One of the specific characteristics of Austrian Foreign Direct Investment (FDI) abroad is that a large part is carried out by firms, which are affiliates of foreign Multinational Enterprises (MNEs). This paper distinguishes two types of investors, namely "direct FDI" (made by Austrian MNEs) and indirect FDI (made by affiliates of foreign MNEs in Austria) in Central and East European Countries (CEECs). The relatively better employment performance of the former group of firms when compared to the latter seems to contradict the stylized facts, namely that direct FDI in CEECs re-export a large proportion of their output back to EU markets, indirect FDI sell a high proportion of their output locally in CEECs.
Hence the objective of this paper is to analyse, whether the relatively better domestic-employment performance of the former group compared to the latter group of investors can be linked to FDI abroad and job creation, in particular. Based on an analysis of the sales and trade structure of a representative sample of Austrian investors in CEECs, we expect that these two groups of investors have different motives to invest in CEECs and therefore the type of FDI (sales affiliate, production abroad) and consequently the effect on their domestic employment differs.
Regression results using information on the motives of firms to invest abroad confirm the hypothesis of a different behaviour of the two groups: Direct FDI are relatively stronger determined by labour costs and exhibit an employment pattern related to a deeper international division of labour (including production). Indirect FDI is based relatively more on market seeking investment and it is conceivable that foreign parents use their Austrian affiliate to tap into CEECs based on its market knowledge, while investing in production units in CEECs via their headquarter from abroad.
While the overall results are in line with other research, the results suggest that dividing the overall effect by different types of investors yields important additional information. In particular, domestic employment growth appears in another light. Results further suggest that current fears of job relocation through FDI to CEECs are largely exaggerated.
To do: market und supply based abstimmen; Struktur checken und deskriptiven Teil kürzen (insbes. Verbalisierungen von Tables vermeiden. Literatursurvey: Motive und Wirkungen genauer!
JEL Nos: F1, F23, L2, L60
Keywords: FDI, Multinational Enterprises, Employment, Industrial Policy
Contents
Introduction
Theoretical Background and Hypotheses
Data and Methodology
Results
Discussion and Conclusions
References
1wilfried.altzinger@wu-wien.ac.at
2 christian.bellak@wu-wien.ac.at
3 Earlier versions of this paper were presented at the Vienna University of
Economics, Forschungsschwerpunkt "Employment and Growth in Europe"; Draft Paper
presented at the INFER Workshop, Speyer (Germany), September 1999; and the "Erster
Österreichischer Arbeitsmarktworkshop" (Vienna).
Stylized facts on Austrian FDI in CEECs show at least three particular structural features with respect to the question of employment and trade effects of FDI which make Austria an ideal country to be studied.
First, Austrian firms have been early investors. Austrian FDI in CEECs have immediately after 1989 reached high levels. Austria’s share in FDI stocks ranges from 3 to 22 per cent in neighbouring CEECs and it is an important location (bridgehead) for regional headquarters responsible for CEECs. (cf. Table 1)
Table 1. Share of Austrian FDI in total FDI of selected CEECs
Share of FDI stocks in % |
1990 |
1997 |
Czech Rep. |
1.3 |
14.7 |
Slovakia |
- |
18.2 |
Hungary |
22.0 |
9.8 |
Poland |
5.4 |
2.4 |
Slovenia |
13.2 |
13.8 |
Croatia |
- |
11.2 |
CEECs |
10.3 |
5.8 |
Source: Stankovsky, J. (1999) Österreichische Direktinvestitionen in Osteuropa, WIFO Monatsberichte, No. 2, p. 117.
The large market share also stimulated (public) concern about job losses through relocation.
Second, Austrian investors in CEECs are only partly Austrian firms, as Austrian affiliates of foreign MNEs account for a large share of Austria’s overall FDI in CEECs.
"A great deal of Austria's FDI abroad is in fact carried out by enterprises which are in turn affiliates of foreign investors" (Neudorfer, 1997, p. 56). During the early transition period (1989-91) this type of investment had come to about 40% in CEECs. Despite this share decreased to 26% in 1996, it remains an important part of Austria's overall FDI in CEECs. (see below)
Table 2: Main Differences of Austrian Direct and Indirect FDI in CEECs, 1991 and 1996 |
|||||||
Direct FDI |
Indirect FDI |
Total FDI |
|||||
1991 |
Total Capital Stock (ATS mn) |
5314.7 |
47.1% |
5973.416 |
52.9% |
11288,12 |
100,0% |
Parent Employment |
33893 |
73.4% |
12301 |
26.6% |
46193 |
100,0% |
|
Affiliate Employment |
13437 |
54.4% |
11244 |
45.6% |
24681 |
100,0% |
|
Affiliate/Total Employment |
28.4% |
47.8% |
34,8% |
||||
Number of Investments |
311 |
75.1% |
103 |
24.9% |
414 |
100,0% |
|
Total Capital per Investment (ATS mn) |
17.1 |
58.0 |
27,3 |
||||
1996 |
Total Capital Stock (ATS mn) |
29042.75 |
73.6% |
10435.52 |
26.4% |
39478,27 |
100,0% |
Parent Employment |
92485 |
65.7% |
48301 |
34.3% |
140786 |
100,0% |
|
Affiliate Employment |
59419 |
69.6% |
26006 |
30.4% |
85425 |
100,0% |
|
Affiliate/Total Employment |
39.1% |
35.0% |
37,8% |
||||
Number of Investments |
682 |
75.0% |
227 |
25.0% |
909 |
100,0% |
|
Total Capital per Investment (ATS mn) |
42.6 |
46.0 |
43,4 |
||||
Source: Austrian National Bank; own calculations |
Third, during the early investment period (1989-95) employment growth as well as sales growth of the parent firms (cf. Table 3), direct FDI achieved better outcomes than indirect FDI (employment growth: 8.0% versus 0.0%; sales growth: 25.0% versus 17.4%). This result holds true even for the post-investment period (1995-98). A comparison of sales and employment performance of direct and indirect FDI is shown in Table 3. Sales growth of direct FDI has been considerably stronger for both periods. In accordance with sales growth figures even parent employment has grown considerable stronger for direct FDI than for indirect FDI. Hence it seems to be the case that the economic performance of direct FDI is much better than for indirect FDI.
