III. Does Asia Need Rebalancing?
- International Monetary Fund. Asia and Pacific Dept
- Published Date:
- May 2010
Despite the strong increase in regional trade over the last decade, Asia has remained heavily dependent on external demand—more so than other regions. However, in the post-crisis world, Asia will need to rely more on domestic sources of demand to sustain high growth and improve economic resilience. Rebalancing toward domestic demand will require many regional economies to act across a range of areas. Some countries may need to increase consumption, others will need to increase investment, and many will need to boost productivity in service sectors. A comprehensive package of measures—including fiscal measures, reforms in product, labor, and financial markets; and more exchange rate flexibility—will also contribute to the rebalancing of global demand, in particular when implemented simultaneously across the region.
Asia and the world economy need a rebalancing of demand to return to sustained high growth.8 The global financial crisis has put an end to the credit-fuelled consumption boom that started in the United States and other advanced economies in the early 2000s. Long-lasting damages to households’ and financial institutions’ balance sheets, the need for a sizeable fiscal consolidation, and the likely steps toward more stringent financial regulation, mean that domestic demand in these economies is unlikely to return to pre-crisis growth rates. Other economies, particularly those emerging market economies that have saved in excess of their investment over the last decade, will need to step up and fill that vacuum—otherwise global demand will not be sufficient to sustain world growth.
For Asia, weaker demand from advanced economies means that an important source of growth would remain subdued. The increase in trade integration across regional economies over the last decade had raised hopes that Asia could become more resilient to business cycles in the advanced economies, and that regional economies could accelerate their pace of convergence by trading more with each other. But the disproportionate response of export-dependent economies in the region to the collapse of global demand at the end of 2008 shows that Asia’s fortunes remain closely linked to the performance of the global economy. And this region, more than any other, would benefit most from a global rebalancing of demand.
Against this background, this chapter addresses three main questions:
- Is Asia more dependent on external demand than other regions?
- What is needed to strengthen domestic sources of growth in Asia? Are there differences in the region?
- Which policies, both in Asia and abroad, would achieve a successful rebalancing of global demand?
To answer these questions the chapter follows a three-step strategy. First, it quantifies Asia’s growth dependence on external demand by looking at a value-added concept of trade flows—capturing not only direct, but also indirect trade linkages between Asia and advanced economies (e.g., Malaysia’s export exposure to the United States via its exports of intermediate goods to China). Second, it assesses whether consumption and investment in Asia are “too low,” and the amount of resources employed in the tradable sector “too high,” by comparing them with model-based and empirical benchmarks. And third, it uses the IMF GIMF model to assess the impact of macroeconomic and structural policies that could be adopted in Asia and abroad, and that could strengthen domestic sources of growth in the region.
The main findings of the chapter are twofold:
- First, growth in Asia depends on external demand more than in other advanced economies and emerging market regions. This dependence has led to an unbalanced production structure that is heavily tilted towards industry, and more generally, the tradable sector. This means that rebalancing growth will require boosting the productivity of the services sector in most Asian economies. However, only a few economies seem to have excessively low consumption, most notably China, while several others may well be able to increase their investment-to-GDP ratios from current levels.
- Second, reforms in product, labor, and financial markets, and in fiscal and exchange-rate policies, have the potential to produce a successful shift in the pattern of growth. However, these measures will need to be taken by all economies in the region. If only a few of them implement reforms, the rebalancing effort still may have some positive domestic and regional spillovers (especially if taken in larger economies, such as China), but it will not sufficiently compensate for weaker external demand from advanced economies.
B. Assessing Asia’s Export Dependence
Structural changes in the nature of trade complicate the assessment of external dependence. Measures based on simple export-to-GDP ratios will overstate the role of exports as a source of growth, as increasing vertical trade integration means that exports include a declining share of domestically produced value added (see for example, Cui and Syed, 2007, on China). Vertical trade integration also masks the true exposure of an economy to the final source of demand (see April 2008 Regional Economic Outlook). For example, ASEAN economies depend on U.S. demand not only because they export to the United States directly, but also because they export intermediate goods to other countries that then reexport to the United States after processing them (Figure 3.1). To capture both these direct and indirect exposures, we use Asian international input-output (AIO) tables that describe how Asian economies combine domestic and foreign inputs to produce goods. This provides a measure of the extent to which the value added produced in an economy can be attributed to domestic, intraregional, and extra-regional demand (see Technical Appendix for details).9
Figure 3.1.Effect of Vertical Integration on Trade
Using this methodology yields the following results:
- Asia’s export “exposure”—defined as the share of value added linked to external demand—is high by international standards. For all Asian economies in the dataset (except Japan) it exceeds the OECD average, and for most of them this export exposure is even higher than for the highly integrated European OECD members (Figure 3.2).10 Moreover, the gap generally has increased over time. About two-thirds of the exported value added goes to destinations outside Asia—but, in some cases such as Korea, the Philippines, and Taiwan Province of China, the exposure to intra-Asian demand has increased.11 By contrast, China’s export exposure to the region has remained stable at a smaller base.12 While India is not part of the value-added dataset used here, its very low ratio of gross exports to GDP (20 percent in 2008) suggests that it is far less exposed than most of developing and emerging Asia.
- Even where export exposure has been relatively low, exports often have been the main engine of growth (Figure 3.3). In Japan, for example, where export exposure is only about 10 percent of value added, exports contributed about 40 percent to growth between 2001 and 2007 (Figure 3.4)—when domestic demand growth was very weak. In this context, it also is noteworthy that when exports are expressed on a value-added basis, their contribution to growth in Asian economies is much greater than what is suggested by the national accounts—often more than twice as high. The reason is that value-added based measures capture the share of income generated by exports. To the extent that this income is spent on imports, this will reduce the contribution of net exports to growth shown in the national accounts. Nevertheless, using this latter measure as a widely recognized international benchmark, the contribution of net exports to growth across Asia has been consistently higher than in other emerging market regions, particularly Latin America.
Figure 3.2.Share of Export Value Added in GDP
Sources: Japan External Trade Organization (JETRO), Asian Input Output Table (2000); OECD; UN COMTRADE; CEIC Data Company Ltd.; and IMF staff estimates.
