DOFs, [[epsilon].sub.it] the stochastic disturbance term that we assume to be independently distributed. Table 1 summarizes the explanatory variables used in estimations.
Table 1 Definition of variables used in Eq. (1) Variables HERFI SCALE FDISECTOR HSPILL BACKSPILL FORSPILL Description Herfindahl index of turnover, used as a proxy for the level of concentration and thus competition within the sector and year. It is constructed as: [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] It can be readily deduced that HERFI is bound between 0 and 1 and that higher HERFI values indicate greater market concentration, i.e. less competition. Minimum efficient scale of the industry, measured as the ratio between firms' sales above the average sales for the industry, divided by total industry sales. It is employed as a proxy for economics of scale (Comanor and Wilson 1967) Sum of the number of employees at time t by all foreign-owned firms operating in sector j. Share of foreign firms' output in total sector output. It accounts for the foreign presence in the same sector: [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] Foreign presence in linked downstream sectors (to which a DOF supplies its inputs): [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] where [gamma]jkt is the proportion of the or j's output supplied to sourcing sectors k obtained from the input-output table for domestic intermediate consumption (i.e. excluding imports). Forward vertical spillovers to DOFs that buy inputs from foreign firms: [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] where [delta]1jt is the proportion of sector j's inputs purchased from upstream sectors l.
The empirical analysis has been conducted by using firm-level data from the AIDA database (Analisi Informatizzata Delle Aziencle) provided by the Bureau Van Dijk. The AIDA database collects annual accounts of Italian corporate enterprises and contains information on a wide set of economic and financial variables, such as sales, costs and number of employees, value added, fixed tangible assets, R&D, start-up year, as well as the sector of activity and the ownership status. In order to study the spillover effects of FOFs on DOFs, we have identified as foreign those firms that are majority owned, wholly owned, or whose main shareholder is foreign. By omitting all observations for which the necessary data are incomplete, we obtained an unbalanced panel of about 563,000 observations over the period 2002-2007. The advantage of using this dataset is twofold. Firstly, it is highly representative of the entire universe of corporate companies (e.g., in 2007, our sample covers about 87 % of total employees declared by the Italian Institute of Statistics--ISTAT 2008). Secondly, it reflects quite well the geographical and size distribution of firms in the Italian economy, which is characterized by a high weight of small and medium-sized enterprises. All nominal variables included in the database were deflated using an appropriate producer price index provided by ISTAT, which also provided the input-output matrix adopted to test for the presence of vertical spillover.
Tables 2 and 3 report some descriptive statistics about our sample. They widely confirm the figures from ICE (2010) and 1STAT (2010a) reported earlier in the paper. In particular. Table 2 contains the means of the variables for the whole sample distinguished by ownership type, as well as tests for the comparison of means of two groups of firms (DOFs versus FOFs). All figures presented in the table are averages over the sample period. Focusing our attention on some firm and industry level variables, we observe that FOFs are on average larger, more productive, and more profitable than DOFs. They also operate in more concentrated industries and with a higher minimum efficient scale.
Table 2 Mean statistics by ownership status and t-test for the comparison of two means for the distribution (DOFs versus FOFs) Definition SIZE Firm size measured by the number of employees TFP Total Factor Productivity WAGE Firms' average wage TECH R&D intensity as the ratio of R&D expenditures on sales NET PROFIT Firms' net profit SCALE Minimum Efficiency Scale of Industry HERF Herfindhal concentration ratio at industry level Mean Diff. t DOFs FOFs 27.5 216.1 -188 -40.5*** 9.5 10.4 -0.9 -72.9*** 24925 35056 -10130 -0.2 0.0123 0.0024 0.0098 0.1 152991 1732627 164529 6.3*** 0.006 0.015 -0.008 -20.9*** 269 456 186 -20.3*** Source: own elaboration based on the A IDA database *** = statistically significant at 0.01 % level Table 3 Distribution of Italian firms by size, ownership status and regional location (percentages, sample average) DOFs FOFs TOTAL SIZE_1_49 99.7 0.3 89.9 SIZE_50_249 96.7 3.3 8.8 SIZE_>250 88.5 11.5 1.3 TOTAL 99.3 0.7 100.0 NORTH-WEST 98.8 1.2 34.4 NORTH-EAST 99.4 0.6 28.7 CENTRE 99.6 0.4 19.6 SOUTH 99.8 0.2 17.3 TOTAL 99.3 0.7 100.0 Source: own elaboration based on the AIDA database
Table 3 compares the distribution of Italian firms by ownership status, regional location and size (small, medium, and large firms), the latter measured by the number of employees.2 According to the figures, DOFs represent the largest percentage of Italian firms (99.3 %), and are mainly of small size, while the share of FOFs is very small (0.7 %). It also appears that FOFs arc mainly of large size and are mostly concentrated in the northwestern region of Italy.