These stylized facts give rise to a number of questions, in particular concerning (labour) cost related relocations, given the large gap of direct and indirect labour cost between "East and West" or the emergence of new markets. The reorganisation of production triggered off by the opening of Eastern Europe has brought new interest to questions of employment relocation and the impact of FDI on the balance of payments. While the public discussion in home countries is based on fears of workers in marginalised industries and of politicians in regions with high unemployment, evidence on the substitution of jobs by FDI in CEECs is still rare. A separation of the total employment effect of both groups of firms should yield important insight which helps to explain this pure descriptive pattern of domestic employment on the basis of the underlying motives and the trade structures that exist between parent and affiliate. Consequently this paper tests hypotheses concerning the links between domestic employment and job creation abroad using regression analysis.
The paper is organised as follows: First, the notion of direct and indirect FDI is explained and the existing literature on domestic employment effects of outward FDI is surveyed briefly with respect to trade and motives. Then the sample and methodology is described and the results are presented. The results point to the importance of the distinction of direct and indirect FDI, concerning the sales and trade structure as well as the structure of the motives for FDI. There is a short concluding section on the micro-economic logic of the aggregated results derived in this paper. Some policy conclusions are outlined.
Definitions
A considerable number of MNEs investing in Central and East European Countries (CEECs) channel their FDI through a affiliate (e.g. a regional headquarter) set up in a third country. Such FDI are termed indirect FDI in order to distinguish them from those FDI set up by the parent (headquarter) directly: FDI by a foreign affiliate is indirect FDI, signifying that the resulting asset-stock is owned by the parent firm via the foreign affiliate, and that it represents, therefore, an indirect flow of FDI from the parent’s home country (and a direct flow of FDI from the country in which the affiliate is located). (UN 1998, p. 145) Since FDI refers to a capital flow or stock, we use "direct FDI" synonymously for Austrian-owned firms and "indirect FDI" for foreign-owned affiliates in Austria.
Figure 1. (a) Definition of Direct vs. indirect FDI
.....
headquarter (white: Austrian firm, shaded, foreign firm)
... affiliate (white: Austrian firm, shaded: foreign firm)
Graph 1 illustrates the notion of indirect FDI. Therefore, three different geographical entities are distinguished. In principle, all three should be included in the analysis, since when analysing the relation between Austria and the CEECs, these are not independent of the activities of the MNEs in "the rest of the world". Taking the viewpoint of one country has to focus on the first, while the latter cannot be included here in a meaningful way.
The question whether outward FDI substitutes or complements domestic production and consequently jobs, has been the subject of a large number of studies. Blomström et al. (1992), Agarwal (1996) and Andersen & Hainaut (1998) provide excellent surveys. Employment effects can be derived from direct and indirect effects of outward FDI on the home country (Andersen and Hainaut 1999). One way to approach the problem is to derive substitution elasticities (relative wages) between employees in parent and affiliates. Since this road of investigation is not open to us, because of the lack of data on wage levels in CEECs by industry, we focus our analysis on indirect linkages. Trade relations between parent and affiliate are an important indirect linkage between domestic and foreign employment. The macro-economic trade effects depend on the particular value-added activities carried out in the affiliates abroad. Trade and FDI are linked in multiple ways (see Gray 1992). According to this strand of argument, the intra-firm trade balance, the geographical distribution of exports, the importance of local sales are important determinants of the demand for labour at home resulting from outward FDI. Most empirical studies, however, lack information about these important indicators. Blomström et al. (1997) argue, that rivalry for markets is one of the main reasons for a positive relationship (p. 1798), therefore we focus on the firms’ motives to invest in CEECs. Consequently, the argument here is based on the relationship between the motives for FDI and their effect on trade (cf. Agarwal 1996). If the international division of labour differs between direct and indirect FDI, such differences can be found by looking at the indicators mentioned. The reasons for differences in motives are first the specific competence of the Austrian affiliates of foreign MNEs and second the experience of Austrian firms in CEEC markets (Bellak 1997).
The implications of the motives for trade differ for direct and indirect FDI. This is mainly due to the different linkages between parent firms and affiliates of these two types of FDI. Strategies are elaborated and accomplished by the parent firm which in the case of indirect FDI happens to be located outside Austria.
According to established theory, the activities of affiliates can be related to the motives of FDI, namely efficiency seeking, market seeking and strategic asset seeking. The impact of these types of FDI on trade patterns are explained by distinguishing four kinds of trade linkages between the parent firm and her affiliates: the substitution of former exports through FDI, growing (re-)imports of goods and services produced abroad, FDI associated exports of goods and services and FDI induced exports of other product lines neither produced by the foreign affiliate nor exported earlier by the parent firm (Agarwal, 1996; Altzinger 1998). The overall impact of FDI on trade (and consequently on domestic employment) is the sum of negative (export substitution, re-imports) and positive effects (associated and induced exports) and can be tested only empirically.
.....
headquarter (white: Austrian firm, shaded, foreign firm)
... affiliate (white: Austrian firm, shaded: foreign firm)
We examine, how FDI in CEECs are linked to the domestic employment performance by analysis of the sales / trade structure and the motives for FDI. We distinguish between two types of investors.
To analyse the impact of foreign affiliates on domestic employment we make use of two different sets of data: The first comprises the results of the annual FDI survey, covering all firms with an investment above ATS 1 million, carried out by the Austrian National Bank. This data set provides excellent information about the total amount and structure (geographical, industrial) of investment as well as on employment. However, these data do not provide sufficient information on trade figures. Therefore we use a second data set which presents results of a survey conducted in summer 1997. Among others, this data set includes sales, export and employment data for 1989, 1995 and expected figures for 1998.
