Policy Complementarities and the East Asian Recovery

3.1 Memo on the East Asian Crisis

Although this paper is not intended to make a detailed description of the East Asian crisis of 1997–1998 or to discuss its mechanisms, it is useful, for the purposes of this paper, and before delving again into its core issue, to recall and highlight some key aspects of that devastating episode in Asia's recent economic past.⁹

As noted above, the economy in these countries seemed to work well. Growth was high and macroeconomic indicators were sound. The Asian economies had neither high budget deficits nor were pursuing expansionary monetary policies. Also, public debt was very low (even compared with many Organisation for Economic Co-operation and Development [OECD] countries)—for instance, public debt in Indonesia was 24% of GDP and it was 4% of GDP in Thailand.¹⁰ However, a more careful look at the balance sheets of banks and companies would have revealed a different situation, as the crisis was caused by an excessively large short-term external borrowing of the private sector.

Thus, in the years preceding the crisis, private capital was flowing into these economies at a high rate, while the relative importance of bank loans was very large and growing (Figure 1 [ PDF 104.7KB | 1 page ]). Accordingly, reserves were growing as well. The total amount of reserves in the four countries studied in this paper—Indonesia, Korea, Malaysia, and Thailand (hereafter referred to as the Asia-4)—was almost three times larger in 1996 than it had been in 1990.

There were several reasons why these private capital inflows were so high. First, economic growth was strong, which gave international investors more confidence. Fairly liberalized capital accounts made it much easier for domestic banks and corporations to finance domestic investments with foreign capital. Second, interest rates were significantly higher in these countries than they were in developed economies. And, third, nominal exchange rates were effectively pegged to the US dollar, which reduced the perceived risk to investors.

The "moral hazard" effect is another factor that should be considered in order to understand the more important aspects of the Asian crisis. That is, investors felt protected by either explicit or implicit guarantees. However, there is no consensus on the importance of moral hazard in the crisis. While some have held that the Mexican bailout in 1995—a US$50 billion rescue operation prepared by the International Monetary Fund (IMF) and the Clinton administration—created moral hazard on a global scale,¹¹ others have claimed that much of the lending was directed to non-bank enterprises and was not protected by any kind of guarantee. Nevertheless, it is indeed true that many loans in Mexico were not covered by guarantees; in addition, Mexico's package of loans was seven times its quota, or lending limit, at the IMF. The operation was unprecedented in size. Therefore, the assumption that a bail-out does not have to rely necessarily on an explicit promise or policy by the government is not an unrealistic one. If formal rules were broken before, why could they not be broken again?¹² Hence, it seems reasonable to consider moral hazard as an important ingredient in the complex mix of interlinked factors at the heart of the Asian crisis.

Moreover, the Asia-4's private debt was increasingly short-term. In seven years, the proportion of short-term debt, which was already very high, rose by almost eleven percentage points. The last two columns of Table 3 [ PDF 122KB | 1 page ] show that the ratio of short-term debt to international reserves increased very rapidly in only three years. In Indonesia, Korea, and Thailand the ratio was largely over one, which is potentially very dangerous (particularly if creditors decide not to roll over the debt). In sum, there were two sources of vulnerability: (i) domestic banks borrowed in foreign exchange and lent in local currencies, implying a greater exposure to losses in the event of depreciation; and (ii) these banks borrowed in short-term maturities and lent with longer payback periods, which implies a greater exposure to the risk of a run. The rest of the story is known—a quick listing of the facts of the events is found in Box 1 [ PDF 122KB | 1 page ] and Box 2 [ PDF 110.9KB | 1 page ].

According to Radelet and Sachs (1998), the existing macro and microeconomic imbalances were not strong enough to warrant a crisis of the magnitude seen in East Asia. They blamed a mixture of panic on the part of the international lenders, policy mistakes by local governments at the very beginning of the crisis, and poorly designed international rescue programs, for triggering a full-fledged financial panic and massive withdrawal of foreign capital,¹³ which deepened the crisis more than was either necessary or inevitable.