Figure 3.3.Selected Asia: Export Dependence and Exposure
Sources: Japan External Trade Organization (JETRO), Asian Input Output Table (2000); OECD; UN COMTRADE; CEIC Data Company Ltd.; Haver Analytics; and IMF staff estimates.
Figure 3.4.Selected Asia: Average Contribution to Real GDP Growth1,2
Sources: Japan External Trade Organization (JETRO), Asian Input Output Table (2000); OECD; UN COMTRADE; CEIC Data Company Ltd.; Haver Analytics; and IMF staff estimates.
1 Average of contributions to 3-year growth rates during 2001-07. Exchange rate adjusted deflator.
2 Latin America includes Argentina, Bolivia, Brazil, Chile, Colombia, Mexico, Paraguay, Peru, Uruguay, and Venezuela.
Asia’s true dependence on external demand is even greater than suggested by these measures for two reasons:
- Export-related investment. Because a significant share of investment in Asia is tied to the export sector, external demand also contributes to growth by affecting investment (see Guo and N’Diaye, 2009a, on China). Simple estimates based on the share of exports in manufacturing output and the share of machinery investment in total gross fixed capital formation suggest that about 30–40 percent of investment in Asia depends ultimately on exports. Taking this into account, our measures of external exposure increases by about 10 percentage points of GDP on average in the region—and China’s export exposure rises to more than 45 percent of GDP (Figure 3.5). Thus, our measures of export dependence (reflecting the contribution of exports to growth) increase by about 5–10 percentage points on average.
- Compositional effects. Asia’s dependence on exports is compounded by its specialization in highly cyclical manufacturing sectors—in particular IT goods13 (Figure 3.6). As noted in the October 2009 Regional Economic Outlook, the global financial dislocation at the end of 2008 led to a collapse in demand for these goods, which is generally more dependent on credit market conditions and consumer confidence. As a result, over the four quarters up to 2009:Q2, U.S. imports from Asia fell about 10 times as much as U.S. consumption. Using our measures of export dependence, the fall in exports may have subtracted between ½ and 2 percentage points from GDP growth across Asian economies over that period14 (Figure 3.7).
Figure 3.5.Selected Asia: Share of Export Value Added in GDP
Sources: Japan External Trade Organization (JETRO), Asian Input Output Table (2000); OECD; UN COMTRADE; CEIC Data Company Ltd.; Haver Analytics; and IMF staff estimates.
Figure 3.6.Selected Asia: Share of Medium and High-Tech Goods in Total Exports
Source: UN COMTRADE database.
1 Latin America includes Argentina, Bolivia, Brazil, Chile, Colombia, Mexico, Paraguay, Peru, Uruguay, and Venezuela. Emerging Europe includes Bulgaria, Croatia, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovak Republic, and Slovenia.
Figure 3.7.Selected Asia: Impact on Value Added of Actual Decline in U.S. Private Final Demand1
1 Decline over 2008-09 in U.S. private durables expenditure, impacting imports in proportion to share of imports in durables consumption and investment in machinery and equipment.
2 Decline in U.S. durable imports impacts U.S. imports from Asian countries in proportion to share of Asia in U.S. imports of consumer durables, and machinery and equipment respectively, based on COMTRADE data.
3 Decline in U.S. durable imports concentrated in imports from Asian countries.
C. What Does Rebalancing Mean for Asian Economies?
A different way of gauging whether Asia depends “too much” on external demand arises from assessing whether its domestic consumption and investment are “too low”. In this section, these questions are tackled by using two different benchmarks for consumption and investment in Asia, one based on the values predicted by simple neoclassical growth models and one based on the results of standard empirical models of investment and consumption.
Is Consumption Too Low?
Excessively weak consumption is often cited as the main driver of current account surpluses in Asia (see Bernanke, 2005). But consumption-to-GDP ratios vary substantially across Asia, ranging from less than 40 percent in China to about 70 percent in the Philippines (Figure 3.8). For many regional economies, consumption is close to OECD or other emerging market averages. At the very least, this argues for some caution in a general characterization that consumption is weak in Asia. What do model-based and empirical benchmarks suggest?
Figure 3.8.Private Consumption
Source: IMF, WEO database.
- Model-based estimates. For a number of Asian economies, actual consumption does not appear to be out of line with our estimates of steady-state consumption levels (see Technical Appendix for details) (Figure 3.9). When compared to deviations for other emerging market economies, they tend to be within one standard deviation of the worldwide sample. However, there is an important exception: China’s consumption stands out as too low—indeed, nearly two-thirds of the region’s (that is, Asia excluding Japan) gross national savings in recent years has been accounted for by China (Prasad, 2009).
- Empirical benchmarks. Cross-country evidence from a structural consumption equation (Guo and N’Diaye, 2010) suggests that weak consumption in China has relatively little to do with country-specific behavioral factors. Instead, it largely can be explained by the low and declining share of household disposable income and a rising saving rate (Figure 3.10). While the share of household disposable income in GDP also has fallen in Japan and Korea between 2000 and 2007, households there responded by maintaining consumption and lowering their saving rates.15 The high household saving rate in China mainly reflects the need for precautionary savings to offset a lack of social safety nets and services, and demographic factors such as gender imbalances.16 Inadequate access to financial services (including consumer credit and housing finance) also may have played an important role. As explained in further detail in Box 3.1, reforms that increase social insurance and services, notably health care coverage, could prove effective in boosting consumption in China. Moreover, measures that improve access to financial services, such as Korea’s reforms to develop a mortgage market, also hold the promise of strengthening private consumption (Box 3.2).
Figure 3.9.Consumption Relative to Steady-State1
Source: IMF staff estimates
1 Uses savings-investment balance in the CGER’s Macroeconomic Balance approach.
2 Other emerging economies include Argentina, Brazil, Chile, Colombia, Egypt, Israel, Mexico, Morocco, Pakistan, Peru, South Africa, and Turkey.
Figure 3.10.Cumulative Change in Ratio of Private Consumption-to-GDP (2000–07): Estimated Contribution of Disposable Income and Savings Rate
Source: Guo and N’Diaye (2010).