A simple model which analyses the different impact of parent and affiliates sales on domestic employment (Blomström, et al., 1997) is used.
Firstly we present the main differences and developments of Austrian direct and indirect FDI between 1991 and 1996 using the first data set in a simple descriptive way. Secondly, after comparing the structure of our survey data to the structure of the first data set, we present the main information on sales, exports and employment patterns of the affiliates and parents by distinguishing between direct and indirect FDI. Finally we test our above-formulated hypotheses with these data.
Development and Structure of Austrian Direct and Indirect FDI in CEECs
Table 2 (cf. above) shows the main differences of Austrian direct and indirect FDI in CEECs for 1991 and 1996. This period is characterised by the opening up of the East European economies and their initial economic integration with the European Union. As can be seen the amount of capital invested has increased fourfold and was ATS bn 39.5 in 1996. Starting from zero in 1989 this amount accounted for nearly 30% of Austrian total outward FDI stock in 1996. Hence the importance of the CEECs for the internationalisation of Austrian enterprises is clearly discernible. Even the number of investments has increased from 414 to 909. However, this increase is less pronounced than the growth of capital invested. This is mainly the result of different investment patterns during the early 1990s and thereafter (see analysis below). At the beginning of the economic integration between Austria and her adjacent Eastern European neighbours even many small and medium-sized enterprises (SME) used their first-mover advantages which can be attributed to long lasting historical and cultural ties (Altzinger 1998). However, due to weak financial capabilities of these firms capital per investment has been rather small. However, after this initial period the integration process has proceeded rather quick and the economic stabilisation of the transition countries has improved. These changes have attracted even some very large investments of Austrian firms. Therefore the amount of capital invested has increased from ATS 27 million in 1991 to ATS 43 million in 1996 and the sharp difference between direct and indirect FDI has vanished. Since 1995 the amount of capital per investment do not differ any longer between direct and indirect FDI. Another reason may be the cumulative learning processes of early movers (Porter 1990), i.e. small Austrian firms which expand their operations based on the positive experience after their early entry. This may also reflect the competitive advantaged vis-à-vis latecomers.
Employment figures increased in accordance with total capital invested. The large increase is evident for both parent and affiliate employment. Parent employment growth can be explained mainly by the strong increase of number of parent firms investing in CEECs. Therefore the share of affiliate employment in total (affiliate and parent) employment remained relatively stable throughout the period 1991-1996. However, we can see strong differences between direct and indirect FDI. Whilst the former group of firms have increased their share of affiliate employment rather strongly the latter group has not. Indeed their share of affiliate employment decreased considerably.
Table 4 provides evidence about the industry structure of direct and indirect FDI.(4). The sectoral distribution of total FDI shows that nearly 60% is allocated to the non-manufacturing sector and only 40% to manufacturing. Within the service sector the largest sectors are finance and insurance (20.5%), trade (18.1%), real estate and business activities (7.9%), which includes holding companies and construction (6.0%). However, there are considerable differences between direct and indirect FDI. It is in particular the trading sector where indirect FDI accounts for a very large share. Two other important differences between direct and indirect FDI exist in finance and insurance and real estate and business services. In both of these sectors direct FDI shows considerably larger shares. Two very large banks (RZB and Creditanstalt/Bank Austria) who are 'market leaders' in CEECs have achieved the dominance of Austrian owned firms in finance and insurance. The difference in business services can be explained only by holding companies in this sector. (5).Within the manufacturing sector there are three sectors which show strong differences between direct and indirect FDI. Whilst food and beverages are dominated by Austrian-owned firms, chemicals and petroleum and electrical equipment is strongly dominated by indirect FDI.
(4). The sectoral distribution of Austrian
investment in CEECs in Table 2 is displayed by the classification of the host
country.
(5). Indirect FDI will in fact only invest in holding companies by their parent
firm which is located somewhere outside Austria and therefore is not be displayed
in Table 2.
Table 4: Structure of Austrian Direct and Indirect FDI in CEECs, 1996
Direct FDI |
Indirect FDI |
Total FDI |
||||
ATS mn |
% |
ATS mn |
% |
ATS mn |
% |
|
mining |
504 |
1.7% |
34 |
0.3% |
538 |
1.4% |
food & beverages |
2434 |
8.4% |
156 |
1.5% |
2590 |
6.6% |
textiles |
262 |
0.9% |
262 |
0.7% |
||
wood products |
442 |
1.5% |
13 |
0.1% |
455 |
1.2% |
paper & publishing |
1305 |
4.5% |
197 |
1.9% |
1502 |
3.8% |
chemicals & petroleum |
560 |
1.9% |
2470 |
23.7% |
3031 |
7.7% |
non-metallic products |
2241 |
7.7% |
1139 |
10.9% |
3380 |
8.6% |
metal |
929 |
3.2% |
114 |
1.1% |
1043 |
2.6% |
machinery |
293 |
1.0% |
239 |
2.3% |
532 |
1.3% |
electrical equipment |
1120 |
3.9% |
1101 |
10.6% |
2221 |
5.6% |
transport equipment |
38 |
0.1% |
122 |
1.2% |
160 |
0.4% |
other manufacturing |
245 |
0.8% |
40 |
0.4% |
285 |
0.7% |
Engineering |
2380 |
8.2% |
1576 |
15.1% |
3956 |
10.0% |
Manufacturing |
9868 |
34.0% |
5591 |
53.6% |
15459 |
39.2% |
construction |
1804 |
6.2% |
578 |
5.5% |
2382 |
6.0% |
trade |
4026 |
13.9% |
3117 |
29.9% |
7143 |
18.1% |
hotels & restaurants |
1956 |
6.7% |
22 |
0.2% |
1978 |
5.0% |
transport & communication |
309 |
1.1% |
309 |
0.8% |
||
finance & insurance |
7071 |
24.3% |
1006 |
9.6% |
8077 |
20.5% |
business services |
3097 |
10.7% |
31 |
0.3% |
3128 |
7.9% |
other services |
407 |
1.4% |
57 |
0.5% |
464 |
1.2% |
Non-Manufacturing |
18670 |
64.3% |
4811 |
46.1% |
23482 |
59.5% |
Total |
29043 |
100.0% |
10436 |
100.0% |
39478 |
100.0% |
Source: Austrian National Bank; own calculations |
Finally a few interesting patterns of intra-firm trade (IFT) flows emerge (see Table 5). First, both direct and indirect FDI display large intra-firm trade (IFT) surpluses seen from the Austrian perspective. Total IFT surplus increased from ATS 1.2 billion in 1991 to ATS 5.5 billion in 1996. This surplus is primarily due to indirect FDI. Although indirect FDI accounts only for 26.4% of total capital in 1996 (see Table 2) these firms have achieved 51.3% of total intra-firm exports and 35.6% of total intra-firm imports. Hence these firms realised nearly two thirds of IFT surplus total. As can be seen in Table 5 there are only two sectors which display (small) IFT deficits (food and beverages, other manufacturing) again viewed from the home country perspective. All other sectors achieved IFT surpluses. Not surprisingly, the trading sector accounts for the largest share of IF surplus. In particular this investment is undertaken with the purpose of expanding sales and improving the distribution of export goods. However, this surplus has been achieved to a very large extent by indirect FDI. If these exports are accompanied by similar imports to the Austrian firm (which is in fact an affiliate of a foreign MNE) it can be possible that this surplus did not help to improve the Austrian trade balance.