3.2 A Complementarities-Based Approach to some Aspects of the Asian Crisis

Not long before the crisis struck, Azis and Wescott (1997: 16) wrote prophetically, "We suspect that it is better to make policy progress on some fronts rather than do nothing, but national authorities must recognize that policy gaps in critical areas can cause improvements in economic growth to be imperceptible, and, in a worst case scenario, could cause problems if reforms are not staged carefully. In particular, countries that have liberalized their capital markets and that receive large amounts of foreign capital inflows must be careful to make sure that they do not backslide in other critical areas."

The Asian crisis can thus be regarded as a case of an abrupt end to an unbalanced or unsustainable growth trajectory, in the sense that some important—and complementary—key policy areas were not taken into consideration before the crisis.

An obvious lack of complementarity in the Asia-4 countries was apparent in the decision to liberalize their capital accounts—Williamson, Griffith-Jones, and Gottschalk (2003) classified the capital accounts in these four countries, as of the first half of 1997, as being "largely liberalized"—and to allow for strong credit creation without having put in place good supervision mechanisms. Banks operated in a weak institutional environment and, as a result, were under-supervised. What is more, many banks maintained very close relations with large companies and governments. And, if one includes the moral hazard effect in the decision-making process of banks operating in such an opaque context, one should expect the amount of expected over-investment to be even larger¹⁴. Put simply, "there were ample conditions for excessive risk taking, poor banking judgment, and even outright fraud" (Radelet and Sachs 1998: 16).

However, while this seems a rather straightforward remark to make, it was not until recently that many economists realized that institutions do matter. Institutions establish a stable structure for human interactions or incentives. But, in many cases, stability does not coincide with efficiency—i.e., bad institutions fail to align private incentives with social welfare; thus, poor institutions can generate incoherent policies that only serve to benefit networks of private interests, which will not coincide with (long-term) social interests. Importantly, these incoherent policies will possibly materialize either into low long-run growth rates (as may be the case in Brazil or the Philippines), or into high but unsustainable growth rates (as happened in the Asia-4 countries).

This sort of higher-level or primary incoherence—the lack of a good institutional environment, in broad terms—may then generate an endless number of policy inconsistencies. The Asian crisis is fertile in providing examples. For instance, the monetary response to the capital inflows between 1994 and 1996 was given mainly through sterilization (i.e., to counterbalance the effect of reserve inflows on the monetary base through open-market operations). However, there was an alternative available: to increase reserve requirements for banks (in order to reduce the money multiplier). Also, higher reserve requirements ensure that a bank will have enough money to cover bad loans, thus reducing its vulnerability. But, when the president's family owns banks—as was the case in Indonesia—and, moreover, there is the feeling that the economy is growing and that possible bad loans are somewhat guaranteed, why would the (non-independent) central bank want to put limits on the business of commercial banks? And, in fact, reserve requirements were imprudently low (not only in Indonesia, but also in Korea, Malaysia, and Thailand).

Another missing link is related to the growing specialization in non-tradable sectors. Figure 2 [ PDF 103.1KB | 1 page ] shows an important appreciation of the real exchange rate in the Asia-4 economies until the very beginning of the crisis (June 1997). This appreciation was caused by the appreciation of the US dollar—particularly since 1994—to which these countries' currencies were effectively pegged.¹⁵ Such bad specialization resulted in a loss of competitiveness, as well as a smaller capacity for those economies to attract reserves. The financial sector amplified this specialization as much of the credit was being directed to speculative investments in real markets (e.g., shopping centers and luxury office buildings), rather than to the exports sector.

As Radelet and Sachs clearly pointed out, the utilization of short-term foreign currency borrowing to finance domestic investments in real estate and other non-tradable sectors was dangerous. However, this inadequate allocation of capital in unproductive activities did not draw the attention of policymakers—not even that of the international community, namely the IMF. Nothing was done to counterbalance that detrimental specialization. If a country decides that, in a given moment, it needs a strong currency—in order to, say, attract foreign capital, fight inflation, and even to be able to buy capital goods and top-level education abroad—then it must put in place complementary policies that stimulate the investment and competitiveness of the export sectors and reduce the (short-term) attractiveness of non-tradable sectors.

A key feature in more coherent economic systems¹⁶ is the existence of good exit mechanisms. Good bankruptcy laws make crises less likely, or even prevent them from happening, and surely contribute to faster recoveries and sound growth.