Is Investment Too Low?
Asia’s high current account surplus also may be the result of an investment slump. In emerging Asia, investment-to-GDP ratios collapsed by about 10 percentage points of GDP or more immediately following the Asian crisis and have remained low (in particular for economies hit most severely by the crisis), but savings remained broadly stable (see Box 3.3). While at least part of the decline can be attributed to overinvestment in the years leading to the crisis,17 there are large differences in investment-to-GDP ratios across the region, suggesting again some caution against generalizations (Figure 3.11). With the exception of Malaysia and the Philippines, most emerging Asian economies have overall investment ratios that are higher than the OECD average and closer to the average for middle-income countries, consistent with their lower stage of development. For industrialized Asian economies, the investment ratio is close to the OECD average (about 20–25 percent), except for Korea. Moreover, there are important differences between public and private investment. Indeed, in economies hit by the Asian financial crisis, the collapse of investment was mainly due to private investment. Public investment has tended to play a relatively greater role after the crisis (notably for Korea, Malaysia, and Thailand), reflecting programs to upgrade infrastructure, thereby partly offsetting the fall in private investment.
Sources: World Bank, World Development Indicators; and IMF, WEO database.
- Model-based estimates. In countries where the capital-output ratio is below its long-run level, investment rates should be above their long-run level—these countries are catching up with their peers and need to invest more. If they fail to do so, investment rates are “too low”—these countries are in the underinvestment quadrant (Figure 3.12). Based on our estimates of steady state levels of capital and investment ratios for a set of about 30 (mostly) emerging market economies, some ASEAN economies (with the exception of Indonesia) seem to have moved in recent years deeper into the underinvestment quadrant than other countries (see Technical Appendix for details). By contrast, China and India appear farthest out in the overinvestment quadrant. Having said that, these model-based estimates should be treated with some caution, mainly because they do not reflect structural changes (such as a shift away from growth that is intensive in capital, infrastructure, or residential investment).18
- Empirical benchmarks. Estimates based on an empirical model of investment (see Box 3.3) suggest that weaknesses in the investment climate (notably governance) and macroeconomic uncertainty may hinder private investment in the region, particularly among the ASEAN-4 economies (Figure 3.13).19 Indeed, efforts to improve the business climate—including strengthening the rule of law, creditor rights, and transparency of government operations; reforms that facilitate access to finance; and efforts to level the playing field between foreign and domestic or private and public investors—have all been on the reform agenda (see Box 3.3) in many economies in the region as a way to boost both domestic and foreign investment. However, despite progress, investor perceptions seem to improve only slowly (see also Hori, 2008 and May 2006 Regional Economic Outlook).
Figure 3.12.Capital-to-Output and Investment-to-Output Ratio
Sources: Penn World Tables; IMF, WEO database; and staff calculations.
By contrast, already high levels of investment in China and India present different challenges. China’s unusually low level of consumption and dependence on external demand point to a need to shift the composition of investment away from manufacturing and export industries and more toward the social and consumer sectors in order to reorient growth toward consumption. On the other hand, India, with a level of overall investment between that of China and the ASEAN-4, and with demand already well balanced between external and domestic sources, would benefit from financial reforms and fiscal consolidation to catalyze private sector participation in infrastructure investment and support faster growth.
D. Are There Any Supply-Side Imbalances?
The supply-side counterpart of high dependence on external demand could be represented by an over-reliance on the tradable sector. For example, at about 50 percent, China’s share of industry in GDP is nearly twice as high as the OECD average, and more than 10 percentage points of GDP above the world average for low- and middle-income countries (Figure 3.14). ASEAN economies, Japan, and Korea, however, also have a relatively high share of GDP in industry. As a mirror image, the share of services in GDP is generally lower in Asia, with the notable exception of India. Regression estimates confirm that the share of industry in GDP is above the “norm” in China and the ASEAN-4 economies (Figure 3.15). The same also appears to be true even for Singapore (although model estimates are not statistically significant), which as a financial center already is a mainly services-based economy. Similar results are obtained when looking at employment shares, broadly confirming the over (under-) exposure of Asian economies to industry (services).
Figure 3.13.Model Based Determinants of Deviation of Investment-to-GDP Ratio from Sample Mean
Source: IMF staff estimates.
1 ASEAN-4 includes Indonesia, Malaysia, Philippines, and Thailand.
2 Latin America includes Argentina, Brazil, Chile, Colombia, Mexico, and Uruguay.
3 Emerging Europe includes Bulgaria, Croatia, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Serbia, Slovak Republic, and Slovenia.
Figure 3.14.Selected Asia: Excess Share of Industry in GDP1
Sources: World Bank, World Development Indicators; and IMF staff calculations.
1 Peer-group definitions are based on World Development Indicators. Countries were assigned the following peer-groups: Hong Kong SAR, Japan, Korea, and Singapore (OECD); China, Indonesia, the Philippines, and Thailand (low-and-middle income countries); India (lower middle income countries); and Malaysia (upper middle income countries).
Figure 3.15.Selected Asia: Model Based Excess Share of Industry in GDP1
Source: IMF staff estimate.
1 Estimated by country-specific dummy variable. Explanatory variables for industry share include per capita income (PPP), population and geographic size. An asterik mark next to the economies indicates statistical significance at 5 percent level.
At the same time, productivity in the service sector generally has been low in Asia. Industry and services broadly reflect the tradable and nontradable sectors, respectively. However, productivity growth in Asia’s service sector has stagnated relatively to the United States in recent years (Figure 3.16). According to empirical studies, deregulation and further opening to foreign competition would help in unlocking the services sector’s growth potential (Nicoletti and Scarpetta, 2003; Conway and others, 2007). Indeed, in many regional economies policy efforts already have been directed at allowing greater competition in infrastructure-related services, further opening the retail and financial sectors, and lifting restrictions on entry into social services, such as health and education.
Figure 3.16.Asia: Productivity Levels
Sources: World Bank, World Development Indicators; and IMF staff estimates.