Empirical Results and Interpretations
As shown in Table 2, FOFs outperform DOFs in terms of productivity levels, thus we expect to detect some productivity spillovers in our analysis. Table 4 presents the results of the estimation of Eq. (1).
Table 4 Estimation of the Eq. (1) Regressors Dependent variable: TFP Coefficient Robust Stand. Err. Cons 9.523*** 0.291 HERFI -0.012** 0.006 SCALE 0.122 0.220 FDISECTOR 0.001 0.011 HSPILL 0.109 0.070 BACKSPILL 0.241 0.390 FORSPILL 0.007** 0.003 Time dummies Yes Adjusted [R.sup.2] 0.636 n OBS 562.745 All regressions are restricted to domestic firms. Robust standard errors in brackets. Areg estimation was performed to tit a linear regression absorbing one categorical factor. *** = statistically significant at 0.01 % level. ** - statistically significant at 0.05 % level
Our main findings can be summarized as follows. First, the negative and significant coefficient of the concentration level (HERFI) suggests that less concentrated sectors (i.e., sectors with more competition) benefit more in terms of productivity increases. On the other hand, the economies of scale (SCALE) and the size of sector (FDISECTOR) are positive and not significant. With regard to the spillover effect, our results suggest, on one hand, the absence of both horizontal and backward spillovers (their coefficients are positive but not statistically significant), and, on the other, the existence of positive forward spillovers. Our findings highlight that the mere fact that being a customer of foreign companies has a beneficial effect on DOF productivity. In other words, the Italian DOFs arc able to improve themselves once they are offered products and services from MNEs from upstream sectors. In general terms, such results are consistent with the findings of other studies which argue that it is more likely that FDI spillovers will take place through vertical linkages (i.e., backward and/or forward spillovers) as opposed to horizontal ones (Smarzynska 2004; Crespo and Fontoura 2007; Blalock and Gertler 2008). This is because MNEs have an incentive to prevent information leakage to their competitors, including DOFs, thus reducing the possibility of horizontal spillovers. By contrast, the existence of forward spillovers is plausible since MNEs in upstream industries may provide inputs to domestic firms that were previously unavailable in the country, or make them technologically more advanced or less expensive, or ensure that they are accompanied by the provision of complementary services (sec Smarzynska 2004). Thus, when MNEs arc involved with DOFs in downstream sectors, the latter gain technological benefit from the former.
With regard to similar studies applied specifically to the Italian case, these have often produced ambiguous results, probably because they employed a different dataset, adopted dissimilar econometric methodologies, and explored different periods. Our results seem broadly to confirm the lack of horizontal spillovers as in Imbriani and Reganati (2004) and Reganati and Sica (2007) who find evidence of positive but not statistically significant intra-industry spillovers. Different results were found by Castellani and Zanfei (2007) who found positive and significant spillovers when they controlled for the size of the industry.
After exploring our general findings. Tables 5, 6, and 7 show the results obtained when we split our sample to detect differences in the pattern of spillovers across different groups of firms (so-called conditional spillovers). In particular, we employ breakdowns by (i) technological gap (Table 5), (ii) firm size (Table 6), and (iii) regional distribution of enterprises (Table 7).