More interestingly, IFT is particularly pronounced in the engineering sector. This sector accounts for one third of total IF exports and 50% of total IF imports. This fact is true for direct as well as indirect FDI which indicates that the international division of labour is well developed within this sector. In contrast to the IFT patterns of the trading sector, this pattern can be associated with a vertical production structure where intermediate goods are shipped forwards and backwards. Surprisingly, even this sector achieved considerable IFt surpluses. It should be emphasised that the industry classification is that of the host country, therefore IFT does not include a large proportion of finished goods as they would be classified under the industry "trading sector".
Table 5: Intra-Firm (IF) Trade, 1996 (in ATS mn)
Main characteristics of the survey data
The following analysis
covers 150 firms which have engaged in at least one investment in CEECs. 112
of these firms (74.7%) were Austrian-owned and 38 non-Austrian owned, i.e. affiliates
of foreign MNEs (25.3%).(6).
(6). Since not each firm replied to all questions the number of respondents in all the following tables is always below the total of 150.
Throughout the further analysis we distinguish between an ‘initial investment’ period (1989-95) and a ‘post-investment’ period (1995-98). It is indeed the ‘initial investment’ period when almost all Austrian investors have achieved their first investment. Most of the proceeding investments have been an enlargement of the first investment or an additional investment. As mentioned above, during this post-investment period the stabilisation of the transitional economies has proceeded and some very large investments have been carried out. As explained below we expect distinctive changes on the FDI-exports relationship over these periods. We proceed in two stages, first we assess parent firms and second affiliates.
P a r e n t F i r m s
The following tables show the differences of regional sales structure for these two groups of firms. We analyse the regional structure of turnover of parent companies and affiliates. Following the considerations above we would expect that efficiency-based FDI show rather low exports to CEECs by Austrian parent companies and large export shares to the EU and Austria respectively by the affiliates. On the contrary, we would assume that market-driven affiliates stimulate exports of the parent companies while imports from the affiliates should be only of minor importance.
Table 6 presents the regional sales structure of the parent firms. As can be seen only 56.3% of total output have been sold at Austrian markets. The remaining 43.7% were sold at foreign markets. Thereof the largest share (24.5%) was shipped to the EU and another 12.8% to markets in CEECs. A comparison by direct and indirect FDI reveals that non-Austrian owned firms display larger export activities than Austrian ones. Their export shares are 49.3% and 42.0% respectively. Indirect FDI exports more to the EU and CEECs alike. One explanation for this observation may be the above-mentioned high share of indirect FDI in the trading sector. It is in particular this sector which displays high export activities with CEECs.
The impact of these investments on export activities of parent firms can be shown by calculating the changes of the regional sales structure. Diminishing market shares in CEECs would point to a substitutive relationship between investments and exports and vice versa. As illustrated for aggregated trade by Helpman (1984) and for intra-firm trade by Bergsten, et al. (1978) it might be the case that in the initial stage of investment the vertical integration aspect may dominate and therefore investment will boost exports. In later stages horizontal investment may be more important, leading to a substitutive relationship in subsequent periods. According to these models we would expect rising market shares of CEECs during the initial period of investment and stable or declining market shares in the post-investment period.
As can be seen in Table 7 the initial investment has been accompanied by a considerable shift of the sales structure. On average the parent companies have improved their total export shares by 6.3 percentage-points. The largest increase has been performed on markets in CEECs (+4.0 percentage-points) but even the EU-market shares have grown (+1.8 percentage-points). Hence the internationalisation of these firms was not restricted to CEE markets only. Although the sales shares in CEECs improved stronger than in EU markets, the internationalisation of the firms included in our sample was simultaneous and two-sided. Hence these results show typical patterns of globally acting firms. Moreover, the expected complementary relationship between investments and exports seems to be confirmed. Differences between direct and indirect FDI are not very strong.(7).
(7). Within the period 1989-95 80% of the firms exhibit growing turnovers whereas 20% of the sample had declining turnovers. However, a separate analysis by these two groups does not provide severe differences.