However, India’s vibrant service sector is an interesting counterpoint. In India, the services sector has been among the economy’s most dynamic (in part reflecting higher productivity growth), leading GDP growth for the last two decades. Nevertheless, employment in the formal sector has grown slowly compared to countries with growth led by manufactured exports. A broadening of infrastructure investment to improve connectivity to markets and liberalization of labor laws and other regulation would raise the prospects for growth in industry and agriculture, which are traditionally more labor intensive.
E. Impact of Rebalancing: Illustrative Model Simulations
How much can policies across the region contribute to sustained growth and what are the spillovers? This section uses multicountry simulations, based on the IMF’s GIMF model, to help illustrate how a successful implementation of structural and macroeconomic policies across the region could boost domestic sources of growth and bring about rebalancing. The set of policies considered here (see Box 3.4) ranges from structural reform to boost productivity and thereby raise investment in the nontradable sector, to financial sector and fiscal policies reducing the incentives for precautionary savings, to moves in the REER. The combination of policies across economies in the region is guided by the evidence from the imbalances in demand and resource allocation presented in the previous section. The policies also generally are consistent with medium-term reform plans already being implemented or envisaged by the authorities of the selected economies.
Main Scenario: Simultaneous Rebalancing By All Economies
The central scenario considered is one where Asia successfully rebalances in the face of weak demand from the United States. This scenario envisages an increase in the private U.S. saving rate of 2½ percent of GDP above the baseline (current IMF World Economic Outlook projections) and a protracted decline in private investment.
This is combined with a full-fledged rebalancing scenario in Asian economies, whereby all policy measures (Box 3.4) are simultaneously implemented across regional economies to boost domestic demand and productivity in the services sector. As most reforms and their benefits will take time to materialize, this section focuses on their medium- to long-term impact.20
- Output rises in all Asian economies relative to the baseline projections presented in Chapter I, reflecting the implementation of structural reforms and positive spillovers throughout the region and the rest of the world (Figure 3.17). U.S. output still would fall below the baseline, but the decline would be less than without rebalancing efforts in Asia as a whole or some of its parts. For example, GIMF model simulations by Blanchard (2010) show that rebalancing efforts in emerging Asia (involving lower savings and a more flexible exchange-rate regime) would boost U.S. output by about 1 percent over a scenario without rebalancing.21
Source: IMF staff estimates.
- Investment rises relative to GDP in all economies in the short and medium term, except in China (Figure 3.18). Also, investment rises, primarily in the nontradable sector, as structural reforms lift productivity. In China, while firms in the nontradable sector invest more, those in the tradable sector invest less as the cost of capital increases and the exchange rate appreciates. The fall in investment in the tradable sector dominates initially. However, over time aggregate investment rises as there is more and more investment in the nontradable sector to meet growing demand.
- Private consumption-to-GDP ratios rise in Australia, China, emerging Asia, and New Zealand (Figure 3.19). The higher share of private consumption reflects the effects of financial sector reforms that result in better access to finance and reduce precautionary savings, higher labor income, and wealth effects from lower payments on foreign liabilities (in Australia and New Zealand, through lower interest rates and some currency appreciation).22 Consumption falls in relation to GDP in Japan and Korea but for different reasons, at least in the short to medium term. In Japan, the projected increase in the consumption tax rate to reduce the level of public debt dampens consumption initially, notwithstanding the positive impact of fiscal consolidation on risk premia and the lowering of precautionary savings that result from such a consolidation. However, the level of consumption rises in the longer term, as the benefits of structural reforms are felt and the tightening fiscal stance ends. In Korea, the fall in the consumption–to-GDP ratio simply reflects the fact that the reforms stimulate mostly investment and that GDP rises faster than private consumption.
- Rebalancing Asia’s growth would also lower global imbalances. Current account balances are lower in most Asian economies, owing to stronger domestic demand, lower export demand from the United States, and an appreciating currency (Figure 3.20). In the short run, the lower current account balances reflect lower exports and more imports relative to the baseline; while in the longer term, stronger import demand from China raises exports in Japan, Korea, and emerging Asia (Figures 3.21 and 3.22). Economies that benefit most from the stronger demand from China are those whose production structure is ready to produce the final consumption goods that China will demand (notably Korea).
Figure 3.18.Investment-to-GDP Ratio
Source: IMF staff estimates.
Figure 3.19.Consumption-to-GDP Ratio
Source: IMF staff estimates.
Figure 3.20.Current Account Balance-to-GDP Ratio
Source: IMF staff estimates.
Source: IMF staff estimates.
Source: IMF staff estimates.
For India the spillovers from rebalancing in Asia likely would be more limited. As noted in Section B, India is much less export dependent than the rest of Asia and thus would benefit less from strengthening domestic demand in the region. Moreover, it is still less integrated with Asian trading partners. For example, the combined share of China, Japan, and Korea in India’s capital and consumer goods exports is only about 2–3 percent, 5 to 10 times less than the respective shares for the United States. At the same time, spillovers from India’s rebalancing to other Asian economies also are likely to be more limited. As noted above, India’s growth already is mainly reliant on domestic drivers, and rebalancing in India will not as much entail a shift from external to domestic sources of growth but more narrowly focus on addressing structural bottlenecks, notably infrastructure. Improving connectivity to markets, domestic and foreign, would help redress India’s overreliance on services and boost overall employment through better prospects for labor-intensive industry and agriculture. However, for other Asian economies, spillovers from the likely rise in capital goods imports would again be limited, as India is only a marginal buyer on the world stage, accounting for 2 percent or less of their capital goods exports.
Alternative Scenarios: Partial Rebalancing
Rebalancing in some individual economies alone will have positive spillovers to other Asian economies, but it will not fully offset the lack of external demand.
- We focus on China, which often has been at the center of the debate on Asia’s contribution to a global rebalancing of demand. This may reflect (in addition to the large size of its current account surplus) two factors: first, the scope for boosting China’s low consumption rate appears large by international standards (as shown in section C); and second, China’s economic dynamism, including the fact that China has recorded very high import growth rates for consumer goods—about 15 percent per year over the last 15 years, compared with a world average of 10 percent.
