Catalog

Looking ahead

Developing Asia has an unrivalled record of growth and economic catch-up. It has, compared to other parts of the developing world, deftly navigated difficult changes and transformations. Yet this aggregate picture masks individual country examples of stunted growth, reversals, and weak performance. Such a record can also breed a sense of complacency among economic decision makers about the future. Even where countries are catching up, there is no guarantee that the process will continue indefinitely, and the gaps to be closed are daunting. Much more is still to be done in terms of the transfer of labor and other resources across sectors; technological upgrading; and building a competitive edge in new activities. Change will be needed to sustain and accelerate momentum, and complexity is likely to increase. These linked evolutions will not happen automatically. They require societies to develop and deploy effectively a broad range of capabilities.

This section approaches the challenge of catch-up by looking ahead to see where developing Asia could be in two decades. How much of the remaining productivity gap could be closed? But it also considers what the future might hold for job creation. Successful catch-up requires not just closing the gap on labor productivity, but also creating sufficient new jobs to absorb the new workers. Unless developing Asia can create jobs for its burgeoning labor force, growth could come unstuck. The social fabric could also be at risk.

Closing the productivity gap

Figure 3.1.16 shows the historical evolution of labor productivity to 2000–2004, and then extrapolates ratios out to 2020–2024. These are not forecasts, just a way of focusing attention of the magnitude of the productivity gaps that remain to be closed. In all panels, the ratio of a country's labor productivity to that in OECD is measured. OECD productivity is assumed to continue to grow as in the past, at a rate of about 1.5% a year. Economies in developing Asia are assumed to spurt, and to grow somewhere between 4% and 6%. This is an optimistic range. Korea, Malaysia, and Thailand are set to grow at 4%, the PRC and India at 6%, and all others at 5%. These rates are at the upper bounds of those that have been observed in the historical data and anticipate (slow) convergence within the region. No trajectories are shown for Hong Kong, China; Singapore; or Taipei,China because it can be reasonably assumed that they will complete the catch-up process before 2024, and indeed may overtake OECD productivity levels. Korea's trajectory is shown alongside ASEAN-4 countries.

On the basis of these assumptions, Korea's aggregate labor productivity may reach 80% of the OECD average within two decades. (For the period 2000–2004, it was just slightly less than half the OECD average.) By 2025, this would place productivity in Korea where Singapore was relative to OECD in 1990–1994, or in level terms where Singapore might be expected to reach sometime during 2005–2009. It is clear that even on optimistic assumptions the dynamics of catch-up evolve slowly. Each year the frontier itself moves away, "taxing" catch-up by 1.5 percentage points.

By 2025, these simple extrapolations place Malaysia's productivity at about 35% of the OECD average, with a still-substantial gap to bridge. This compares with where Korea was in about 1985–1989 in relative terms, and would be in 2000–2004 in terms of productivity levels. On the same metric, by 2025 Thailand would reach where Malaysia was in relative terms in 1985–1989 and in level terms where Malaysia was in around the mid-1990s. This would seem to point to a gap of about 20 years between Malaysia and Korea, and a gap of about 30 years between Thailand and Malaysia. Of course, this calculation is somewhat artificial because it assumes fast convergence for all countries from 2004.

The calculations are revealing about the PRC and India and underline the fact that behind these countries' gigantic size and torrid growth, there is still a yawning gap with OECD, and indeed with other countries in developing Asia. Even at an accelerated growth rate of 6%, 1–2 percentage points higher than assumed for other countries, neither the PRC nor India would have 10% of OECD's labor productivity by 2025. By that year, in terms of level productivity these calculations place the PRC where Malaysia was around 1987, and India where Thailand was around 1997. These would be considerable achievements, but still serve as stark reminders of the gaps that remain. In 2000–2004, labor productivity in the PRC compares with where Thailand was at the end of the 1970s. For India the comparison-still with Thailand-is the 1960s.