Table 8 shows the expected changes of the parent firm's sales structure for the post-investment period 1995-98. Interestingly, the general pattern of the initial period of investment seems to be extended. On average it is expected that the internationalisation process will continue into both directions, Eastern and Western Europe alike. However, for this period there are strong differences between direct and indirect FDI. Direct FDI show much stronger patterns of internationalisation than indirect FDI. The complementary relationship between investments and exports to the CEECs holds for direct FDI only. This might be an indication that direct FDI still expands whilst indirect do not.
A f f i l i a t e s
Next we analyse the sales structure of the affiliates (Table 9). For that purposes it makes sense to distinguish between two different motives of investment: ‘efficiency-based’ and ‘market-driven’ FDI. Presumably the first one would indicate that the dominant factor of investment is to get access to a cheap industrial workforce. Such a scenario would presumably substitute exports from the home country and encourage reimports to the home country. These investments are usually associated with ‘relocation’ and display high export shares to EU markets. In contrast to this approach is the market-driven approach. Such a strategy is enforced by a considerable expansion of local demand or diversification of the investor. Therefore the production of the affiliates should be sold to a large extent on local markets.
Table 9 shows the regional sales structure of the affiliates for 1995. On average the local markets account for 65.5% of total sales. Hence the predominance of local markets is obvious. However, even 23.2% were shipped to EU markets, thereof 10.6% to Austria. Furthermore, some important differences between direct and indirect FDI can be detected. Trade relations between direct FDI and their Austrian parent firms are more intense than those of indirect FDI are. Whilst the former group exports 12.3% of their sales to Austria this share is only 4.2% for indirect FDI. Interestingly, the local market share of indirect FDI is quite larger than for direct FDI, which is an indication that the division of labour is more advanced by direct FDI.(8).
(8). Expected changes of affiliates sales structure for the period 1995-98 are not very strong. Therefore these data are not presented here.
Are motives of FDI of any significance in explaining the trade structure of parent and affiliate firms?
Previous analysis has demonstrated rather clearly the predominance of market-driven motives for Austrian FDI in CEECs (Altzinger and Winklhofer 1998, Neudorfer 1997). Supply-based motives, low wage costs in particular are only of minor importance.(9). However, it has been shown that the ranking of supply-based and market-driven motives is rather different for particular industries. In the analysis that follows we want to evaluate the impact of these different motives on the export performance of parent and affiliate firms alike. We expect that market-driven motives improve exports to CEECs by the parent firms whilst the affiliates sell most of their output at local (foreign) markets and export only small amounts to the EU. In contrast, if supply-based motives dominate, we expect strong exports to the EU by the affiliates and only small exports to CEECs by parent firms. For two particular industries (engineering and trade) we have also included dummies in the regressions. The first dummy has been chosen since engineering is one sector where efficiency-based FDI dominates whilst in trade it is foremost market-driven FDI.
Table 9: OLS Regression Results for Parent Exports to CEECs, 1995
Constant |
Market |
low |
Strategic |
Dummy for |
Dummy for |
R² |
|
Potential |
wage costs |
Motives |
Engineering |
Trade |
|||
Total FDI |
9.0 |
9.1* |
-10.3** |
7.4* |
-1.6 |
13.1 |
0.22 |
(t-values) |
0.37 |
1.91 |
-2.59 |
1.85 |
-0.16 |
1.49 |
|
Direct FDI |
19.5 |
6.5 |
-12.7*** |
8.6** |
1.6 |
21.0** |
0.27 |
(t-values) |
0.75 |
1.28 |
-2.81 |
2.02 |
0.15 |
2.05 |
|
Indirect FDI |
-16.2 |
25.1 |
-4.2 |
-4.1 |
-23.4 |
-17.0 |
0.28 |
(t-values) |
-0.23 |
1.71 |
-0.44 |
-0.34 |
-0.89 |
-0.73 |
***, **, * indicate significance at the 1, 5 and 10% significant level;
(9).Market-driven motives are among others market potential and proximity to customers whilst supply-based motives are low wage costs, availability of skilled labour, intermediate inputs and procurement.
Results of the regressions for parent exports to the CEECs are shown in Table 10. (10). It can be seen that the ranking of the motive market potential indeed increases exports to CEECs of the parent firms whilst low wage costs decreases exports considerably. The value of the coefficient shows that a one-point higher ranking of the motive market potential improves the export performance of the parent firm by 9.1%-points. Furthermore, even strategic considerations are of importance for the export performance of the parent firm. Interestingly, both coefficients for the industry dummies are insignificant.
However, there are some differences for direct and indirect FDI. Low wage costs and strategic motives can only explain the parent export performance for direct FDI. For direct FDI even the dummy for trade is significant. The results for indirect FDI are rather disappointing. The only variable, which is close to the 10% significant level, is market potential.
Table 10: OLS Regression Results for Affiliates Exports to EU, 1995
Constant |
Market |
low |
Strategic |
Dummy for |
Dummy for |
R² |
|
Potential |
wage costs |
Motives |
Engineering |
Trade |
|||
Total FDI |
98.0*** |
-21.4*** |
4.6 |
-5.7* |
17.4** |
-2.4 |
0.49 |
(t-values) |
5.54 |
-6.11 |
1.58 |
-1.96 |
2.36 |
-0.38 |
|
Direct FDI |
89.3*** |
-19.4*** |
6.6* |
-7.2** |
20.4** |
-2.2 |
0.52 |
(t-values) |
4.47 |
-4.97 |
1.88 |
-2.21 |
2.37 |
-0.28 |
|
Indirect FDI |
118.8** |
-29.0** |
-3.8 |
4.2 |
6.8 |
-1.4 |
0.52 |
(t-values) |
2.43 |
-2.87 |
-0.57 |
0.50 |
0.38 |
-0.09 |
***, **, * indicate significance at the 1, 5 and 10% significant level;
(10).Questions on motives were of a close-ended variety where the degrees of importance were based on a four-point scale using irrelevant (1), of minor relevance (2), important (3), and very important (4). Results are shown only for selected motives.