According to the analysis of Radelet and Sachs, the crisis resulted from vulnerability to financial panic that arose from certain weaknesses in these economies, combined with policy missteps and accidents that triggered a full-fledged financial panic and massive withdrawal of foreign capital. In general, if creditors believe that they will not be able to recover what they have loaned to a bankrupt (or even only temporarily illiquid) borrower in a reasonable amount of time (e.g., because the bankruptcy laws are bad, de jure or de facto), then a creditors' run will probably occur, as each creditor will rush to be the first to demand full repayment.¹⁷ Furthermore, the absence of appropriate debt workout mechanisms that allow for the coordination of creditors will help to create a situation of panic.

Good bankruptcy laws are important in an (at least) equally important way: they do not impede plant dynamics and, therefore, lead to a faster and more efficient reallocation of resources, which is of particular significance in the context of a post-crisis recovery process.¹⁸

In recent years, various studies have shown that the resource re-allocation process from exiting producers to entering producers explains a substantial portion of total factor productivity changes at the aggregate level. Exiting producers exhibit persistently declining productivity, while entering producers that survive the market selection process exhibit rapidly increasing productivity. Thus, policies that prevent the reallocation of resources via entry and exit could be potentially very costly for the economy (see Lim and Hahn 2003).

Although this is an aspect that will be developed later in this paper, it is worth mentioning here that Indonesia, the country with the slowest recovery, and Korea, the country with by far the fastest recovery, have, respectively, the worst and best bankruptcy systems among the Asia-4.

Indonesia's bankruptcy law, drafted by the Dutch in 1905, remained unchanged when the crisis struck in 1997. Remarkably, in the nearly 50 years since the Dutch left the archipelago, Indonesia had never translated its bankruptcy law from Dutch into the native language. Declarations of bankruptcy were extremely rare in Indonesia. As a consequence of the non-use of the bankruptcy law, judges and lawyers lacked experience in bankruptcy matters (Walker 2000). On the contrary, Korea had a medium-quality bankruptcy system at the time of the crisis and improved it in the immediate post-crisis years.

A crucial piece in the intricate puzzle of economic coherence is the existence of social safety nets, namely in the form of a system of unemployment benefits. In its absence, the hardship imposed on people who lose their jobs and their families during a crisis is larger and, as a result, aggregate consumption and, consequently, aggregate output will decrease to a greater extent. Therefore, unemployment benefits act as an automatic stabilizer. In poorer countries, where households spend a very large proportion of their budget on food, the adjustment to a job loss situation is likely to be made through the reduction or elimination of education expenses, which is bad for human capital accumulation, and, thus, bad for growth; this is the scenario described by Chetty and Looney (2005). These authors compared large panel datasets on the US and Indonesia and came to find that the mean and median consumption drop associated with unemployment in both economies is roughly 10%. Such a similarity is remarkable, given that Indonesia has no formal unemployment insurance system, whereas the US insures 50% of the pre-unemployment wage for most individuals. However, in the Indonesian sample, the average household devotes nearly 70% of its budget to food, compared with 20% in the US. The authors examined the methods households used to mitigate the income loss associated with unemployment and found that, in Indonesia, parents appear to sharply reduce expenditures on children's education during idiosyncratic unemployment spells. Therefore, the welfare costs of transitory unemployment shocks, which are prevalent in developing economies, could be particularly large and long lived.

What is more, a system of unemployment benefits improves labor allocation, in the sense that it enables workers to find better jobs. In the specific circumstances of a crisis episode that suddenly signals the need to rapidly abandon a given specialization pattern—say, in non-tradable sectors like real estate—the existence of unemployment benefits can facilitate a worker's decision to abandon the non-tradable sector and try to find a job in a more dynamic sector or even to create their own small firm. Workers' resistance to change is lessened by the availability of unemployment benefits.

However, in the Asia-4 and other Asian "tigers," unemployment insurance was probably dismissed as superfluous because growth was high and unemployment rates were low (Lee 1998). It has been argued that the fiscal costs of implementing unemployment benefits systems in developing economies are prohibitive for those countries. But, as Lee pointed out, unemployment insurance is self-financing, with schemes based on contributions from workers, employers, or a combination of both—fiscal costs to the government need not rise unless the government chooses to subsidize the program. Lee made the point that at very modest levels of required contributions, the effects of unemployment insurance on labor costs and, hence, on demand for labor would be negligible. He added that International Labour Organization assessments show that if Indonesia, Korea, and Thailand had introduced unemployment insurance in 1991 (i.e., six years before the crisis' onset), an average required contribution rate of between 0.3 and 0.4% of payroll from 1991 to 2000 would have sufficed to provide 12 months of benefits for all insured job losers over this period, including during the crisis.