- Nevertheless, rebalancing growth in China alone will not fully offset the lack of external demand. For example, the negative impact on Asian exports of a rise in the saving ratio in the United States would not be offset by an equal decrease of that ratio in China. Positive spillovers to the region from greater Chinese demand would at best mitigate one-third of the adverse shock (see also October 2009 Regional Economic Outlook). One important reason is that, despite high growth, China has remained a marginal importer of consumer goods—accounting for only 3 percent of global imports—while the United States still dominates global imports, both in terms of direct and indirect trade linkages. Moreover, countries in the region, especially those currently catering predominantly to demand in advanced economies, would need to adapt their production lines to the new Chinese demand for final goods.23
- Moreover, a faster appreciation of the renminbi than that considered in our central scenario (Box 3.4) would have only limited additional positive spillovers for trading partners, including the United States. A scenario analysis that adds increased exchange rate flexibility (with a 10 percent real effective appreciation of the renminbi over two years followed by a floating exchange rate) to the main scenario presented above, without accelerating reforms by China and other countries, shows that China’s current account would only be lowered by an additional ¾ percent of GDP relative to the central scenario, while the U.S. current account would improve by ¼ percent of GDP. The ultimate impact on growth in China and elsewhere depends on how fast private consumption can be catalyzed. Exchange rate appreciation would help support private consumption by raising the labor share of income, but structural rigidities could slow the transition toward a more balanced economy.
How can Asia achieve a successful rebalancing? Asia’s growth has depended highly on external demand, according to various measures and by international comparison. This dependence has led to an unbalanced production structure, leaving the nontradable sector relatively underdeveloped. From the domestic demand side, ways to make growth more resilient are more varied, as only a few economies need to raise private consumption and several need to promote investment.
Simultaneous implementation across the region of a package of measures, including a combination of reforms in product and labor markets, fiscal and exchange rate policies, and financial markets (where measures to improve access to finance are complemented by prudential policies to avoid excessive risk taking) has the potential to bring about a successful shift in the pattern of growth. When implemented on a standalone basis by individual economies, including China, rebalancing efforts will help increase growth in Asia but will not be sufficient to fill the void created by weaker external demand from advanced economies. In fact, the global slump has motivated a rethink of development strategy beyond the cyclical recovery across the region. For example, in Singapore, a high-level Economic Strategies Committee chaired by the Minister of Finance has laid out recommendations to make long-term growth sustained and inclusive. The 2010 budget implements some of these recommendations through measures to boost productivity growth and the role of domestic workers in the economy. In Malaysia, the government has unveiled a New Economic Model with an ambitious reform agenda to revitalize the economy by addressing long-standing structural and fiscal weaknesses. If implemented, this vision could double per-capita income in a decade.
Box 3.1.Boosting Consumption in China: Is Public Spending on Health the Right Medicine?
Strengthening consumption is a key objective of China’s policymakers, and an integral part of their strategy to rebalance growth. This box reviews a recent IMF working paper (Barnett and Brooks, 2010) arguing that one important dimension to boost consumption is to increase public spending on health. The paper’s key finding is that an increase in public health spending of 1 yuan translates into a 2-yuan increase in urban household spending.
One factor often cited for the high household saving rate is precautionary saving. The reforms in the 1990s—especially to state-owned enterprises—led to the breaking of the so-called iron rice bowl as existing systems for providing education, health care, and pensions were dismantled. This led to an increase in uncertainty facing households, and helps explain the rise and high level of precautionary saving. It also points to a potential solution: Reduce precautionary saving by strengthening the provision of social services, so that households have less need to self-insure. This certainly holds in theory, but is it also true empirically?
Empirical evidence supports the view that increased public social services reduce precautionary saving, at least for health spending in urban households. As shown in Barnett and Brooks (2010), regressions using a time series of provincial data yield a fairly robust finding that a 1-yuan increase in government health spending increases urban consumption by 2 yuan. The result also holds for the higher-income rural provinces. Separate work done with the Fiscal Affairs Department of the IMF (Baldacci and others, 2010) arrives at a similar finding, but using illustrative estimates for the relationship between the saving rate and social expenditures in China based on panel estimates for OECD countries.
The empirical finding is important, as ex ante the impact of higher government spending is ambiguous. On the one hand, the more that government provides, the less that household have to pay themselves. Therefore, government health spending could substitute for private spending and thus lead to lower household consumption. On the other hand, if government spending reduces the need for households to self-insure against the risk of a large future medical bill, then household consumption would rise. Indeed, private health care spending by urban households has fallen in recent years, suggesting that higher government spending on health care has potentially freed up resources for households for other spending or saving. The empirical results indicate that this second effect dominates, and that higher government health spending leads households to save less and spend more.
In contrast to health spending, there seems to be little evidence to support the view that higher government education spending boosts consumption. However, this could reflect that pre-cautionary motives are centered on higher education costs, whereas most of the increase in government spending has been toward primary and secondary education, or that expected household education spending has risen even faster than the increase in government spending.
China: Health Care Spending
Source: China Household Survey.
Focusing on health spending, the results corroborate the conventional wisdom that precautionary motives help explain China’s high saving rate. It is encouraging that Chinese policymakers have made improving health care a priority. Doing so will of course directly benefit the people of China by improving access to and the quality of medical care. It also will have the beneficial side effect of boosting household consumption, and thereby contributing to rebalancing China’s economy.Note: The main author of this box is Steven Barnett.
Box 3.2.Consumption and Financial Development in Asia
Financial reform and the resulting changes in access to financing are closely related to a household’s decision on the level and timing of saving. Theory indicates that the overall effect of financial reform on savings may well be ambiguous. In a more liberalized financial market, households might have easier access to finance and therefore need to save less, as the precautionary motive is less pressing. Yet, at the same time, in a more liberalized financial market, interest rates may rise, which in many instances may induce households to save more. Changes in access to mortgage financing also can have an ambiguous effect, depending on the starting point. Lowering down payment requirements from a high to a moderate level can induce more savings, as households that previously could never meet that high down payment might start to save to meet the lower requirement. Further declines in down-payment ratio eventually may lower savings (by reducing the amount of consumption forsaken in order to meet that requirement). This box analyzes the empirical link between household consumption and the extent of financial development, including for China and India, and asks which of the competing effects prevail.