Results for the explanation of affiliate exports to the EU are more encouraging (see Table 11). Exports to the EU can be quite well predicted. Again, market potential explains most. A one-point higher ranking of this motive reduces affiliate exports to the EU by 21.4%-points. Even strategic motives decreases exports to the EU. Interestingly, the coefficient for wage costs is insignificant. However, the coefficient of the dummy for engineering is positive. It is in particular this sector where most of total sales are sold at foreign (mainly EU) markets.
Separated by direct and indirect FDI we can see that the results differ rather strongly. The general explanation above can be confirmed for direct FDI only. For direct FDI even wage costs are of importance. The expected positive impact on exports becomes clear. This rather diverse explanation of exports for direct FDI contrasts strongly with the pattern of indirect FDI. For this group of firms only market potential can explain the export performance. Neither any other motive nor industry dummy can help to improve the explanation of affiliate exports.
To summarise these different explanations of parent and affiliate exports we conclude that the export performance of indirect FDI can be explained by market-driven motives only. In contrast, for exports of direct FDI even wage costs, strategic motives and industry dummies provide additional explanations. However, even for direct FDI market potential is the dominant explanatory variable.
Impact of parent and affiliate sales on domestic employment
Finally we look at the impact of affiliate sales on domestic employment. For that purpose we have chosen a simple descriptive model (see Blomström et al., 1997). Such an approach tries to evaluate the relationship between foreign sales and domestic employment for a given level of parent sales.
Table 11: OLS Regression Results for Parent Employment, 1995
constant |
Parent sales |
Affiliate |
Dummy for |
R² |
|
Sales |
engineering |
||||
Total FDI |
269.9** |
0.3*** |
-0.7*** |
502.2** |
0.76 |
(t-values) |
2.48 |
10.75 |
-2.88 |
2.43 |
|
Direct FDI |
248.2* |
0.3*** |
-0.9*** |
499.2* |
0.77 |
(t-values) |
1.86 |
9.86 |
-2.95 |
2.00 |
|
Indirect FDI |
-34.7 |
0.6*** |
-0.2 |
84.4 |
0.90 |
(t-values) |
-0.43 |
7.55 |
-0.96 |
0.58 |
|
***, **, * indicate significance at the 1, 5 and 10% significant level; |
The estimation results for total FDI suggests that higher affiliate sales reduce parent employment (cf. Table 12). However, this holds only for a given level of parent sales. The size of the coefficients suggests that an increase of parent sales by ATS 1 million induce 0.3 additional parent employees. In contrast, an additional ATS million of affiliate output reduces parent employment by 0.7 employees, The size of these two coefficients is approximately in line with the productivity differences between parent output in Austria and affiliate output in CEECs. Due to the crudeness of this relationship the interpretation of this result has to be rather cautious. However, it seems to be the case that larger foreign affiliate production is associated with an allocation of labour-intensive value added stages of production to foreign countries whilst capital-intensive production is performed at home. Although such a scenario is associated with a relocation of labour, such FDI are efficiency enhancing and therefore improve the overall competitiveness of the MNE. The dummy for engineering shows that those parent firms are much larger than the average firm.
However, an individual evaluation of this relationship for direct and indirect FDI shows again considerable differences. In particular the negative impact of affiliate production on parent employment holds only for direct FDI. Parent employment of indirect FDI is not significantly influenced by affiliates output. The only variable that matters is parent sales. Moreover, parent sales of indirect FDI are more labour-intensive than parent sales of direct FDI.
The results of this analysis indicate that direct FDI is more of an efficiency-type than indirect FDI. However, this result does not indicate that the absolute impact on domestic employment is negative. It only suggests that direct FDI has a much stronger direct linkage between the activities of the parent firm and its affiliates in CEECs. Indeed, we are able to show empirically that firms of the direct FDI-type have improved domestic and foreign employment simultaneously by improving their internal division of labour. This is mainly due to the fact that Austrian investors in CEECs include many small and medium-sized firms, which actively started to restructure their activities via FDI in adjacent countries after 1989. Geographical proximity is a driver of FDI (Martin and Velazquez 1997), reflected here in the relatively deeper integration of Austrian firms when compared to other MNEs. Proximity is expressed in such terms as easier market entry, prior information, lower transport costs. Moreover, indirect FDI indicates higher export shares to CEEC (CEE-exports/sales) by the parent firms than direct FDI and even the local market shares by the affiliates are higher than for direct FDI. In addition, (re-)exports to Austria from direct FDI are higher than from indirect FDI.
Our regression analysis underpins this pattern by distinguishing between direct and indirect FDI. We explain exports to CEECs by the parent firm as well as exports to the EU by their affiliates by different motives for FDI, namely market-driven, strategic and supply based. The results reveal that direct investment can be much better explained by these motives (i.e. they are statistically significant) than indirect FDI (where motives, except the market motive, lack statistical significance). We explain this difference by the different linkages between parent firms and affiliates of direct and indirect FDI.
In Austria, most affiliates of foreign MNCs were based on market seeking motives (see Table 13).
Table 13. Motives of foreign investors in Austria
Selected home countries |
Labour-cost |
Market |
Sourcing |
Tax |
Germany |
2.4* |
69.0 |
2.1 |
1.5 |
Switzerland |
1.8 |
54.6 |
2.1 |
4.8 |
Netherlands |
0.0 |
73.4 |
0.0 |
1.6 |
USA |
0.5 |
69.9 |
0.5 |
3.2 |
Total |
1.5 |
66.3 |
2.3 |
2.6 |
EU |
1.7 |
69.6 |
2.1 |
1.7 |
* percentage of affiliates
Source: Austrian National Bank, Statistisches Monatsheft, No. 6, 1998, Table 10.2.