Once more, comparisons within the Asia-4 seem to be revealing: Korea, the country where the impact of the crisis was the smallest in terms of falling GDP per capita in 1998, and where the recovery was the fastest, was also the only country that had introduced before the crisis (in 1995) a relatively modest unemployment insurance scheme, which was expanded in 1998 in response to the colossal increase in unemployment resulting from the financial crisis.¹⁹ This aspect will be further developed in a later section (3.4.2.1) of this paper.

3.3 Deep Impact, Different Recoveries

This section gives a brief description of the different impacts of the crisis in the Asia-4 and their respective recovery speeds. While some other variables are mentioned in order to make the picture more complete, the main focus hereafter is on real GDP per capita (GDP per capita at constant year 2000 US dollars).

The hardest hit country was Indonesia. In 1998, Indonesia's real GDP per capita decreased 14.3% (see Table 1 [ PDF 173KB | 1 page ]). In Thailand and Malaysia, GDP per capita decreased 11.4% and 9.6%, respectively. Among the Asia-4 countries, Korea was the least affected, with a diminution of only 7.5%.

In Indonesia, one in every five formal-sector jobs was terminated in 1998, which left five million workers with bleak future prospects (Lee 1998). Poverty rates also grew across the region.

Griffith-Jones and Gottschalk (2006) justly stated that a key cost is forgone output. They have estimated the output loss for the Asia-4 at US$917 billion for 1997–2002. The largest losses in relative terms—that is, adjusted for the GDP sizes of their economies—were incurred by Thailand and Indonesia: 157% and 133% of their 2002 GDP, respectively. Malaysia incurred a loss of 69%. The loss in Korea was no larger than 26%. That simple ratio gives us a general impression of the depth of the crisis and, at the same time, the velocity of the post-crisis growth until 2002.

Figure 3 [ PDF 100.7KB | 1 page ] shows that it was only in 2004 that Indonesia finally recovered its pre-crisis level of output per capita. By then, GDP per capita in Korea was already 27% higher than in 1997; in Thailand and Malaysia, the GDP per capita was 12% and 9%, respectively. Interestingly, Malaysia experienced a faster recovery than did Thailand until 2001, but then Thailand overtook Malaysia and, by accelerating its growth in the sub-period 2002–2004, took second place in the run.

The average growth rate of GDP per capita between 1998 and 2004 in Korea was 5.1%. In Thailand, it was 3.8%. Malaysia and Indonesia both had the smallest average growth rates: 2.7%.

From these elements, a quite stylized picture can be sketched: Korea started and finished in first place; Malaysia started second but finished third, changing positions in 2002 with Thailand; Indonesia started last and reached 2004 in the same position. The main argument of this paper is that such a pattern is related to the coherence of the economic systems in these countries—that is, to the degree of complementarity of the policies that were adopted.

3.4 Coherence and Recovery in the Asia-4

3.4.1 Measuring Coherence

Trying to make a comprehensive portrait of a given economy and its evolution across a number of dimensions and years is not an easy task. There are data limitations: sometimes data do fit exactly in the concept that we want to see described; other times, there is simply not any data.²⁰

To create such a portrait, several data sources were used, namely: the very wide-ranging Economic Freedom of the World reports of the Fraser Institute, the Index of Economic Freedom of the Heritage Foundation, and World Bank data (World Development Indicators and Governance Indicators).²¹

There are 17 dimensions considered here; these can be further divided into three blocks:

1. a basic economic block (EB1), which represents roughly the somewhat typical receipt inspired by the Washington Consensus and has nine policy areas: liberalization of prices, less government intervention, stabilization, financial markets' openness and deregulation, ease of entry mechanisms, labor market deregulation, trade liberalization, access of nationals to foreign capital markets, and foreign access to domestic capital markets (i.e., ease of capital flows and attractiveness to foreign direct investment);
2. an extended economic block (EB2), which consists of adding to EB1 three key dimensions: exit mechanisms, social safety nets (specifically the existence of unemployment benefits schemes), and infrastructure; and, additionally,
3. an institutional quality block (IQB) with five dimensions: property rights, political stability, voice and accountability, control of corruption, and government effectiveness.