The numbers tell a compelling story about how large the gaps are for other countries, too. Take Viet Nam, one of the fastest-growing economies in Southeast Asia. Even if the growth of its labor productivity outruns its historical advantage, by 2025 it will have attained a level of productivity that compares approximately with Indonesia in around 2005. And if Pakistan could engineer a productivity U-turn, and emulate the performance of the fast-growing economies of the region, by 2025 its productivity levels would compare with those of Thailand in around 2005.

These comparisons-based as they are on optimistic assumptions about future labor productivity growth-size up the challenge for developing Asia and place in context aspirations for growth. But a bigger challenge still is lurking: creating sufficient jobs for developing Asia's burgeoning labor force. If fast labor productivity growth comes at the expense of jobs, then growth might prove difficult to sustain as social cleavages widen. The challenge is to lift productivity and create jobs.

Prospects for jobs

Historically, the NIEs followed a pattern of development in which fast economic growth was accompanied by rapid expansion of labor-intensive activities and the creation of jobs in the organized sector, particularly in manufacturing. By creating formal sector jobs, poverty was quickly reduced and economic gains were widely spread.

The patterns of development observed now in the PRC and India, and elsewhere, do not readily conform to this "model". In both countries, open unemployment rates have risen and, at least in India, there are high levels of underemployment and informality. Estimates of open unemployment are given in Table 3.1.3.

Various elements are probably contributing to these trends on unemployment. Structural factors may play a role. In India for example, high-end services have contributed significantly to growth, but provide few jobs. Manufacturing activities are also comparatively sophisticated, and presumably capital intensive, for India's level of per capita income. But many other factors are also likely to be playing a role, including labor market regulation, bad infrastructure that raises business costs, and the poor quality of India's education system.

In the PRC too, open unemployment is rising despite impressive growth. Again, institutional features of the economy may help explain this, but structural influences may also be present. The technological sophistication of manufacturing is far above what was observed in countries like Korea at comparable levels of income. The specialization that comparative advantage would seem to predict does not appear to be present (see the previous section, Looking back), and output structures, at least in manufacturing, are becoming increasingly diversified at low income levels.

Countries that are growing more slowly are also having a tough time creating decent jobs. Indonesia and the Philippines see a large exodus of workers abroad each year. Similarly, a lack of decent employment opportunities at home spurs high levels of emigration from Bangladesh, the countries of Central Asia, Nepal, Pakistan, Sri Lanka, and the Pacific island countries. Emigration and jobs overseas would appear to be an important "safety valve" for many young Asians, but deprives countries of expensive human capital.

Reliable estimates of underemployment are hard to come by, but Table 3.1.4 provides some information. "Underemployment" is, by its very nature, difficult to measure. Available estimates tend to focus on "time-based" underemployment, that is, on workers who would willingly work more hours or more frequently. For example, a construction or farm worker wanting to work 45 hours a week but finding employment for only 20 hours is underemployed in a time-based sense. Some indication of the extent of this type of underemployment can be gleaned from labor force surveys. These calculations suggest that time-based underemployment is prevalent. For example, in 2003, 34% of those in work in Indonesia were involuntarily working less than 35 hours a week. Although this type of underemployment in the Philippines has been falling since 2000, it remains substantial, at 17% of total employment. In South Asia, too, a significant proportion of workers seems to be underemployed in the time-based sense. (Estimates of time-based underemployment for the PRC are not considered reliable.)

A simple way of looking at the impact of GDP growth on jobs is to calculate "employment elasticities". If jobs grow at the same rate as output, the employment elasticity has a value of 1. But if growth is entirely attributable to improvements in labor productivity, the employment elasticity is 0. Values between 0 and 1 are compatible with job creation and improved labor productivity, but growth that is labor intensive (values closer to 1) necessarily curbs gains in labor productivity (values closer to 0).

Table 3.1.5 presents empirical estimates of employment elasticities for a sample of developing countries in Asia. As the estimates are all greater than 0 and less than 1, growth in this sample has been associated both with job creation and labor productivity gains (see Looking back). But the responsiveness of employment to economic growth has varied widely across countries and over time. Economic growth has had least traction on job creation in East Asia and most in Southeast Asia. Estimated elasticities for the PRC are consistently on the low side and suggest that labor productivity gains have been an important part of its growth experience (see the previous subsection, Closing the productivity gap). In South Asia, the responsiveness of new jobs to growth generally lies in between the estimates for East Asia and Southeast Asia.