These results are also in line with other research (e.g. Taggert and Hood, 1999) on the role of decisions taken by Headquarters (or in other words, the extent of affiliate autonomy): The HQ’s influence is usually dominant "where central resources are directly affected or drawn upon; where long-term obligations result; and where the decisions involve standardisation and a common framework of organisational routines and practices." Most of these factors apply to marketing and sales activities, which points to a low autonomy of the Austrian affiliate. This in turn might imply, that production decisions concerning CEECs are taken by the parent firm directly. The propensity of Headquarters (HQs) to use their affiliates for the collective benefit of the enterprise as a whole increases with the unique "skills" of the unit in question. (11).(see e.g. Taggert and Hood, 1999, p. 228)
The underlying structure of the motives indicates that the parents use their Austrian affiliate in particular for sales-related activities not only in Austria, but also in CEECs, utilizing their specific market know-how. This is consistent with Porter’s explanation (1990, p. 47), that "early movers gain advantages such as being first to reap economies of scale, reducing costs through cumulative learning, establishing brand names abroad and customer relationships without direct competition, getting their pick of distribution channels, and obtaining the best locations for facilities or the best sources of raw material or other inputs." The changes in CEECs after 1989 created unique opportunities for foreign parent firms to enter CEEC markets through their Austrian affiliate.
In CEECs, Austrian MNCs intensified their international division of labour over time and moved from sales to production. The foreign MNCs located in Austria did not follow this sequence. This does, however, not imply that both groups’ activities in CEECs will differ, as the latter group organises their strategic and efficiency-oriented FDI not via their Austrian affiliate. This possibility is shown in graph 1b, but is excluded here as mentioned above.
(11).It is important to emphasise, that we do not argue for a difference in activities between Austrian firms and foreign MNEs, but rather, that they organise their activities in CEECs differently.
Discussion and Conclusions
The paper started from a few stylised facts on Austrian FDI in CEECs. At first glance, the descriptive results presented here seemed to contradict the stylised facts as presented in the introductory section about the relatively better employment performance of Austrian-owned firms. Further investigation led to the conclusion that not only the sales and trade structure, but also the underlying motives of FDI provide a consistent picture of direct and indirect FDI. We have found that FDI in CEECs by direct and indirect investors are motivated by different factors. It is important to emphasize, that both types of FDI have positive repercussions on the Austrian part of both groups of investors, yet the mechanism differs: The fact that a large wage gap still exists, points to the possibility that direct FDI reflects a re-allocation of low-skill activities to CEECs, while indirect FDI stimulates further exports of the Austrian firm. These results portrait "an average" investor, as of course, both groups of firms engage in more than one type of FDI, based on more than one exclusive motive. In particular, Austrian-owned firms, by relocating single (labour-intensive) value-added stages to CEECs, create sustainable competitive advantage at home, resulting in an overall (net) increase in domestic employment. Overall, the results of this analysis confirm that direct FDI is more of an efficiency-type than indirect FDI. However, this result does not indicate a negative impact on domestic employment, but rather that direct FDI has a much stronger direct linkage between the activities of the parent firm and its affiliates in CEECs. Indeed, we are able to show empirically that firms of the direct FDI-type have improved domestic and foreign employment simultaneously by improving their internal division of labour. In what follows the results are presented in greater detail.
The results show that the underlying ownership structure of foreign investors is an important determinant of employment effects of outward FDI on home countries. The empirical evidence suggests that direct and indirect FDI use different strategies for the expansion of their activities in CEECs. In particular, Austrian firms (direct FDI) are operating along an efficiency-enhancing strategy with many direct linkages between the parent firm and its affiliates. By contrast, foreign owned firms (indirect investment), use their Austrian affiliate as a bridgehead and engage primarily in market-oriented FDI in CEECs. Although both types of FDI exert a positive effect on Austrian employment, the effect of the former is relatively smaller compared to the latter, which is due to the deeper integration of direct FDI into the international division of labour with CEECs. These results are also in line with Blomström et al. (1997) who derive a similar pattern with respect to two different host countries.
The results derived here on an aggregated level do also bear some micro-economic logic as argued in studies about the division of labour between parent and affiliates, and the autonomy of affiliates in particular. Such studies take the value-added chain as a starting point. The number, importance and quality of the activity of an affiliate within the internal division of labour of the MNC determines the degree of autonomy and inter alia determines whether the affiliate itself engages in affiliates in foreign countries (indirect FDI). The affiliates’ activities range from simple assembly operations to World Mandate status. Young, Hood and Peters (1994) describe "developmental affiliates" as firms which are able to combine their competitive advantage (provided by the parent) and indigenous resources (comparative advantages) in a unique way. The market know-how of Eastern Europe is such a factor which in combination with the product or service results in a unique position of the Austrian affiliate within the whole group of the MNE.
From the viewpoint of a national economy, the basic difference between direct and indirect FDI is related to the aspect of the locus of decision making (Bartlett and Ghoshal 1989; Birkinshaw and Morrison 1995; Dunning 1990). Sometimes, a "metropolitan – hinterland" relationship is suggested, where the headquarter is responsible for the strategy, the affiliate for operational sales targets. The trade structure of the Austrian headquarter and her FDI compared to the foreign affiliate in Austria and her FDI resembles this type of relationship. In particular, the Austrian affiliate can be used by the foreign parent for transit trade to her affiliate in the CEEC (indirect FDI). Where substantial business decisions are concerned, MNEs seem to manage their foreign affiliates on a very short leash. Thus, the impact of FDI on employment in the home country depends on the division of labour between domestic headquarter and its affiliates abroad and therefore is idiosyncratic.
The evidence about a change in the effect over time presented here is also consistent with evidence published by the Austrian National bank (see Table 2 above). A possible explanation could be that the specific experience of the affiliate located in Austria has been lost towards the affiliate in CEECs and that foreign HQs have already built up their own CEEC-competence. Moreover "Austrian FDI in the CEECs has become controlled more and more by Austrian firms." (Neudorfer 1997, p. 57) Indirect FDI nevertheless remain an important determinant of the domestic employment effect of Austrian FDI in CEECs. Also, our survey included a question on the future strategies of investors in CEECs and although the plans of firms need not materialise in praxi, the answers point into the same direction. Seven percent of Austrian investors intend to engage in "production" in the future, while the ratio is only 5 percent for indirect investors. On the other hand, an expansion of "distribution and sevices" is intended by 7.4 percent of indirect FDI, but only 2.4 percent of direct FDI.
