Evaluation of Korea's Fiscal Policy Response in 2009

In this section, I evaluate Korea's fiscal policy response in 2009 in relation to the empirical analysis discussed above.

4.1 How big was it?

As in most other countries, Korea's fiscal balance deteriorated substantially in 2009 as a consequence of economic slowdown and discretionary policy responses by the government. However, preliminary estimates indicate that Korea's GDP growth rate in 2009 was not as bad as had been thought previously: it was slightly positive at 0.1%-0.2%.. While clearly lower than the country's historical average, the growth figure is higher than most people had initially predicted. In fact, according to my definition, the Korean economy was not even in recession in 2009. Since fiscal balance is normally expected to change systematically with economic activity, one may wonder whether the sizeable government budget deficit in 2009 can be justified by the GDP growth rate.

In order to evaluate the magnitude of Korea's fiscal response in 2009, I examine changes in the fiscal balance to GDP ratio after making cyclical adjustments to the series. In particular, I regress changes in the government budget balance to GDP ratio on GDP growth and examine how the residual changed over time.¹³ The correlation between the dependent variable (changes in the government budget balance to GDP ratio) and the independent variable (GDP growth) is about 0.53 in Korea for both the consolidated budget and the operational budget. Changes in the government budget balance to GDP ratio after cyclical adjustments are shown as a graph in Figure 6. As Figure 6 [ PDF 64.1KB | 1 page ] shows, the adjusted series is at its lowest in 2009, indicating that the sharp turn to government budget deficit in 2009 was more than can be warranted by the GDP growth rate. The expansionary change in 1998, in contrast, turns out to be fully justified by the low GDP growth rate at that time: the cyclically adjusted series is almost exactly zero for 1998. This suggests that Korea's fiscal response in 2009 was unprecedentedly large, especially when the severity of the economic downturn is taken into account.

Korea's fiscal response was large by international standards as well. When compared with other OECD economies in 2009, Korea appears to have been more keen to provide fiscal stimulus. The first graph in Figure 7 [ PDF 49.2KB | 1 page ] makes a comparison between OECD and Korea in terms of cyclically-adjusted changes in the government budget balance to GDP ratio, with the cyclical adjustment made in the same way as in Figure 6. To make an international comparison possible, fiscal balance here is defined as total outlays minus total tax and non-tax receipts of the general government as reported in the OECD Economic Outlook. Forecast values for 2010 and 2011 are from the same source. It is clear from the graph that the absolute magnitude of the fiscal response in 2009 was greater in Korea than in other OECD economies on average. A similar pattern is observed in the second graph in Figure 7 where I adjust for country-fixed effects instead of cyclical components in the fiscal response series. After adjusting for country-fixed effects, fiscal measures in Korea in 2009 are shown to be more counter-cyclical than the same measures in other OECD economies.

The large magnitude of Korea's fiscal response in 2009 can be associated with several reasons. First, as was clearly shown in the 1997-98 Asian crisis, concerns about exchange rate stability can substantially limit the scope for counter-cyclical monetary policy, especially in developing economies. With limited flexibility in monetary policy, fiscal policy should play a greater role in stabilizing the economy. Second, fiscal policy has relatively short outside lags. The long inside lag of fiscal policy can be reduced through the practice of supplementary budgets and front-loading of expenditures. These two measures have been particularly useful in Korea. Since 1990, only in two years, 1993 and 2007, was a supplementary budget not introduced. In 1991, 1998, 1999, 2001, and 2003, two supplementary budgets were introduced in each year. Also, the rate of implementation of fiscal spending was as high as 84% at the end of the third quarter of 2009. With such flexibility in its implementation and excution, a fiscal stimuls would naturally be favored by policy makers. Third, compared with other OECD economies, Korea has relatively small automatic stabilizers. This means that Korea needs to rely more on discretionary measures for stabilization. These explanations, however, do not provide an answer to the question of why fiscal the government budget deficit should be so large in 2009 compared with other years.