Table 5 Group estimation according to the technological gap Regressors Dependent variable: TFP (i) High gap (ii) Medium gap (iii) Low gap Cons 0.008(0.013) 5.444*** (0.517) 11.111*** (0.461) HERFI 0.018(0.13) 0.043*** (0.004) 0.054*** (0.012) SCALE 0.237 (0.292) -2.326*** (0.401)1.631** (0.645) FDISECTOR 0.065** (0.031) 0.155*** (0.020) -0.040** (0.018) HSPILL 0.193(0.138) 1.537***(0.I36) 1.105*** (0.159) BACKSPILL 0.304 (0.763) 6.381*** (0.699) 2.326*** (0.864) FORSPILL -0.057*** (0.008)0.185*** (0.022) 0.0187** (0.008) Time dummies Yes Yes Yes Adjusted [R.sup.2] 0.655 0.707 0.770 nOBS 169.951 262,151 130,643 All regressions are restricted to domestic firms. Robust standard errors in brackets. Areg estimation was performed to fit a linear regression absorbing one categorical factor. *** = statistically significant at 0.01 % level. ** = statistically significant at 0.05 % level Table 6 Group estimation according to the firm size Regressors Dependent variable: TFP (i) Small firms (ii) Medium firms (iii) Large firms Cons 8.608*** (0.327) 11.809*** (0.548) 10.770*** (1.619) HERFI -0.014** (0.006) 0.0004 (0.0080) 0.0159* (0.009) SCALE 0.247 (0.240) 0.311 (0.245) -0.675 (0.522) FDISECTOR 0.035*** (0.013) 0.067*** (0.021) -0.010 (0.064) HSPILL 0.047 (0.076) 0.316** (0.135) 0.620 (0.428) BACK.SP1LL 0.438 (0.423) -0.538 (0.776) -1.184 (2.320) FORSPILL 0.008** (0.007) 0.004 (0.003) 0.010 (0.055) Time dummies Yes Yes Yes Adjusted [R.sup.2] 0.6114 0.725 0.801 n OBS 505.293 50.688 6,764 All regressions arc restricted to domestic firms. Robust standard errors in brackets. Areg estimation was performed to fit a linear regression absorbing one categorical factor. *** = statistically significant at 0.01 % level. ** = statistically significant at 0.05 % level. * = statistically significant at 0.10 % level Table 7 Group estimation according to the geographical area Regressors Dependent variable: TFP (i) South (ii) Centre (iii) North-East Cons 8.244*** (0.863)8.591*** (0.688) 8.854*** (0.512) HERFI -0.002(0.010) -0.004(0.016) -0.186* (0.010) SCALE -0.038 (0.523) 0.010(0.405) 0.268 (0.366) FDISECTOR 0.040 (0.034) 0.035 (0.027) 0.030 (0.020) HSPILL 0.054(0.162) -0.012(0.160) 0.118(0.139) BACKSPILL 0.744(1.082) 0.874 (0.880) 0.049 (0.744) FORSPILL 0.008*(0.005) -0.007(0.010) 0.014*** (0.004) Time dummies Yes Yes Yes Adjusted [R.sup.2] 0.556 0.603 0.652 n OBS 94,851 109,105 164.255 (iv) North-West 9.339*** (0.427) 0.0175** (0.008) -0.108(0.323) 0.014 (0.017) 0.182* (0.107) 0.358 (0.601) 0.036* (0.021) Yes 0.654 194.534 All regressions arc restricted to domestic firms. RobusI standard errors in brackets. Areg estimation was performed to fit a linear regression absorbing one categorical factor. *** = statistically significant at 0.01 % level. ** = statistically significant at 0.05 % level. * = statistically significant at 0.10 % level
In this study, technological gap has been defined in terms of the relative productivity performance of DOFs vis-a-vis FOFs in the same sector. Thus, the technological gap [TG.sub.ij] for a firm i is measured in terms of TFP gap, i.e., as the difference between the productivity of the average FOFs in the sector and each firm in the sector (sec Flores et al. 2007; Jabbour and Mucchiclli 2007). It is worth noting that, following the main literature, we use the terms 'productivity gap' and Technological gap' interchangeably, although the concepts are not exactly the same. Indeed, technological gap can be defined as the difference in the techniques available for production, whereas productivity gap represents the difference in productivity when the same technology is used (Kathuria 2010). Since determining the technological gap is often tricky, most of the empirical work (including ours) has proxied the technological gap through measures of productivity gap, the general idea being that a more productive FOF is a reflection of the technological gap between FOFs and the DOFs.