The above-mentioned data sources allow us to rank the Asia-4 countries for all EB1 and IQB policy areas, and also for infrastructure availability in EB2.²² Given the absence of readily usable data for unemployment benefits schemes and bankruptcy laws for the totality of the time period considered in this paper, these two areas were rated by me, using available data as well as information contained in several reports and papers. All elements were converted to a 0–10 scale.

As in Macedo and Martins (2008), I calculated, for each block, the respective RL (reform level—the simple average of the respective sectoral indicators Ri) and an index RC of reform complementarity (captured through the inverse of a Hirschmann-Herfindahl indicator). Therefore:

where N is the number of policy areas (for EB2, for example, N = 12). As before 2000, the Fraser Institute data, in its more detailed form, are available on a five-year basis, and so it was possible to calculate RC and RL for 1995. However, because it is important to isolate the immediate pre-crisis moment, I calculated RC and RL for 1997, using 1995 data to give a rating to five of the policy areas in EB1 (fully, in two cases, and partially, in the other three)²³. However, 1995 can also be regarded as a good snapshot of the pre-crisis situation, as policies did not change much between 1995 and 1997. RC and RL were also calculated for 2000, 2002, and 2003. RC was also converted to a 0–10 scale.

A country could have i.e., an economic system can be highly coherent, but have extremely market-unfriendly policies; one can think of autarchic, state-planned economies as an example. Conversely, RL can be high and RC can be low. So, it is appropriate to calculate an indicator that captures both the reform level and complementarity—a reform level indicator adjusted for complementarity:

Additionally, and having again in mind that institutions do matter, I also computed an indicator that intends to measure the general quality of a given policy environment:

²⁴

3.4.2 Looking at Data

3.4.2.1 Individual Policies

With a quick glance at the data (Table 4 [ PDF 417.7KB | 1 page ]), one can already highlight some aspects. Before the crisis, (direct) government intervention in those economies—measured by the relative importance of government enterprises and investment in total investment—was low (except for Malaysia). Inflation was low in all countries. Trade liberalization was high in Korea and Malaysia and medium in Indonesia and Thailand. Financial systems were only medially deregulated and open to competition; the respective indicator ranged from 4.5 (out of a maximum of ten) in Malaysia to 6.1 in Korea. Ease of capital flows was above 7.3 in Indonesia, Malaysia, and Thailand; only in Korea was that indicator not high, reaching no more than 4.7.

In 1997, Korea had the best infrastructure (6.3), exit mechanisms (4.5), and unemployment safety net (4.0). Malaysia ranked second and first (ex aequo) in the former two areas, respectively; on the contrary, Indonesia ranked last.

It is also easy to observe that Malaysia and Korea had the most proper institutional environment: property rights were well protected (6.8 and 7.8, respectively) and controls of corruption were the highest in the Asia-4 (6.5 and 5.2). Once more, Indonesia was in the worst situation: poor protection of property rights (3.9), poor control of corruption (3.1), and a lack of political stability (2.1).

It is interesting to notice that Malaysia—the only country that did not follow the IMF's prescription for crisis recovery—responded to the crisis in a very specific manner, clearly diminishing the importance of market mechanisms in its economy. Between 1997 and 2000, price controls in Malaysia increased, the labor market lost a great deal of its prior flexibility, regulations on economic activity increased (also in the financial sector), and restrictions on capital flows were imposed (conversely, Korea quite sharply augmented its ease of capital flows). Also in Malaysia, the degree to which property rights were protected decreased from 6.8 in 1997 to 5.3 in 2000 (meanwhile, in Thailand, property rights protection grew from 5.9 to 6.1).

As expected, by 2000, as compared to 1997, all the Asia-4 countries had increased their degree of trade liberalization. This was a logical way to profit from the strong devaluation of Asia-4 currencies that came along with the crisis.

