Editor's note. Professor Michael Pettis lays out the finite possibilities for China responding to the fact that it cannot rely on US consumers in the future for its full employment project.
China’s debt problems have emerged so much more rapidly and severely this year than in the past that, combined with swirling rumors about the country’s leadership, a growing number of analysts believe that this may be the year that China’s economy breaks. As always, I am agnostic. There is no question that China will have a difficult adjustment, but it is likely to take the form of a long process rather than a sudden crisis. In this essay, I lay out, as methodically as I can, the dynamics that have driven the Chinese economy for over a decade.
[Michael Pettis | September 14, 2018 | Carnegie Endowment for International Peace]
When we consider the Chinese economy as a system, it turns out that the problems China faces are easier to understand than most analysts suggest. Like any deeply unbalanced economy or system, China’s imbalances must reverse, and one way or another, they will, but there are various paths by which this reversal can occur.
My July 2018 article in Barron’s sets out one case of how events force China to choose among these paths. If the global trade environment forces a contraction in China’s current account surplus, I argue, by definition it also forces a contraction in the gap between Chinese savings and Chinese investment. This means that either the country’s investment share of GDP must rise or the savings share must decline (or some combination of the two). There are literally only four ways that either of these outcomes can happen. Consequently, there are also only four ways that Beijing can respond, each of which would drive the economy to one of the four possible outcomes (or some combination of them):
- Raise investment. Beijing can engineer an increase in public-sector investment. In theory, private-sector investment can also be expanded, but in practice Chinese private-sector actors have been reluctant to increase investment, and it is hard to imagine that they would do so now in response to a forced contraction in China’s current account surplus.
- Reduce savings by letting unemployment rise. Given that the contraction in China’s current account surplus is likely to be driven by a drop in exports, Beijing can allow unemployment to rise, which would automatically reduce the country’s savings rate.
- Reduce savings by allowing debt to rise. Beijing can increase consumption by engineering a surge in consumer debt. A rising consumption share, of course, would mean a declining savings share.
- Reduce savings by boosting Chinese household consumption. Beijing can boost the consumption share by increasing the share of GDP retained by ordinary Chinese households, those most likely to consume a large share of their increased income. Obviously, this would mean reducing the share of some low-consuming group—the rich, private businesses, state-owned enterprises (SOEs), or central or local governments.
Notice that all four paths either raise investment or reduce savings, thereby reducing the country’s excess of savings over investment. This is what is meant by a contraction in the current account surplus.
MORE INVESTMENT MEANS A GREATER DEBT BURDEN
In terms of the first path, an increase in investment technically can occur in two forms. The additional investment can be productive. For that to be true, such an increase would have to result in a rise in debt-servicing capacity equal to or greater than the rise in associated debt-servicing costs. Or, to put it another way, the present value of the future increase in production generated by the investment must be equal to or greater than the cost of the investment. If that is the case, the increase in investment is sustainable and will not add to the country’s debt burden, even if it expands the country’s debt. By contrast, if the investment is not productive, it automatically increases the country’s debt burden. This means that either debt rises faster than debt-servicing capacity or that debt-servicing capacity is forced to decline, usually because productive resources were deployed nonproductively.
As an aside, if a nonproductive investment is not written down, it also results in what is effectively a capitalization of what should be an expense. In other words, an illusory increase in wealth is recorded. This is the main difference between money borrowed to fund consumption and money borrowed to fund nonproductive investment (aside from the fact that the former at least gives you temporary pleasure): neither increases wealth (that is, productive capacity) or income, but until it is correctly written down the latter allows you to report higher wealth and income that is wholly illusory. It would be as if you spent $100 on dinner and then rather than record an expense that is deducted from your total income, which would leave you $100 poorer after dinner, you instead record an asset, leaving your total wealth unaffected.