We check for the sensitivity of the model to alternative ranges of gap by adopting a sub-sample strategy, i.e., by splitting the sample into three groups according to the level of absorptive capability. By employing an exogenous grouping model, we select some ad hoc values from the observations to divide the sample into three sub-samples (low, medium, and high gap). In particular, the group with low TG consists of firms with TG below the 25th percentile of the TG distribution across all domestic firms; the medium TG group contains firms with TG between the 25th and 75th percentiles, while the high TG group includes firms with TG above the 75th percentile. Table 5 provides the estimates of Eq. (1) for the three above-defined sub-samples.
When the technological gap between FOFs and DOFs is wide (column (i)), we find evidence that DOFs in downstream sectors receive a negative externality from FDI. This may indicate the hypothesis that inputs produced locally by FOFs are more expensive and less adapted to local requirements, since MNEs are too technologically advanced by comparison with local enterprises.
When the technological gap between FOFs and DOFs is both low and medium (columns (ii) and (iii)), DOFs enjoy positive spillovers from FOFs operating in the same sector. This result is in line with the literature (Imbriani and Reganati 1999 for Italy, and Girma and Wakelin 2000 for the UK) and shows that domestic firms find it easier to benefit from FDI when the foreign presence strengthens an already existing domestic technological capability. In addition, in the same groups of firms, we find negative spillovers from FOFs to their local suppliers in the upstream sector. The explanation for this effect could be that FOFs benefit from their ability to diversify the supply network through their knowledge of the market at both national and international level. Moreover, they can exert a sort of a monopsony power, by imposing lower prices on their domestic suppliers (Driffield et al. 2004).
Finally, the existence of positive forward spillovers suggests that when the technological gap is either low or medium, DOFs benefit from supplies of intermediate goods and machinery from MNEs since, for example, the latter provide better quality products and lower costs that enhance the productivity of Italian firms using these inputs and/or because DOFs may receive support in the form of training in sales techniques and supply of sales equipment from MNEs, thus generating more positive externalities. Overall, with regard to the general case, we can conclude that the level of technological gap matters considerably for significance and sign of spillovers in the Italian case.
Table 6 presents the estimation results by considering the firm size of DOFs. As we can see, only small-sized companies (column (i)) are able to benefit from forward spillovers, the coefficient being positive, whereas only medium-sized companies (column (ii)) benefit from the presence of MNEs in the same productive sector (HSPILL is positive and significant).
The situation for smaller companies echoes the overall results reported in Table 4 with only positive forward spillovers. These results seem to suggest that the existence of positive forward spillovers at a national level is mainly due to small DOFs. In other words, MNEs in upstream industries make DOFs technologically more advanced by providing them with inputs that are not available in the country or by ensuring them complementary services. These results could reflect the structure of the Italian productive system discussed previously in the paper, which is typically founded upon small firms. It should be noted, however, that in the case of small firms, two opposite effects are generally possible. On one hand, they have only limited sources for improving their technologies; on the other, they may be more flexible and able to adjust more quickly to a new market situation. With regard to our analysis, the results seem to suggest that smaller firms' flexibility prevails over the effect of their limited sources.
Finally, the existence of horizontal spillovers in the case of medium-sized enterprises highlights the ability of such a group of Italian firms to interact better than larger and smaller firms with FOFs, and thus to exploit the technology brought by MNEs. As Schiliro (2011) stressed, medium-sized Italian films with their high-flexible business model have an organizational structure that is much more able to withstand the challenges of globalization. Moreover, despite the diversity of territories and sectors to which they belong, Italian medium-sized firms have similar characteristics to MNEs and share common strategies (Musca and Schiliro 2012).
Table 7 provides the estimates for productivity spillovers at sub-national level.3 The table suggests the presence of positive and significant forward spillovers from FDI in almost all the Italian sub-regions (specifically in the South, Northeast, and Northwest, as well as the absolute lack of backward spillovers in any Italian sub-region. Such results perfectly reproduce our overall findings at the national level (Table 4), meaning that the presence of positive forward linkages between FOFs and DOFs (but also the lack of any backward spillover) does not have a geographical dimension.