How does financial sector development affect private consumption? We estimate a panel regression based on a cross-country database (24 advanced and emerging market economies) which relates the consumption share of GDP to measures of financial sector reform, controlling for other possible determinants of consumption including per capita income (measured in terms of purchasing power parity), share of household disposable income in GDP, private sector credit, money supply, and the real interest rate. The measures of financial development include indicators for (1) entry barriers (e.g., licensing requirements and limits on the participation of foreign banks), (2) interest and credit controls, (3) privatization in the financial sector, (4) international financial transaction restrictions, (5) operational restrictions on securities markets, (6) banking supervision, and (7) an overall financial reform index (Abiad, Detragiache, and Tressel, 2008).1 The main findings are:
Financial Reform Index, 2005
- Financial reform is associated with a higher household consumption share in GDP. As an illustration, the results suggest that raising China’s financial reform index in 2005 to that of Korea2 would raise household consumption by 2.4 percent of GDP, while raising that index to the maximum level (in line with the G-7) would raise it by 5 percent of GDP. While this would represent a substantial improvement, a large gap would remain between China’s consumption level of about 35 percent of GDP and the regression based norm (which would call for a consumption level of about 50 percent of GDP).
- Could India benefit as much as China? While most attention is focused on China, given its unusually low consumption, India’s consumption also is relatively low. Based on the regression model for consumption used here, in 2005, India’s consumption level was about 9 percent of GDP lower than that suggested by cross-country comparisons. However, since India is relatively more financially developed (under the financial reform index used, in 2005, India and China’s index value was 87 percent and 68 percent of Korea’s index level respectively), raising India’s financial reform index in 2005 to that of Korea would only increase consumption by 1 percent of GDP.
- Financial depth also plays a role. In addition to financial reform, financial depth, which measures not only the underlying institutions but also the supply and demand for credit, is considered. A 1-percentage point increase in the ratio of private sector credit-to-GDP would raise consumption by 0.05 percent of GDP.
What are the policy implications? Financial reform can play an important role in boosting consumption and can usefully complement other initiatives. Improved household access to financial services, particularly consumer credit or housing finance, lessens incentives to save or hold precautionary cash balances, thereby boosting consumption. Policymakers should identify financial reforms, which could boost consumption without contributing to systemic risks. Collateralized consumer loans (such as through payroll loans in Brazil) tend to be relatively safe, but could have limited effect on consumption unless there is a latent demand for such loans. An expansion of access to housing-related loans (for example reforms in Korea from 1999 onwards to make the mortgage market more competitive and deeper) could have a larger potential impact on consumption, but would involve more systemic risk.Note: The main authors of this box are Marcos Chamon and Sonali Jain-Chandra.1 Due to the correlation of the financial reform indices among themselves, we consider only one reform indicator at a time. The financial reform database ends in 2005.2 China’s current level of financial sector development as measured by the Abiad, Detragiache, and Tressel database is roughly equivalent to Korea’s level in 1985.
Box 3.3.Rebalancing Growth in Asia: What Role for Investment?
Since the Asian financial crisis, investment has remained low in large parts of emerging Asia. Moreover, investment across the region remains largely export driven, while there are still significant infrastructure needs. This box reviews some of the impediments that may have hampered investment in Asia, and how policies can help in addressing them.
Emerging Asia: Savings and Investment1
Souce: IMF, WEO database.
1 Excluding China and India.
Investment spending in much of emerging Asia has declined sharply since the late 1990s, and has since remained at relatively low levels. The average investment-to-GDP ratio in emerging Asia (excluding China and India) has fallen by nearly 10 percentage points since the Asian financial crisis, and now settles at around two-thirds of its pre-crisis peak. This mainly reflects a collapse in private investment, while public investment held up. With savings relatively stable, this collapse in investment has been an important contributor to the region’s large current account surpluses.
Selected Emerging Asia: Change in Investment Share of GDP
Source: IMF, WEO database.
1 Pre-crisis average calculated over 1986-1996, and post-crisis average calculated over1999-2009.
India and China are notable exceptions to this trend, with investment-to-GDP ratios steadily rising. In India, in particular, private investment spending has become an increasingly important component, accounting for nearly three-fourths of overall investment. Notwithstanding continued infrastructure development needs, sustaining high rates of private investment appears to be less of a concern in these countries than in the rest of the region. Accordingly, most of the subsequent discussion focuses on developments outside China and India, and emerging Asia is used to refer to economies excluding these two countries.
While the decline in overall investment ratios in emerging Asia can be in part explained as a correction of the excesses that led to the financial crisis at the end of the 1990s, the composition of private investment in the region also has changed over the last decade, shifting toward larger, export-oriented firms in the manufacturing sector. Domestically oriented and service-sector firms have exhibited much less of a recovery since the crisis. In particular, investment by smaller firms has lagged behind across emerging Asia, reflecting weaker fundamentals since the crisis, notably lower profitability and liquidity. As rebalancing emerging Asia’s growth model involves reorienting the production structure and pattern of spending away from external to domestic drivers of growth, there may be scope for shifting capital spending toward firms and sectors more directly linked to the domestic economy—even in countries such as Korea and Singapore, where aggregate investment does not seem abnormally low.
What explains these trends?
Staff estimates suggest that part of the decline in aggregate investment can be explained by structural changes following the Asian crisis.1 Relative to the early 1990s (1) lower returns have decreased investment by 5 percent of GDP; (2) greater uncertainty by slightly more than three-fourths of a percentage point; and (3) deterioration in perceptions about the business climate is associated with a further decline of three-fourths.2
Emerging Asia: Elasticity of Investment Rate to Fundamentals
Sources: Worldscope; and IMF staff calculations.
A more disaggregated firm-level view points to lack of external financing as a major constraint, especially for small firms.3 Since the Asian crisis, investment in the ASEAN economies has become more sensitive to the availability of internal funds. This effect is especially acute for firms that are small in size, more domestically oriented, and operating in the services sector. One possible explanation is that all these factors hinder the ability of firms to offer collateral (e.g., physical assets and hard-currency receivables). However, in the NIEs, financing constraints appear less binding and are confined to domestically oriented firms and those with relatively higher labor intensity. In addition, the firm-level analysis suggests that higher leverage holds back investment, both in the export and services sectors. By contrast, the association between investment and fundamentals is much weaker in China, and there is less evidence of differentiation in the determinants of investment across firms in India. This seems to suggest less scope for rebalancing through reforms to further boost private investment in China or to reorient private investment away from exports toward domestic sectors in India.