To summarize, the paper provides evidence, that the composition of the overall employment effect is important and a division between different types of direct investors sheds new light on the question of production relocation. In particular, the specific argument used here, is that (a change of ) the employment effect depends not only on (a change of) the amount of FDI, but also on (a change of) the structure of investors.
The plausibility of the aggregated results on the micro-economic level show that a distinction between the two groups of foreign investors may give rise to reconsider orthodox arguments concerning the restriction of outward FDI as frequently raised by governments and trade-unions. First, the positive effect of relocating parts of the value-chain of firms and its positive effects on domestic employment and the overall competitiveness of the remaining domestic value-added stages should be actively supported (provision of information on host countries, joint production etc.). Second, fears of welfare loss of the developed region when integrating with a less developed region seem to be unjustified, concerning trade and FDI relationships. On the contrary, the evidence points to a positive welfare effect.
Agarwal, J. (1996) Does Foreign Direct Investment Contribute to Unemployment in Home Countries?, Kiel Working Papers, No. 765.
Altzinger, W. (1998) Austria’s Foreign Direct Investment in Central and Eastern Europe: ‚Supply Based‘ or ‚Market Driven‘? Working Paper No. 57, Vienna University of Economics and B.A, Department of Economics.
Altzinger, W., Beer, E. and Bellak, C. (1998a) Fallstudien zu den Auswirkungen der Ostöffnung auf Beschäftigung und Zahlungsbilanz Österreichs, Research Report of a Project of "Jubiläumsfonds der OeNB", No. 6700.
Altzinger, W., Beer, E. and Bellak, C. (1998b) Arbeitsplatzverlagerung nach Osteuropa, Wirtschaft und Gesellschaft, Vol. 24, No. 4, pp. 475-502.
Andersen, P.S. and Hainaut, P. (1998) Foreign Direct Investment and Employment in the Industrial Countries, BIS Working Paper, No. 61, November.
Data provided by the Austrian National Bank
Barell, R. and Pain, N. (1997) The Growth of Foreign Direct Investment in Europe, in: National Institute of Economic Review, No. 160, April, pp. 63-75.
Barell, R. and Pain, N. (1997) Foreign Direct Investment, Technological Change and Economic Growth Within Europe, The Economic Journal, Vol. 107 (November), pp. 1770-1786.
Bartlett, C. and Ghoshal, S. (1989) Managing Across Borders: The Transnational Solution, Boston, Mass: Harvard Business School Press.
Bellak, C. (1997) Austrian Manufacturing MNEs: Long Term Perspectives, Journal of Business History, Vol. 39, No. 1, pp. 47-71.
Bellak, C. (1998a) The Measurement of Foreign Direct Investment - A Critical Review, The International Trade Journal, Vol. XII, No. 2, 1998, pp. 227-257.
Bellak, C. (1998b) Lessons from Austria’s Postwar Pattern of Inward FDI for CEECs, Research Report, No. 251, The Vienna Institute of Comparative Economic Studies, Vienna, 1998.
Bergsten et al. (1978) ??
Birkinshaw, J.M. and Morrison, A.J. (1995) Configurations of Strategy and Structure in Subsidiaries of Multinational Corporations, Journal of International Business Studies, Vol. 26, No. 4, Fourth Quarter, pp. 729-753.
Blomström, M., Fors, G. and Lipsey, R.E. (1997) Foreign Direct Investment and Employment: Home Country Experience in the United States and Sweden, The Economic Journal, 107 (November), pp. 1787-1797.
Buckley, P.J. and J.-L. Mucchielli (eds, 1997) Multinational Firms and International Relocation, Edward Elgar: Cheltenham.
Dunning, J.H. (1990) Multinational Enterprises and the Global Economy, Addison-Wesley: Wokingham.
Gray, H.P. (1992) The Interface between the Theories of International Trade and International Production, in Buckley, P.J. and Casson, M.C. (eds) Multinational Enterprise and International Direct Investment: Essays in Honour of John H. Dunning, Edward Elgar Publishers, Cheltenham, pp. 41-53.
Helpman (1984) ??
Iammamino, S. and Pitelis, C. (1999) Foreign Direct Investment and "Less Favoured" Regions: Greek FDI in Bulgaria and Romania, mimeo. ??
Markusen J. R. (1995) The Boundaries of Multinational Enterprises and the Theory of International Trade, Journal of Economic Perspectives, Vol. 9, No. 2, Spring, pp. 169-189
Martin, C. and Velazquez, F. (1997) The Determining Factors of Foreign Direct Investment in Spain and the rest of the OECD: Lessons for the CEECs, CEPR Discussion Paper, No. 1637.
Neudorfer, P. (1997) The Opening of Central and Eastern Europe, Focus on Transition, No. 2, Austrian National Bank, pp. 52-68.
Porter, M. (1990) The Competitive Advantage of Nations, The Free Press: New York.
Sleuwaegen, L. and Pennings, E. (1998) International Relocation: a Firm-Perspective Empirical Study, Paper presented at EIBA, Jerusalem, Israel, mimeo.
Taggert, J. and Hood, N. (1999) Determinants of Autonomy in Multinational Corporations Subsidiaries, European Management Journal, Vol. 17, No. 2, pp. 226-236.
Young, S., Hood, N. and Peters, E. (1994) Multinational Enterprises and Regional Economic Development, Regional Studies, Vol. 28, No. 7, pp. 657-77.
UN (1998, ed.) World Investment Report 1998: Trends and Determinants, United Nations: New York and Geneva. (esp. Ch. V).