The presence of positive and significant horizontal spillovers in the northwestern area of Italy broadly confirms the findings of Imbriani and Reganati (1999, 2003), whose studies provided evidence for the existence of intra-sectoral spillovers in the northwestern region of Italy and, at the same time, rejected the presence of horizontal spillovers at the national level. Such a result is strongly concerned with the structure of the Italian productive system discussed earlier in the paper, which is characterized by an economic dualism even within the northern area, the Northwest being more advanced in terms of productive and innovative systems compared to the Northeast. Thus in the Northwest, which is characterized by a typical network enterprise model of production, through FDI, local firms are able to capture the benefits arising from the spillovers because the foreign presence strengthens the already existing domestic technological capability.
Finally, the reason for the lack of spillovers in northeast, central, and southern of Italy needs to be differently addressed on the grounds of the deep socio-economic differences across the three areas. The northeastern fans, organised mainly in typical industrial districts, are generally SMEs characterized by a self-propelling capacity to achieve efficiency and to be competitive at an international level: the possibility of horizontal spillovers for such firms is consequently weak because of the different model of organisation and production. However, the lack of horizontal spillovers in the central and southern regions is mainly concerned with/due to the different basic conditions (above all in social terms), which make the localisation of investments unattractive both to domestic and foreign capital. Moreover, FOFs, when present, crowd out the domestic firms, so mat no company is able to absorb the potential spillovers.
This paper aims to verify whether the effect of FDI on the productivity of Italian DOFs is dependent on their absorptive capacity, through the analysis of inward FDI spillovers at both an intra-industry (horizontal) and inter-industry (backward and forward level) in the Italian manufacturing sector. In particular, given the specific characteristics of the Italian productive system, we test the capacity of DOFs to absorb external knowledge and technology from MNEs by focusing on three different dimensions, i.e., (i) the technological gap between FOFs and DOFs, (ii) the size of DOFs, and (iii) the regional distribution of Italian companies. Overall, our findings can be summarized as follows:
(*.) The strongest channel through which Italian firms benefit from the presence of foreign companies is represented by forward spillovers. Being a customer of foreign companies has a beneficial effect on a firm's productivity: domestic firms seem in fact to benefit from supplies of intermediate goods and machinery from MNEs in the upstream sectors, probably because they provide better quality products at lower costs, as well as providing support to DOFs in the form of training in sales techniques and supply of sales equipment.
(*.) The level of technological gap between FOFs and DOFs matters considerably for the spillover effect: when it is high, DOFs in downstream sectors receive a negative externality from FDI, probably because inputs produced locally by FOFs are too expensive and less adapted to local requirements, given that MNEs are too technologically advanced compared to local enterprises. When the gap is medium-low, we find evidence of positive spillovers at a horizontal level and negative spillovers from FOFs to their local suppliers in the upstream sector. The existence of intra-industry spillovers can depend upon the ability of foreign presence to strengthen the already existing domestic technological capability, while the reason for negative forward spillovers is founded on the FOFs' capacity to diversify the supply network or the possibility of imposing lower prices on their domestic suppliers, given their monopolistic power.
(*.) Similarly, firm size matters for the spillover since small firms take advantage of positive externalities from MNEs in the upstream sectors, and medium-sized companies at intra-industry level. These results reflect the structure of the Italian productive system, typically founded on small firms other than the organizational structure of medium-sized Italian DOFs, which exhibit similar characteristics and share common strategies with MNEs despite the diversity of territories and sectors they belong to.
(*.) Finally, the presence of positive and significant forward spillovers from FDI in almost all the Italian sub-regions (specifically in the South, Northeast, and Northwest), as well as the absolute lack of backward spillovers in any Italian sub-region reproduce our overall findings at national level, thus suggesting that the presence of positive forward linkages between DOFs and FOFs (but also the lack of any backward spillover) does not exhibit any geographical dimension.
Published online: 9 October 2013
[c] International Atlantic Economic Society 2013
(1.) It is important to consider that productivity gains for domestic firms might be transmitted through the market mechanism (pecuniary externalities). Multinational entry may lead to a higher degree of competition which, as a result, might induce firms to reduce inefficiencies, and thus increase productivity. However, the increase in competition through FDI may also reduce the market share of domestic firms, which could reduce productivity in the presence of scale economics.