What do these employment elasticities suggest about developing Asia's capacity to create jobs? As the rate of labor force growth slows in most countries, estimated employment elasticities suggest that developing Asia should be able to create sufficient jobs for new workers, provided that growth does not stall and that growth continues to create jobs as it has in the past.

Table 3.1.6 illustrates this point by comparing upper- and lower-bound estimates of employment elasticities with projected labor force growth rates. These calculations suggest that if India can grow at 6%, it can create sufficient jobs for new workers. And even if labor intensity of growth is diluted, emulating say the experience of the PRC, India could still absorb all new workers in jobs if it could grow at 8%. But this is probably close to India's current potential growth rate, and leaves little room for slowdowns or, for that matter, crises.

In the PRC, slowing labor force growth should ease the challenge of job creation. If the historical relationship between income growth and job creation is a good guide, the PRC can create enough jobs for all its new workers if it grows by just 4% each year. This is well below the 10% growth that the PRC has enjoyed over the past decade and above the 8% target that the PRC authorities have set for the medium term. Again, even if labor productivity growth was to spurt (say, halving the employment elasticity), growth of 6% would still generate new jobs sufficient for labor force entrants.

However, not all countries have such a favorable alignment of growth prospects, labor absorption capacity (i.e., aggregate employment elasticities), and expected flows of new workers into the labor force. In Bangladesh, Malaysia, and Pakistan, the labor force is set to expand quickly, and fast growth will be needed to create sufficient new jobs. This could also be a predicament in parts of Central Asia and the Pacific islands. In the Philippines, although the labor force is also set to expand quickly, economic growth has been highly labor intensive in the past. If labor productivity there were to pick up, unless it is accompanied by faster economic growth, this would pose difficulties for future job creation. In some countries, outmigration may continue to release pressures.

There is a crucial rider to these calculations: they tell us nothing about the quality of the jobs being created. As the definition of employment includes workers who are underemployed, it is entirely possible that only modest rates of open unemployment could coexist with extensive and chronic underemployment. Indeed, poverty data and migration statistics strongly suggest that jobs may be of low quality, and that underemployment could pose a significant challenge for developing Asia well into the future.

One way to think about prospects for underemployment is to make the link to poverty incidence. Table 3.1.7 (above) illustrates two projections based on this approach. One paints an optimistic scenario for poverty reduction, which is based on fast growth that is broadly inclusive. The other is based on a less sanguine outlook for growth and income distribution (ADB 2005a).

The optimistic projection for poverty suggests that it may be possible to cut underemployment by 150 million workers between 2005 and 2015. In the pessimistic scenario, projected numbers barely change relative to 2005. The presence of such a large pool of unproductive workers would constitute a tragic underutilization of resources and would pose a clear danger to social stability. In countries like India, failure to tackle the underlying causes of underemployment could yet turn a potential "demographic dividend" into a "demographic curse".

Summary

The challenge for developing Asia is to catch up with OECD productivity levels and to create jobs. Although in a narrow arithmetic sense, faster labor-productivity growth means the creation of fewer jobs for a given rate of output growth, this entails a fallacy: output growth is unlikely to be independent either of productivity growth or, in the long run, of the capacity of an economy to create decent jobs for its workers.

Asia's stellar economic growth disguises wide variation and distracts attention from the full extent of the gap that is still to be bridged. Many countries still have a long road to travel. For most, attaining OECD productivity levels is a distant ambition if only in the sense that, if the past is any guide, it will take the best part of this century to achieve it. For some countries, closing the productivity gap will require a complete turnaround in performance. For others, it will mean sustaining rapid growth for decades to come-itself a difficult challenge. And if there are unpleasant shocks, gaps could easily widen, rather than narrow.

The prospects for job creation are tightly linked to those for productivity catch-up. But what is the likely location of productivity gains and job creation? Is it industry or services that is going to play the most important role in the decades ahead? These questions are examined in the next section.