In China’s case, it is by now well agreed upon that a significant amount (and perhaps nearly all) of most public-sector investment in the past few years has been nonproductive; there has been no rise in the country’s debt-servicing capacity commensurate with the rise in debt. If Beijing were to engineer a further surge in investment, it is extremely unlikely that doing so would not result in an increase in the country’s debt burden.
BEIJING’S THREE OPTIONS
Consequently, if there is a forced contraction in China’s current account surplus, Beijing’s only possible response must involve one (or some combination) of three things: an increase in unemployment, an increase in the debt burden, or wealth transfers.
Beijing wants to avoid at all costs an increase in unemployment, which it has been able to do so far by maintaining high growth in economic activity. The Chinese national government also wants to rein in the growth in debt, but so far it has been unable to do so. Moreover, Beijing will not be able to rein in debt growth if it wants to avoid unemployment at least until it has managed a sufficient transfer of wealth to ordinary Chinese.
Unfortunately, rising debt is unsustainable and, at some point, Beijing will no longer be able to increase the debt fast enough to avoid a collapse in growth. It could take several more years before this happens, of course, but once it does the result will be an automatic surge in unemployment.
Put differently, over the longer run, Beijing can only avoid a sharp rise in unemployment or stagnant wage growth to the extent that it has managed to achieve significant wealth transfers. Beijing has engineered a small increase in the household share of GDP since 2012, but this has occurred far too slowly to have prevented a destabilizing surge in debt. Beijing must speed up the transfers, but it has had trouble doing so because of political opposition from so-called vested interests.
A LIMITED SET OF OPTIONS
To recap, all of the plausible policy choices available to Beijing are limited to one or some combination of the following three options: more unemployment, more debt, or more wealth transfers. This is because China (and indeed most economies) is limited to six economic paths, of which only three are plausibly available if Beijing wants to avoid surging unemployment:
- A rise in unemployment or stagnant wages. This occurs when an economy is unable to generate sufficient growth to maintain demand for workers. (The remaining five options, by definition, do generate sufficient growth.)
- A sustainable increase in investment. This would entail additional investment such that the growth in debt-servicing capacity exceeds the growth in debt. Although this option is technically open to Beijing, achieving it has been much easier said than done over the past decade. We can reasonably assume that Beijing is no longer capable of engineering enough productive investment to keep the economy growing fast enough to prevent a rise in unemployment or wage stagnation.
- An unsustainable increase in investment. This would mean an increase in nonproductive investment (in projects whose value is less than the cost of the investment), a choice that would worsen the country’s overall debt burden. China has followed this path for the past few years but may soon reach its debt limits.
- A sustainable increase in consumption. In China’s case, this would signify an increase in the consumption share of GDP that is driven by a corresponding increase in the household income share. (And this would likely further lead to an increase in sustainable private-sector investment.) This is the goal of Chinese rebalancing— effectively transferring wealth from elites, businesses, or governments to ordinary Chinese households—but achieving it has proven very difficult politically.
- An unsustainable increase in consumption. This outcome occurs when consumption growth is driven by rising household debt, which (obviously) would worsen the overall debt burden. China has followed this path for the past three years but may soon reach its debt limits. Coincidently, this path also seems to have been the main driver of U.S. growth for the past decade or more.
- A rising current account surplus. This option is only plausibly achievable for very small economies whose rising surpluses can be easily absorbed by the global economy.
These six pathways logically cover every possible option open to Beijing. If we exclude the second and sixth options as unrealistic, and if we acknowledge that the third and fifth choices both would lead to a rising debt burden, Beijing is effectively left with the same three aforementioned options as described in the Barron’s article: an increase in unemployment (option 1), an increase in the debt burden (options 3 and 5), or greater wealth transfers (option 4).
This is no coincidence. No matter what economic event China is reacting to or what Beijing’s goal may be, once we exclude the second and sixth options above, every policy Beijing could implement must lead to one or some combination of higher unemployment, higher debt, or greater wealth transfers. This is why Beijing’s only possible response to a forced contraction in the current account surplus is also limited to these options.