(2.) Where small firms have 1-49 employees, medium firms 50 249, large firms more than 250 employees.
(3.) The northwestern region includes Lombardy. Piedmont, Liguria and Vallc d'Aosta; the northeastern region is composed of Friuli Venczia Giulia, Trentino Alto Adige, Veneto, and Emilia Romagna; the centra] region is composed of Tuscany, Marche, Lazio, Umbria, and finally the southern area includes Abruzzo, Molisc, Campania, Calabria. Basilicata, Apulia, Sicily, and Sardinia.
Arundel, A., Smith, K., Patel, P., & Sirilli. G. (1998). The future of innovation measurement in Europe-Concepts. Problems and Practical Directions. IDEA Paper Series n. 03
Becattini. G. (1987). The market and local forces: The industrial district. Bologna: IL Mulino.
Blalock, G., & Gertler, P. (2008). Welfare gains from Foreign Direct Investment through technology transfer to local suppliers. Journal of International Economics, Elsevier, 74(2), 402-421.
Blomstrom, M., & Kokko, A. (1998). Multinational corporations and spillovers. Journal of Economic Surveys, 112, 247-277.
Blomstrom, M., & Wolff, E. N. (1994). In W. J. Baumol. R. R. Nelson, & E. N. Wolff (Eds.), Multinational corporations and productivity convergence in Mexico in Convergence of Productivity: Cross-National Studies and Historical Evidence (pp. 263-284). Oxford: Oxford University Press.
Brusco, S. (1986). Small firms and industrial districts - the esperience of Italy. In E. Wever & D. Kecble (Eds.), New firms and regional development (pp. 184-202). London: Croom Helm.
Cantwell, J. (1989). Technological innovation and multinational corporations. Oxford: Basil Blackwell.
Castellani. D., & Zanfei. A. (2007). Multinational companies and productivity spillovers: is there a specification error? Applied Economics Letters, 14, 1047-1051.
Cohen, W. M., & Levinthal, D. A. (1989). Innovation and learning: the two faces of R&D. The Economic Journal, 99, 569-596.
Cohen, W. M., & Levinthal, D. A. (1990). Absorptive capacity: a new perspective on learning and innovation. Administrative Science Quarterly, 35, 128-152.
Comanor. W. S., & Wilson, T. A. (1967). Advertising, market structure and performance. Review of Economics and Statistics, 49, 423-440.
Crespo, N., & Fontoura, M. P. (2007). Determinant factors of FDI spillovers-what do we really know? World Development, 35, 410 425.
Dimelis, S. (2005). Spillovers from foreign direct investment and firm growth: technological, financial and market structure effects. International Journal of the Economics of Business. 12, 85-104.
Driffield, N., Munday, M., & Roberts. A. (2004). Inward investment, transaction linkages and productivity spillovers. Papers in Regional Science, 83, 699-722.
Findlay, R. (1978). Relative backwardness, direct foreign investment and the transfer of technology: a simple dynamic mode. Quarterly Journal of Economics, 92, 1-16.
Flores, R. G., Jr., Fontoura, M. P., & Santos, G. S. (2007). Foreign direct investment spillovers in Portugal: additional lessons from a country study. The European Journal of Development Research, 19, 372-390.
Girma, S., & Wakclin, K. (2000). Are there regional spillovers from FDI in the UK? GEP research paper 00/16. Hamida: University of Nottingham.
Hamida, L. B., & Gugler, P. (2009). Are there demonstration-related spillovers from FDI? Evidence from Switzerland. International Business Review, 18, 494-508.
ICE (2010) L'Italia nell' economia internazionale Rapporto ICE - 2009-2010.
Imbriani, C., & Reganati, F. (1999). Productivity spillovers and regional differences: Some evidence on the Italian Manufacturing Sector. Centro di Economia del Lavoro e di Politica Economica (CELPE), University of Salerno, Discussion Paper 48.
Imbriani, C., & Reganati, F. (2003). The Effects of Foreign Direct Investments on Local Firms. Some Evidence from Panel Data for the Italian Manufacturing Sector. Paper presented at the 2003 Hawaii International Conference on Social Sciences. June, 12-15.
