Labour

CEUeTDCollection

46 knowledge-intensive sectors. The specific policy suggestions are less straightforward: usually the government is expected to invest in everything from education to provision of start-up capital, besides the funds already committed to attract the multinationals (e.g. Lall & Narula 2004; see also O’Riain 2000). In addition to being extraordinarily expensive, such advice also once again reverses the causality between foreign investment and development, and is not very realistic in developing country settings. With the governments reluctant to undertake industrial investment themselves, and the local firms too weak to participate in technology development, there is little basis on which to create the kind of infrastructure that would inspire interest by global multinationals. More often than not, the only bait they can count on to attract knowledge-intensive investments, if they can raise the quality of offer, is the one resource they have aplenty: the labour.

CEUeTDCollection

47 According to Amsden (1990a), in Korea and Taiwan, as in Japan before them, wages actually grew quite quickly, but never outpaced the productivity gains. All of these countries frequently resorted to currency devaluations to improve external competitiveness, but these also increased the costs of imported inputs and capital equipment. Therefore, to ensure profitability of industrial enterprises, the governments also tried to address the problem of labour costs directly, by suppressing all activities which could interfere with productivity growth. Among other, these included outlawing independent trade unions, suppressing strikes and allowing wage growth to be balanced by exceptionally long working hours.

However, wage control was largely a secondary strategy. The main trust of labour policy in East Asia was to focus heavily on investments in education, especially technical education, which included pressure on the private sector to do the same (Evans 1995; World Bank 1993).

Here growing wages doubled as a kind of “productivity whip”: they forced the firms to increase productivity faster, among other through investments in workforce skills, in order to catch up with the wage growth. Private investment was further encouraged by paternalist labour relations which demanded workforce commitment and wage moderation in exchange for the promise of stable employment (Amsden 1990b; 1989; Dore 1972). At the same time, in industries where productivity gains were slower or where labour costs remained the key source of competitiveness, the costs were kept low through reliance on the more vulnerable segments of the workforce, typically women (Berik 2000; 2006; Amsden 1989).

We find a very similar range of tools in Latin American development states, but in a less coordinated manner. Here the labour force was more militant, and the governments less cohesive, often in need of broad popular support. Outright repression of labour and adversary shop-floor labour relations thus alternated with periods of populism aimed to appease urban industrial workforce (Evans 1979; Murillo 2001; Kohli 2004). Market protection allowed capital owners to pass on excessive wage increases to the consumers, but it also lowered the incentive for skill investments, which was not compensated by greater public investment in education.

The result was a more erratic productivity trend, which turned stagnant, or even negative with

CEUeTDCollection

48 the crisis of the 1980s. Negative impact on external competitiveness was somewhat softened by selective opening through special economic zones, which relied on labour segmentation and tax exemptions to lower labour costs and recoup competitiveness in some sectors. Probably the most famous example are Mexico’s maquillas in the northern border zone, which offered consistently lower wages and worse working conditions than the domestic-market oriented industries around the capital, and employed a very different type of workforce: rural, less skilled, younger and often female (Kopinak 1997; OECD 1997).

The onset of political and economic liberalization altered the tools of labour control to some extent, but the differences in outcomes between “independentist” and “integrationist”

types only became more pronounced. Democratization in Korea and Taiwan stirred the trade union militancy, and the firms responded by rushing into higher value-added activities to prevent the loss of competitiveness from wage increases. In Latin America, governments resorted instead to radical liberalization of labour markets, replacing political repression with labour market flexibility to ensure wage moderation. Combined with trade liberalization and relative scarcity of skilled labour, this choice only further accentuated the potential for the region’s specialization in low-cost manufacturing. In Mexico, for instance, the share of maquilladora sector in total manufacturing employment sharply increased in the 1990s (Dussel Peters 2003).

Reliable productivity and compensation data for developing countries are scarce, but Figure 2.2 offers a striking illustration of this divergence. It shows trends in productivity and labour costs in manufacturing for Mexico, Korea and East Central Europe between 1990 and 2010. For the sake of comparability, all scores are standardised as percentage of US values in the same period. Data for Mexico is only available from 1995, but the starting point is very similar to that of Korea in the early 1990s. However, relative labour productivity in Korea nearly doubled in this period, closely tracked by increase in wages, continuing the steady earlier trend. In Mexico, meanwhile, productivity remained practically unchanged, oscillating around 20% of the US levels. The gap between productivity and costs is also much larger, indicating

CEUeTDCollection

49 smaller unit labour costs in Mexico, and a continued comparative advantage in labour-intensive industries.

Can we attribute this divergence to different incentive structures within the two development models? The experience of East Central Europe would seem to reject this hypothesis. Although it starts from a lower level than either Mexico or Korea, productivity trends in ECE are closer to the Korean pattern, picking up around the start of 2000s and surpassing the Mexican levels by the end of the decade. Productivity increase is even slightly faster than that of wages, signalling improvements in both quality and cost competitiveness of the region. Nevertheless, there are good reasons to be cautions in interpreting these results as a sign that the ECE’s hyper-integrationist development model had found a durable solution to the problem of long-term productivity growth.