STEIN’S LAW: WHAT CANNOT GO ON FOREVER WILL STOP
Eventually, however, one of these policies will no longer be available—the rising debt option. If Beijing does not rein in credit growth in time, it will be forced to do so once debt levels reach the point at which debt can no longer rise fast enough to maintain the country’s targeted economic growth rate. This adjustment can happen quickly, in the form of a debt crisis. Or (what I think is far more likely, at least for now) it can happen slowly, in the form of what is subsequently called a lost decade (or decades) of slow growth, similar to what Japan experienced after 1990.
There is little more to analyzing China’s economic options. Unless Beijing suddenly becomes able to do one of two things that it hasn’t been able to do in more than a decade, its options for preventing a fall in growth will be limited. That is to say, unless Beijing discovers and channels funding to huge new areas of productive investment (even as global conditions deteriorate), or unless China grows its trade surplus by several percentage points of GDP every year, this struggle will remain. And until Beijing manages substantial wealth transfers that shift income from low-consuming entities to high-consuming entities, the only way China can prevent growth from dropping sharply is temporary, by way of an unsustainable rise in debt. Once debt can no longer rise, unemployment will rise or wages will stagnate.
SOME ILLUSTRATIVE EXAMPLES
It is worthwhile to consider how the following conditions or policy proposals would fit into this framework, the result of which is likely to be higher unemployment (via a slowing economy), more debt, or greater wealth transfers:
- China’s current account surplus contracts due to the trade war. If demand is reduced by a contraction in China’s trade surplus, either unemployment will rise or China must significantly increase the growth of its debt burden to make up for the contraction in net foreign demand (as it did between 2009 and 2011). Of course, the growth in the country’s debt burden will decline if the trade surplus expands. This is why the trade surplus is so important to China: it helps determine the growth rate of debt.
- Chinese SOEs increase dividend payments to the central government. Many economists have argued that an increase in the SOE dividend payout is key to China’s economic adjustment. But in fact, from a systemic point of view, it will have very little meaningful impact. SOE dividend payments represent a transfer of resources from SOEs to shareholders, mainly the government. As such, they mainly direct the focus of credit creation needed to generate economic activity. The greater the dividend payout, the more debt SOEs must create to generate any specific level of economic activity and the less the government must create.
- Beijing forces a reduction in credit growth. Because economic activity has become so heavily dependent on continued credit growth—whether that be household credit to boost consumption or corporate or local government credit to boost mainly nonproductive investment—a contraction in credit growth automatically causes a contraction in GDP growth. This is currently the problem Beijing faces today: in recent months, it has put substantial downward pressure on credit growth and, as a result, economic activity is slowing so rapidly that most analysts expect Beijing to panic, relent, and allow credit growth to pick up again before the end of the year.
- Beijing expands the Belt and Road Initiative (BRI).Foreign investment through the BRI can be seen mainly as a way of boosting the country’s trade surplus. To the extent that BRI financing causes an increase in net capital exports from China, it must also cause an increase in the Chinese trade and current account surpluses. As I explained in the first example above, the result will be that, for any given target level of GDP growth, China will be able to achieve it with a smaller increase in debt. The big problem, of course, is that to the extent Beijing has difficulty recovering the financing it has extended, the BRI represents a reduction in China’s net foreign asset position (that is to say that foreign assets decline relative to foreign debt). In that sense, Beijing is mainly exchanging domestic debt creation for foreign debt creation (or declining reserves).
- Beijing forces banks to increase lending to the private sector. This is proving to be much easier said than done, but to the extent that the private sector gains net new financing (that is to say, to the extent that the private sector does not merely use bank financing to pay down more expensive nonbank financing) and invests the proceeds, what matters is whether the new investment is productive or not. If it is, China will be able to maintain targeted growth levels with a smaller increase in the debt burden. If not, debt creation is simply transferred from the public sector to the private sector.