Exit Mechanisms

After having seen in section 3.2 how crucial good exit mechanisms and social safety nets (namely in the form of unemployment benefits) can be in the context of a post-crisis recovery, it is important now to describe the main features of these two areas in the Asia-4 countries. The fact that the respective ratings had to be calculated by the author of this paper only reinforces the need to proceed—although very briefly—to such a description, for the sake of transparency.

As of 1997, Indonesia did not have even a minimally effective bankruptcy law (see section 3.2). In response to the crisis, the Indonesian Bankruptcy Regulation was amended in 1998 (this time in Bahasa Indonesia, the Indonesian language) and a special commercial court to handle bankruptcies was established. However, the revised Indonesian bankruptcy law was never implemented in a way that creditors viewed as transparent. People tended to avoid using the court system because it was considered expensive, unpredictable, and unreliable. There was a pervasive culture of corruption at all levels (Gamble 1998; Walker 2000; Ruru 2001).

By comparing the ease of closing a business²⁵ in 2005 in the United Kingdom (UK) (our benchmark in this domain) with the ease of closing a business in Indonesia in the same year, one has a basis for attributing a "grade" to the Indonesian bankruptcy system (see Table 5 [ PDF 142KB | 1 page ]). In the UK, the average duration to complete a bankruptcy procedure was one year, while in Indonesia it was 5.5 years; the cost of the process, as a percentage of real estate, was 4% in the UK and 18% in Indonesia; and, finally, the recovery rate (how many cents on the US dollar claimants—creditors, tax authorities, and employees—are able to recover from an insolvent firm) was 85.3 in the UK and 13.1 in Indonesia. Therefore, it is reasonable to attribute a grade 3 (out of 10) to Indonesia's system in 2003. The ratings from 1995 to 2002 reflect a supposed linear tendency of improvement (see Table 4 [ PDF 417.7KB | 1 page ]).

The Thai law, enacted in 1940, was also obsolete. Before the 1997 crisis, the liquidation of enterprises was so cumbersome and lengthy that creditors rarely obtained recovery—some bankruptcy cases in Thailand have continued for more than 20 years. Such difficulties were compounded by the lack of a specialized bankruptcy court. By the time judgment was secured, few, if any, assets of the debtor remained to be recovered. Not surprisingly, creditors rarely utilized the Thai bankruptcy regime (Walker 2000).

Legislation that put in place a new bankruptcy court and reformed bankruptcy procedures was enacted in 1998 and in 1999. These laws have given debtors and creditors the alternative of negotiating reorganization plans through the courts. These reforms seemed to be what the Thai economy was waiting for: in tandem with surge in court cases of corporate reorganization, bankruptcy cases also soared from 6,993 cases in 1998 to peak of 42,413 cases in 2002. That year also recorded the largest bankruptcy debt claims from creditors: 1.387 billion baht (Vongvipanond 2004). The comparison with our benchmark results in a grade of 4 for Thailand, an evaluation which suggests that even though some important reforms were successfully undertaken, other measures still needed to be taken.

Malaysia's bankruptcy law is based on English law and comes under the Bankruptcy Act of 1967. It was used with frequency before the crisis, and only a few changes were introduced after the crisis. According to our benchmarking exercise, the Malaysia bankruptcy system is graded 5 out of 10.

Korea made dramatic changes in its bankruptcy law in 1998 and 1999, achieving an almost state-of-the-art law in early 2005 (as reflected by its grade of 9). Lim and Hahn (2003: 11) pointed out that "the most crucial element in the post-crisis court-administered bankruptcy system was the court's establishment and tight enforcement of an economic efficiency criterion in selecting qualified firms for judicial bankruptcy procedures. Instead of basing the system on economic efficiency, the pre-reform system was based on high social value and prospects for rehabilitation.… The new criterion greatly contributed to removing the de facto exit barrier placed on large firms that had existed in the in-court bankruptcy system prior to the crisis." The authors showed that, prior to the crisis, "producers [in Korea] with persistently declining productivity were more likely to be accepted into a rehabilitation procedure as long as they exhibited "high social value" such as a large output or employment share in the economy" (Lim and Hahn 2003: 12).