What can be done to boost investment and shift it towards domestic demand?
- Further improve the investment climate. While the structural reforms implemented since the Asian crisis have potentially made a substantive difference in the region’s investment climate, it appears that perceptions have not yet caught up with the new reality. It will be important to continue with ongoing efforts, such as making product and labor markets more competitive (e.g., adoption of a competition law in Hong Kong SAR), leveling the playing field for foreign investors (e.g., by lowering restrictions on foreign investment in the services sector, as recently done in Malaysia), ensuring contract enforcement, and reducing administrative bottlenecks (e.g., one-stop shops for foreign investors as introduced in Indonesia and Malaysia).
- Better access to finance. Even for larger firms, funding sources can be broadened as only Korea and Malaysia have a sizable corporate bond market in emerging Asia. Moreover, the financial infrastructure for smaller and more service-oriented firms needs improving by more lending on risk-based terms; reforming collateral laws to allow for a wider range of securitization (beyond real estate and other fixed assets); and by widening the pool of venture capital funding through targeted tax breaks of the kind introduced by Malaysia. Deepening credit information and extending the coverage of credit registries, as was done in the Philippines through the establishment in 2008 of Credit Information Corporation, would also improve access to finance by helping banks to better assess credit risks.
- Reduce credit risk by facilitating corporate restructuring of SMEs. For example, the Korea Asset Management Corporation (KAMCO) successfully created a market for distressed Korean corporate debt by purchasing NPLs from banks and repackaging them for eventual sale to investors.4 A similar restructuring and consolidation of the SME sector might be accomplished usefully by promoting asset management companies that specialize in repackaging distressed debt of small firms.
- Increase infrastructure investment. Over the next decade, Asia needs to invest nearly US$7½ trillion in infrastructure, including transport, energy, and communications.5 Despite improvements, stark disparities in the quality of infrastructure remain, with China, India, Indonesia, and the Philippines lagging. Demand for infrastructure in these countries also is being driven by demographics and rapid urbanization. To help meet the resulting financing needs, India, for example, has employed the Public Private Investment Partnership model for investing in roads. The viability of projects is further supported through capital grants, reduced tariffs on imported machinery and equipment, and easier norms for external commercial borrowing.
Selected Asia: Size of the Corporate Bond Market, 2009
Sources: Asian Development Bank; and Bank for International Settlements.
Box 3.4.The Asia GIMF Model and Policy Assumptions
The GIMF model used in this chapter includes eight regions: five Asian blocks (China, emerging Asia, Japan, Korea, and Australia and New Zealand), the United States, the euro area, and the rest of the world.1 Consistent with section B of this chapter, the model also incorporates several layers of production, distinguishing between the manufacturing of intermediate goods, distribution of intermediate goods to domestic and foreign assemblers, and final production of consumption and investment goods. It thus captures a more comprehensive (that is, direct and indirect) transmission of external shocks and spillover effects from rebalancing.
Policy assumptions for rebalancing efforts of Asian economies in the main scenario are:
- In China, the package of reforms assumed is similar to the one envisaged in N’Diaye, Zhang, and Zhang (2008) and includes (1) structural reforms in the services sector that raise productivity (e.g., enhanced competition between domestic and foreign firms, liberalization of retail services) accompanied with a shift in households’ preference toward nontradable goods; (2) fiscal reform aimed at reducing precautionary saving by increasing coverage of education, health care, and pensions, and improving infrastructure in rural areas; (3) further financial development and liberalization (including interest rates) to enable better smoothing of household consumption, capital allocation, and improved risk management by banks; and (4) a gradual real effective appreciation of the renminbi—20 percent over 10 years for illustrative purposes—that supports the transition to greater reliance on the nontradable sector and stimulates private consumption by raising labor’s share of income.
- In Japan, the reform package is geared toward improving productivity in the services sector (by ½ percentage point per year over five years), liberalizing labor and product markets, and putting public finances on a more sustainable path. Fiscal consolidation aims at lowering the fiscal deficit by 5 percent of GDP over five years (similar to the consolidation during 2002–07, albeit a period characterized by healthy growth), with half the burden falling on tax increases and the other on expenditure cuts. The rise in taxes is achieved solely through a gradual hike in the consumption tax rate.
- In Korea and emerging Asia, structural reforms aim at raising productivity in the services sector by encouraging greater competition and leveling of the playing field between the tradable and nontradable sectors, as well as enhancing flexibility in labor and product markets, and further developing domestic capital markets including by reducing credit constraints for households.
Technical Appendix 3.1
Asian Input/Output (AIO) Tables
Updating the AIO 2000 Table and Data Sources
Our updating procedure follows Pula and Peltonen (2009). The AIO table is divided into three blocks: Intermediate demand (A), Final demand (F), and Exports (L), as shown below:
The elements in each of these blocks are updated as follows:
Intermediate Demand block (A)
- Value added : Growth rates of value added are calculated from National Accounts data for each country (data from CEIC).
- Total output : Since total output data for the AIO countries is not available except for 2000 (with the exception of the United States), the ratio of output to value added in the manufacturing sector of each country is used as a proxy to estimate total output from GDP figures. Data on manufacturing value added comes from National Accounts data, while data on manufacturing output is based on industrial production data, (both from the CEIC database).
- Imported inputs : For imported intermediate inputs, we use two data sources. National Accounts data on imports was used to estimate the growth rate of overall imports, including goods and services. In order to capture changes in the direction of trade, we used COMTRADE data for intermediate goods. Since COMTRADE only covers merchandise trade, we assume that the change in imports by direction and type are similar for goods and services. Let i, j stand for country of source and country of demand respectively. Thenwhere int stands for intermediate, M for imports, and NA and COM for National Accounts and COMTRADE respectively. The figures for 2000 are incremented by the growth rate of total imports, weighted by the change in the share of bilateral intermediate imports in total goods imports of country j.