Perhaps the most straightforward explanation may be that the Lee administration has a strong inclination for expansionary policies. From the beginning of his term in 2008, President Lee has been trying to implement various tax cut plans and large scale public projects. Apparently, the global financial crisis provided a rationale for his expansionary stance.

4.2 How effective was it?

Another important issue in the evaluation of a fiscal response is the magnitude of the fiscal multiplier. A multiplier greater than 1 means that a fiscal stimulus can be particularly effective in giving a boost to the economy. A multiplier less than 1 means that there is a “crowding-out” effect: an increase in government spending causes other GDP components to decrease. While studies have found a wide range of values for the short-term fiscal multiplier, it is typically less than 1. For example, Barro and Redlick (2009) found that the defense spending multiplier is 0.6-0.7, substantially smaller than 1. OECD Economic Outlook (2009) provides a summary of existing studies in addition to its own estimates of fiscal multipliers. According to the report, the spending multiplier is smaller than 1 for most OECD economies, including Korea with an estimate of 0.8. It is larger than 1 only for large economies such as the US, Japan, and Germany. Existing studies in Korea also inidicate that the fiscal multiplier is positive but less than 1. Hyun (2009), for example, estimates that a 1 unit increase in fiscal expenditure in Korea leads to a 0.4 unit increase in GDP in the same year and a 0.17 unit increase in the next year. Huh (2007) and Kim (2007) each use a structural vector autoregression model and find that the fiscal multiplier is not significantly different from zero or even negative. Moon (2010) reports larger estimates: 0.9 for the expenditure multiplier and 0.6 for the tax multiplier.

It could be argued that the fiscal multiplier is likely to be larger in the current crisis because the effectiveness of a stimulus package depends on the severity of a recession. It is clear that the fiscal multiplier should be zero for an economy operating at full capacity. When there is slack in the economy, however, an increase in government demand would lead to an increase in output. At the same time, however, balance sheet effects and great uncertainty caused by a crisis can make individuals wanting to save more, thus lowering the fiscal multiplier. Thus, it is unclear what the net effect would be for the current crisis.

In Table 3, I provided a result that is closely related to fiscal multipliers. According to Table 3, a 1 percentage point increase in government consumption growth during a recession period leads to a roughly 0.3 percentage point increase in GDP growth during the first year going into the recovery phase. While not directly comparable to previous estimates of multipliers, the result suggests that the short-term multiplier effect of government spending may be rather substantial.¹⁴ Since government consumption amounts to about 15% of GDP in my sample, a 1 percentage point increase in government consumption growth corresponds to additional government consumption of about 0.15% of GDP. The coefficient of 0.3 in Table 3 thus implies that the short-term multiplier may be as high as 2 (=0.3/0.15). I also examine the medium-term effects of government spending, by extending the analysis in Table 3 and estimating the impact of fiscal policies on the growth rate in later years. I consider up to five years after the trough in my estimation and provide the results in Table 6 [ PDF 18.6KB | 1 page ]. The result for the first post-trough year is the same as in column (1) in Table 3. Results for the other years suggest that effects of government spending may be even greater in the medium-term than in the short-term.

It is not clear why my estimate of multiplier effects is so large. One possible explanation is that, in contrast to previous studies that examine all sample years, I only use recession periods in my estimation. As mentioned above, a demand stimulus can be effective only when there are spare resources. My approach may produce a greater estimate of fiscal multiplier because resources are underutilized during recessions. Another possibility is that the composition of fiscal spending may be structurally different between recession or crisis periods and ordinary periods. For example, capital injection by the government into troubled financial institutions can promote recovery through money and credit creation rather than the traditional Keynesian multiplier effect. Government consumption expenditure during a crisis period may serve as a proxy for the overall fiscal stimulus package that includes capital injection and other similar measures.¹⁵

4.3 Debt sustainability

While Korea's fiscal response in 2009 may have contributed to an early recovery of the Korean economy, it has also raised concerns about fiscal consolidation. As described in Section II, Korea's government debt to GDP ratio, although still relatively low by international standards, has been increasing very rapidly. In addition, it is widely believed that Korea has an unusually large number of “below-the-line” items that are not included in the official fiscal data. Some even argue that Korea's too narrow coverage of fiscal debt makes international comparison almost meaningless (Ok 2007). When the “below-the-line” items are included, Korea's fiscal debt may be substantially greater.