- CPI inflation rises. Rising consumer price index (CPI) inflation (which many in China believe exceeds the official numbers) undermines the real value of disposable income and so reduces the household income share of GDP. This means that rather than wealth transfers from businesses and local governments to households, we are seeing the opposite and, hence, a worsening of existing income imbalances. If that is indeed the case, because the sustainable consumption share of GDP will decline, it will take more credit growth than it otherwise would to meet the country’s GDP growth target.
- The RMB appreciates. Such appreciation has the opposite effect of rising CPI inflation. A stronger renminbi (RMB) effectively transfers wealth from the tradable goods sector and those who are long dollars (mainly the People’s Bank of China) to importers and those who are effectively short dollars (mainly households). By increasing the household share of GDP, and thus the sustainable consumption share, China can achieve its GDP growth targets with a smaller increase in debt.
- Chinese interest rates decline. Such declines have the same effect as rising CPI inflation and the opposite effect of RMB appreciation. Lower interest rates transfer wealth from net savers (mainly ordinary households) to net borrowers (mainly SOEs and local governments) and so worsen existing income imbalances. In this case, it takes more debt to achieve the GDP growth target.
- Chinese real estate prices plunge. A plunge in real estate prices would transfer wealth from those who are speculatively long real estate (mainly the rich) to those who are short (mainly the young and the poor). Apart from the potentially devastating impact on the banking system, this would cause a sustainable increase in consumption, although the adverse effect on consumer confidence could easily overwhelm this effect in the short term.
The point of these nine examples is to show that nearly any event or policy must fit into one of the six aforementioned economic options. In China’s case, this means (practically speaking) that Beijing must continuously choose between more unemployment, more debt, or more wealth transfers to the household sector. There seems to have been a serious attempt to rein in credit growth in recent quarters, but to the extent that it has been effective (and, in many cases, it has simply pushed credit growth from monitored areas to unmonitored areas), it seems to be showing up already in a weaker economy and rising unemployment. The official data do not record any slowdown in economic growth, but there is a spreading beliefthat the data this year seriously overstate economic activity (having long overstated real economic growth).
For the rest of 2018, I expect that we will see different groups in Beijing try to reconcile the need for slower credit growth with greater growth in economic activity. But because these two things cannot be reconciled, one group or the other must win. So far it isn’t clear whether we will see growth in economic activity continue to slow or credit growth pick up. The PMI data released on Friday suggested a stronger-than-expected pickup in manufacturing. One data point tells us nothing, but perhaps it’s a hint.
The students in my PhD seminar have suggested that, from a policy point of view, it might be better to further subdivide some of the six economic options featured in this post. This doesn’t change the underlying framework, but it allows us to evaluate policy options more easily. Wang Xiaotong, for example, suggested that consumption should be divided into household consumption and government consumption. The students also suggested that it might make sense to divide unsustainable increases in investment into investment in rising inventory and investment in nonproductive projects. This would leave us with the following nine economic options, of which only six are plausibly available to China if it wants to avoid surging unemployment or stagnant wages:
- A rise in unemployment or stagnant wages.
- A sustainable increase in investment.
- An unsustainable increase in investment in rising inventory.This, of course, can only be sustained by rising borrowing.
- An unsustainable increase in investment in nonproductive projects.
- A sustainable increase in household consumption. In China’s case, this would mean an increase in the consumption share of GDP that is driven by a corresponding increase in the household income share.
- A sustainable increase in government consumption. For China, this would mean increased government spending on consumer goods and services funded by the sale of government assets or by higher taxes on the rich or on businesses.
- An unsustainable increase in household consumption. In China’s case, consumption growth is driven by rising household debt.
- An unsustainable increase in government consumption. For China, this would mean increased government spending on consumer goods and services funded either by debt or by higher taxes on ordinary Chinese.
- A rising current account surplus.