- Freight, insurance, and import duties : Assuming the share of these items in total imports remains unchanged, the growth rates are set equal to the growth rate of total imports from National Accounts. Domestic intermediate inputs: This is estimated as the residual, equal to total imports less total imported inputs, that is
Final Demand block (F)
The updating procedure for this block is similar to that described above. Final demand is updated separately for consumption and investment demand, as follows:
- Consumption and investment : National Accounts data are used to update these series. Consumption is taken as the sum of private and government consumption expenditure, while investment is defined as the sum of gross fixed investment and inventories.
- Imported final goods and investment goods : growth rates are calculated as follows:andwhere c stands for consumption and cap for capital goods.
- Freight and insurance, and import duties : Same procedure as used for the intermediate block.
- Domestically produced final goods : This is the residual, estimated as total final consumption less sum of imported final goods, for each type of good.
Export block (L)
Exports to Hong Kong SAR, EU-15, and rest of the world (RoW)
Tracing value added to domestic, intraregional, and extraregional demand
Our exercise requires calculating the Leontief inverse matrix, as follows. From the supply side, the basic input-output equation can be written as AX × Y = X where A is the matrix of input coefficients
For any final demand vector(or matrix)fj the impact of fj on total output is given by
The contribution ratio of final demand in country j to the value added in country i is given by
Adjusting for Hong Kong SAR’s Entrepôt Trade.
Reexports account for a high proportion of Hong Kong SAR exports (97 percent). If not accounted for, this will lead to an overestimate of the contribution of Hong Kong SAR final demand to value added in the AIO countries. To adjust for this, a part of the unadjusted contribution of exports to Hong Kong SAR to value added is reassigned to other countries and regions as follows:
We follow the approach set out in the September 2005 World Economic Outlook, Box 2.4 “Is Investment in Emerging Asia Too Low?”
In a neoclassical growth model, the steady-state level of investment (i*) is given by: = k*(g + d)/(1 + g), based on estimates of a steady-state capital-output ratio (k*), the depreciation rate (d), and the rate of potential output growth (g).
The stock of capital is derived from the standard perpetual inventory method. Data on gross fixed real investment during 1950–80 is obtained from Penn World Tables, and from 1980 onwards from the IMF World Economic Outlook.
For a given deprecation rate, k* is found as the maximum value of the capital-output ratio on average over long (15- and 20-year periods) between 1950 and 2008. This helps ensure robustness, particularly vis-à-vis boom and bust periods such as the Asian financial crisis.
Results reported here are the averages for two models, one with a deprecation rate of 5 percent and the other with a rate of 7 percent.
The rate for potential output growth is consistent with medium-term IMF World Economic Outlook projections.
Levels for steady-state consumption are derived from those for steady-state investment (see above) and country-specific saving-investment norms, exploiting the national accounts identity for open economies: Y* - (S-I)* -I* = C*, with Y representing income, S saving, I investment, and C consumption.
The saving-investment norms (S - I)* are based on staff estimates using the IMF’s Consultative Group on Exchange Rates (CGER) Macroeconomic Balance approach. These are structural current account balances based on panel regression estimates that reflect country characteristics such as level of development and demographic factors, and abstract from cyclical variations in current account balances.
See Kalra and others (2009). Input-output tables have been constructed for the years 1995–2008 based on the 2000 Asian Input Output Table provided by the Japan External Trade Organization using the methodology in Pula and Peltonen (2009).
The OECD data are based on total external demand as opposed to final external demand for the Asian economies owing to data limitations. This makes the latter an even more demanding benchmark.
Calculations have been adjusted for Hong Kong SAR’s entrepôt trade which would otherwise lead to an overestimate of the contribution of Asian demand to value added in the region.
China’s exposure to external demand from outside Asia has increased from 15 percent of total value added (1995–2000 average) to over 19 percent (2001–08 average) while its exposure to external demand from Asia has remained stable at about 6 percent of total value added. These trends mainly reflect China’s rapid growth of exports to advanced economies as it has become a hub of the region’s supply chain network for advanced economies.
For many Asian economies the share of medium- and high-tech exports in total exports exceeds the OECD average (about 57 percent), and for all economies in our sample, except Indonesia, it also exceeds the average for Latin America and emerging Europe (about 45 percent). The definition of medium-and high-tech goods is based on International Standard Industrial Classification (ISIC) (Rev. 2) classifications.
The estimated impact would have been even larger if the decline in U.S. import demand affected Asian imports disproportionately (about 2–8 percentage points of GDP under the extreme assumption that it was all concentrated on imports from Asia).
The declining share of household disposable income across Asia has been mirrored by rising corporate profitability and savings. This has been analyzed in the October 2009 Regional Economic Outlook.
Moreover, they do not control for changes in the efficiency of investment over time, although the fact that they are robust to a range of depreciation rates may in part address this issue. Excluding the Asian-crisis period, incremental capital- output ratios have remained relatively stable, averaging about 4.5 in the 1980s for emerging economies in Southeast Asia compared to 4 in the 2000s, suggesting that aggregate efficiency has also remained broadly stable.
This includes an indicator for governance from the International Country Risk Guide, which largely reflects investor perceptions of corruption in both the public and private sector. As such a worsening investment climate does not necessarily reflect an actual deterioration, but could well be the result of worsening investor perceptions that can be triggered by severe economic shocks like the Asian financial crisis.
However, rebalancing growth could entail short-term cost, including for employment. While short-term employment costs are not always present they likely depend on policies put in place to smooth the transition (April 2010 World Economic Outlook; Guo and N’Diaye, 2009b).
According to these simulations, U.S. output after three years would remain about 2 percent below the baseline if rebalancing of global demand were to focus only on emergin Asia (and about 3 percent below the baseline with no rebalancing efforts in Asia).
The valuation effects from currency changes may be overstated in the case of Australia and New Zealand because most banks’ foreign liabilities are hedged.
Measured by an import similarity index of more than 300 types of consumer goods, the consumer goods basket imported by China overlaps with that of other advanced economies by only about 35 percent.