In this section, I assess the long-term sustainability of Korea's fiscal debt by providing a review of existing empirical studies on the issue. Since one of the most distinguishing features of Korea's government debt has been the rapidly increasing trend, most studies focus on the time series behavior, rather than the level, of fiscal variables. While these studies use various empirical methods including Bohn's test, non-stationary tests, cointegration tests, etc., most of them conclude that Korea's government debt position is sustainable. Bohn's test examines how primary balance changes in response to the government debt to GDP ratio in the previous period. If the primary balance improves whenever the debt to GDP ratio increases, government debt is regarded as sustainable. By applying Bohn's test to Korean data, Moon (2010) finds that a 1 percentage point increase in the debt to GDP ratio tends to be followed by a 0.1-0.15 percentage point increase in the primary balancce to GDP ratio. The esimated coefficient of 0.1-0.15 is not particularly small compared to estimates from other countries. Based on these results, Moon (2010) argues that Korea's government debt passes the sustainability test. Some other studies examine whether the debt to GDP ratio is stationary or whether fiscal expenditure and revenue are cointegrated with each other. As long as the debt/GDP ratio does not diverge, one can assume that fiscal debt is not increasing too fast. Similarly, as long as government expenditure does not deviate too much from fiscal revenue, it can be assumed that there is no persistent government budget deficit. Studies that follow these approaches typically find that Korea's government debt is sustainable.

Park et al. (2006) follow a somewhat different approach and focus on the level, rather than time series movements, of government debt. They first relate international differences in the levels of fiscal debt to various economic, demographic, and political variables. After adjusting for the effect of these factors on the magnitude of government debt, they find that Korea's debt is smaller than predicted by the model. This result also supports the view that debt sustainability may not yet be a serious problem for Korea. Another way to determine the sustainable level of government debt is to compute the debt level that is consistent with the ability to generate a primary balance surplus and growth adjusted interest rates. According to simple intertemporal budget accounting, fiscal debt as a % of GDP should be less than p(1+g)/(r-g) where p is primary balance as a % of GDP, g the growth rate, and r the interest rate. For example, given r-g = 4% and p = 2%, the sustainable level of government debt should be less than 50% of GDP. However, it is not clear whether Korea's government debt satisfies this criterion, because the calculation is quite sensitive to assumptions about the relevant interest rate.

Although existing studies tend to conclude that Korea's fiscal debt is sustainable, the sizeable deficit and the rapidly increasing trend in the debt to GDP ratio in recent years still calls for vigilance. As is well known, the government budget balance tends to move asymmetrically over a business cycle, with the deficit recorded during a recession not fully offset by the surplus during a boom. Consequently, an increase in government debt is rarely reversed. In Figure 8 [ PDF 27.7KB | 1 page ], I illustrate how government budget deficits translate into government debt over time using data from OECD. Figure 8 shows that, except for the years arount the Asian crisis, a 1 percentage point increase in the deficit to GDP ratio in a particular year was associated with a 4-5 percentage point increase in the debt to GDP ratio in 2007. Government budget deficits made during the Asian crisis period had an even greater impact on the government debt to GDP ratio in 2007. The lack of a trend in the series indicates that an increase in government deficit has an almost permanent effect on the level of government debt.

Download this Paper [ PDF 354.1KB| 26 pages ].

Comment(s)

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