Imbriani, C., & Reganati, F. (2004). The effects of foreign direct investment on local firms: some evidence from panel data for the Italian manufacturing sector. The Journal of Current Research in Global Business. 9, 18-24.
ISTAT (2008). Archivio Statistico della Imprese Altive. Italian National Institute of Statistics.
ISTAT (2010a). Struttura e attivita delle imprese a controllo estero - Anno 2008. Italian National Institute of Statistics.
ISTAT (2010b). Einnovazione nelle imprese ilaliane - Anni 2006-2008. Italian National Institute of Statistics.
ISTAT (2010c), Conti Economici Regionali. 1995-2009. Rome.
Jabbour, L., & Mucchielli. J. L. (2007). Technology transfer through vertical linkages: the case of the Spanish manufacturing industry. Journal of Applied Economics X. 115-136.
Jordaan, J. A. (2008). Intra- and inter-industry externalities from foreign direct investment in the Mexican manufacturing sector: new evidence from Mexican regions. World Development. 36, 2838-2854.
Kathuria, V. (2010). Does the technology gap influence spillovers? A post-liberalization analysis of Indian manufacturing industries. Oxford Development Studies, 38, 145 170.
Keller, W. (2009). International Trade, Foreign Direct Investment, and Technology Spillovers. NBER Working Papers 15442. National Bureau of Economic.
Kokko, A. (1994). Technology, market characteristics, and spillovers. Journal of Development Economics, 43, 279-293.
Levinsohn, J., & Petrin, A. (2003). Estimating production functions using inputs to control for unobservables. Review of Economic Studies, 70, 317-341.
Musca M. & Shiliro D. (2012). An investigation into medium-sized multinational enterprises. Munich Personal RePEc Archive MPRA Paper No. 43012. December 2012.
OECD (2009). Economic Surveys - Italy. Organisation for Economic Co-Operation and Development.
Olley, G. S., & Pakes, A. (1996). The dynamics of productivity in the telecommunications equipment industry. Econometrica, 64, 1263 1297.
Reganati, F., & Sica, E. (2007). Horizontal and vertical spillovers from FDI: evidence from panel data for the Italian manufacturing sector. Journal of Business Economics and Management, 8, 259 266.
Schiliro, D. (2011). Innovation and performance of Italian multinational enterprises of the "fourth capitalism". Journal of Advanced Research in Management. II(2), 89 103.
Smarzynska, J. B. (2004). Does foreign direct investment increase the productivity of domestic firms? In search of spillovers through backward linkages. American Economic Review, 94, 605 627.
Spencer, J. W. (2008). The impact of multinational enterprise strategy on indigenous enterprises: horizontal spillovers and crowding out in developing countries. Academy of Management Review, 33, 341-361.
Takii, S. (2005). Productivity spillovers and characteristics of foreign multinational plants in Indonesian manufacturing 1990-1995. Journal of Development Economics, 76, 521-542.
Wang, Y., & Blomstrom, M. (1992). Foreign investment and technology transfer: a simple model. European Economic Review, 36, 137-155.
Zhang, Y., Li, H., Li, Y., & Zhou, L. A. (2010). FDI spillovers in an emerging market: the role of foreign firms' country origin diversity and domestic firms' absorptive capacity. Strategic Management Journal Early View, 31, 969-989.
Although this work was jointly conceived and produced by the four authors. "Introduction" section was written by Filippo Reganati, "Absorptive Capacity and FDI Spillovers" and "Estimation Strategy and Data Used" sections by Rosanna Pittiglio, "The Italian Case: Some Stylized Facts" and "Empirical Results and Interpretations" sections by Edgardo Sica, and "Conclusions" section by Cesare Imbriani.
C. Imbriani * F. Reganati
University of Rome 'Sapienza', Rome, Italy
R. Pittiglio (*)
Second University of Naples, Naples, Italy
University of Foggia, Foggia, Italy
|Printer friendly Cite/link Email Feedback|
|Author:||Imbriani, C.; Pittiglio, R.; Reganati, F.; Sica, E.|
|Publication:||International Advances in Economic Research|
|Date:||Feb 1, 2014|
|Previous Article:||Does labor market status influence self-assessed health?|
|Next Article:||The impact of emu enlargement on structural reforms: a political economy approach.|