Unemployment Protection

As of 2003 (the last year of our time sample), neither Malaysia nor Indonesia nor Thailand had unemployment benefits systems.²⁶ Korea established an unemployment benefits scheme in 1995 and improved it not only as an immediate response to the crisis (International Labour Office 2000),²⁷ but also in the following years.²⁸ It expanded the system to all firms, included temporary workers, shortened the compulsory contribution for eligibility, and extended the duration of unemployment benefits.

Korea has developed a medium-level unemployment insurance system much better than those of the other Asia-4 countries, but not as generous or wide-covering as in many EU countries, for instance. A very rapid, but still elucidative comparison helps to better understand this point. In Korea, the benefit is equal to half of the insured's wage earnings and it is payable for a period ranging from three to eight months (only for those with more than 10 years of enrollment in the system or aged 50 and over). In Portugal, for example, the unemployment benefit corresponds to 65% of wage earnings and its duration varies between one year and 30 months (for those aged 50 or more).²⁹ The way this policy area was classified is therefore consistent with this comparison, and also with the resolute tendency for improvement undertaken by Korea after the crisis. All the other Asia-4 countries were rated zero.³⁰

3.4.2.2 Indicators

By analyzing the computed indicators RC, RL, and ARL (see section 3.4.1) for 1997—a snapshot of the immediate pre-crisis situation—, it is possible to have a somewhat different look that can help us to understand why the crisis was deeper in terms of diminution of GDP per capita in some countries than it was in others.

The "do as much as you can" approach applied to EB1 is not very discriminating, but nevertheless would signal Korea and Indonesia as the "worst students," because their RLEB1 were the smallest among the Asia-4. However, the alignment between RLEB1 in 1997 and GDP per capita growth in 1998 was very low (Table 6 [ PDF 139.4KB | 1 page ]).

Conversely, ranking these four countries by their RLEB2, RCEB2, or ARLEB2 and the smallness of their GDP per capita losses in 1998 results in exactly the same order: Korea in the first position, then Malaysia, Thailand, and finally Indonesia. Quite interestingly, concerning the extended economic block (EB2), Korea had by far the highest level of complementarity. It was the most coherent economic system. That is, even if it was not as "market friendly" as Thailand and Malaysia, it was better complemented with good infrastructure, and also with a modest—but nevertheless existing—unemployment safety net and exit mechanisms.

Things become even more interesting when looking at the entire 1997–2003 period. In every single year between 1998 and 2002, Korea had the highest GDP per capita growth rate among the Asia-4. In the post-crisis years Korea was able to maintain very high levels of complementarity: its RCEB2 was 9.4 in 1997 and 9.3 in 2003, while its ARLEB2,(that is, its reform level adjusted for complementarity) increased from 5.3 in 1997 to 6.4 in 2003 (Figure 4 [ PDF 120.5KB | 1 page ]). Also, Korea was the only country that augmented its RLEB1 during the 1997–2003 period, which, in 1997, was the lowest among the Asia-4, and ended up being the highest in 2003. This suggests that the key to a faster recovery is not to retrocede on typical market-friendly policies, but to complement them with others. Implementing such policies can imply considerable amounts of public investment, but one should compare the long-term costs of these investments with those of not executing them; such costs could include stronger vulnerability to crises, deeper recessions, and slower recoveries.

On the other hand, Malaysia, as seen in section 3.4.2.1 on individual policies, reduced the relevance of market mechanisms in the functioning of its economy, as RLEB1 decreased from 6.3 in 1997 to 5.4 in 2000 and 5.3 in 2003. Also, and in a quite marked manner, Malaysia reduced its levels of complementarity, from an RCEB2 (RCEB1) of 8.3 (9.2) just before the crisis to 7.0 (7.3) in 2003.