Figure 2.2 Evolution of labour cost and productivity in manufacturing, 1990-2010

5 15 25 35 45 55

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

ECE

Productivity Cost 5

15 25 35 45

55⁵ Korea 15

25 35 45

55 Mexico

Source: own calculations based on OECD productivity database

Note: Labour cost and labour productivity (value added per person employeed) in manufacturing as % of US values

CEUeTDCollection

50 One reason is that productivity growth in ECE since the 1990s is partly the result of a one-off change resulting from the enormous influx of new capital investments, and the gradual collapse of less productive domestic industries. Once this substitution process is complete, and the outdated capital stock has been fully replaced by state-of-the-art technologies, further improvements will have to come from better utilization of existing capacities, including workforce skills. It is, however, unclear, whether the hyper-integrationist model provides the right structure of incentives to motivate further productivity convergence. As we have seen in the Mexican case, left to its own devices FDI seems likely to follow comparative advantages, in this case lower labour costs. Whether the model can be pushed further along the high-wage, high-productivity path will depend on the future balance between strategies of skill formation and cost control. There are, however, several differences between ECE and Latin America which suggest that the outcome of the “hyper-integrationist” model might be more promising.

First, keeping the costs low in the long run might be more difficult in East Central Europe. The legacy of industrialization, together with the collapse of former socialist industries and the dissolution of workers’ organisation in the early transition years have created a large pool of cheap, skilled manufacturing workforce. But unlike in Latin America, which has a much larger young population and an equally sizeable reserve labour force in the informal sector, this situation in the ECE’s labour markets is likely to be temporary. Stronger inflows of foreign investment relative to the size of these economies have helped to quickly absorb the available workforce. In the most competitive manufacturing industries the upward pressure on wages is already palpable, although it was somewhat relieved by the recent crisis. Greater flexibility, which in principle contributes to cost control, can backfire in tight labour markets because it decreases workforce loyalty, and although there are clear signs of attempts to use segmentation in order to combine the benefits of skilled and committed core workforce with the more flexible labour arrangements, the scope for such measures is limited simply by the lack of sufficient number of cheap, unskilled workers.

CEUeTDCollection

51 This leads us to the second major difference between the ECE and the Latin American integrationist types: the legacies of state-led development, which have left East Central Europe with an overall higher level of education. In fact, unlike in most developing countries, the limiting factor on further upgrading of the region’s industries is not so much the lack of skills in general as the lack of specific manufacturing skills which would allow them to move into higher value-added activities. This also means that moving ECE onto the “high road” of higher productivity and wages would be easier, because the skill gap is much smaller and would require less extensive policy intervention. Bearing in mind the long-term wage pressures, it is not impossible that the incoming firms might consider the benefits of further upgrading to keep ahead of the cost curve.

However, while tighter labour markets and greater skill potential both suggest upgrading as a plausible alternative, this does not mean that it will be easy to convince the foreign firms to take it. First of all, unlike basic investments in education, the formation of specific manufacturing skills requires complementary investments by the private sector.

Adequate training in new technologies is too costly for the governments to implement on their own, and will almost certainly not be as effective. So far, however, although there have been many complaints in the region about skill shortages, foreign investors have done little to support public vocational training efforts, except for putting pressure on the governments to do it for them. Moreover, relying on the “productivity whip” to bring mobile transnational firms to invest in more complex activities may be potentially more dangerous. Unlike domestic champions, who draw their comparative advantages from the home markets, transnational firms do so by distributing their activities in places with different structure of advantages.

Therefore, to the extent that the investments in ECE had been motivated primarily by the need to lower production costs along the value chain, foreign companies may not be interested in investing in higher value-added activities, which they can still do profitably at home, and may prefer to move elsewhere in search of lower costs. In theory, this could also result in improved productivity in the host economy, if they manage to attract instead different companies with

CEUeTDCollection

52 different motivations, but this is a more disruptive path than upgrading with ones who are already there.

This creates an additional problem in that individual investments in specific manufacturing skills are unlikely if employment possibilities are so uncertain. The very factors which had made the region so attractive for manufacturing FDI in the first place: the collapse of socialist industries and low wages – are the same ones which make manufacturing employment less appealing for the future generations. The atmosphere of permanent restructuring, together with additional efforts to make employment relations even more flexible in order to control the costs, have decreased the attractiveness of manufacturing jobs in spite of the growing wages. In addition to this, rapid structural shift towards services opened up new opportunities in other sectors, and the regional labour markets still offer very high premiums on higher education.

The governments are thus facing a choice between continuing to support upgrading in complex manufacturing and hope that the firms will play along, and the even more uncertain strategy of allowing the cost advantage to slowly wear off while the region’s skill profiles shift elsewhere. Similar to the problem of technology development, as they have fewer tools to influence the behaviour of foreign firms and force them to cooperate in skill development, governments in the region have been trying to convince the students to commit to vocational training path, mostly by promoting cooperation between schools and companies, and by limiting enrolment in higher education. The latter in particular, however, is proving to be a politically problematic issue, creating a new source of instability in the hyper-integrationist development model.