In the late nineties, many economists seemed to rush to praise the somewhat unorthodox reaction of Malaysia to the crisis. It is widely believed that this nonconformist country was less hit than others in the region because it refused to fall under the tutelage of the IMF. Although the immediate interventions of the IMF were indeed inappropriate, I believe that such a vision misses a key point. The Malaysian economy reacted well to the crisis (when compared to Thailand and Indonesia) because it had a relatively more market-friendly environment and a more coherent economic system. However, in the subsequent years, Malaysia sharply reduced its ARLEB1 and ARLEB2 (and RLIQB, that is, the quality of its institutional context) because of reductions in both RC and RL. Thailand, on the contrary, adopted a different policymaking approach, keeping high levels of complementarity for EB1 and increasing RCEB2 to 8.4 in 2003 (Figure 5 [ PDF 88.6KB | 1 page ] compares the trajectories of the two countries). Before the crisis, ARLEB2 in Malaysia and Thailand was 4.53 and 4.07, respectively; by 2003, the scenario was the opposite, with 4.4 for the Thai economy and only 3.46 for Malaysia. In 2001—a bad year for the global economy—complementarity was already higher in Thailand; interestingly, the country then experienced no more than a slowdown in its GDP per capita growth, which continued to be positive (from 3.8 to 1.2%), while Malaysia experienced negative growth (-1.8%).

In fact, adding coherence to its economy seemed to have resulted in good dividends for Thailand; the Thai economy accelerated its growth and reached GDP per capita growth rates in 2002 and 2003 of 4.4% and 6.1%, respectively. In 2002 and 2003, growth in Malaysia was only 2.3% and 3.4%, respectively.

Indonesia was the country with the lowest real GDP per capita in 1997 (US$906, less than one-tenth of real GDP per capita in Korea), but it was also the country with the slowest recovery (it was only in 2004 that it recovered its pre-crisis GDP per capita level). Therefore, the Indonesian economy seemed not to have benefited from any kind of catching-up effect. The analysis also appears to shed some light on this. In Indonesia, ARLEB2 was always the lowest among the Asia-4 countries, and it did not grow from 1997 to 2003—in 2003, it was 3.2 (while in Korea, it was 6.4); furthermore, ARLEB1 decreased from 5.5 to 4.3. In short, Indonesia did not undertake significant market-friendly reforms (on the contrary, its RLEB1 decreased from 6.0 in 1997 to 5.0 in 2002, while Korea increased the same indicator from a similar initial level to 6.5) and did not improve significantly the already intrinsically low coherence of its economy.

In addition, Indonesia had very poor institutions. This represents a broad higher-level or primary incoherence, to which I make reference above (section 3.2), as it is reflected in a very low ARLIQB of 2.7 for 2000 (compared with 5.3 in Thailand, 5.4 in Malaysia, and 6.3 in Korea). There was also poor protection of property rights (3.0 in 2000, 7.1 in Korea) and, furthermore, political stability was practically nonexistent. Not surprisingly, the GQ indicator for Indonesia was the lowest among the Asia-4 countries in 1997 (3.1) and also in 2003 (3.2).

Remarkably, Figure 6a [ PDF 106.4KB | 1 page ] shows that ARLEB2 and GDP per capita experienced twin evolutions. Figure 6b [ PDF 106.9KB | 1 page ] is much less convincing in suggesting that RLEB1 can help to explain, to a similar extent, the different post-crisis growth trajectories within the Asia-4.

To sum up, this short comparative analysis provides suggestive evidence not only for the importance of preserving or achieving high levels of economic complementarity during a post-crisis recovery process (Korea and, to a smaller extent, Thailand are good examples), but also for how pernicious it can be to implement a nonconformist policy agenda that results, essentially, in a strong reduction of the relevance of market mechanisms in a given economy (the path chosen by Malaysia and Indonesia). There is no need to opt between a "do as much as you can" approach focused on an orthodox agenda and, on the other hand, a reflexive anti-market policy package (eventually, an "undo as much as you can" approach), which will be equally incomplete and myopic.

A faster and more correct and complete reallocation of resources is crucial to any fast growth trajectory, especially in a post-crisis context. In order to achieve that, market mechanisms, such as free prices or free financial markets, will play a key role. While this invisible hand should not be impeded, at least not in a persistent way, it must be complemented with a more visible hand—made of social safety nets, public investment in infrastructures, or transparent institutions, for example—which is to play an equally vital role.

The views expressed in this paper are the views of the authors and do not necessarily reflect the views or policies of the Asian Development Bank Institute (ADBI), the Asian Development Bank (ADB), its Board of Directors, or the governments they represent. ADBI does not guarantee the accuracy of the data included in this paper and accepts no responsibility for any consequences of their use. Terminology used may not necessarily be consistent with ADB official terms..

Comment(s)

There are [0] comment(s) for this entry. Post a comment.