(This post reproduces a thread I wrote on Twitter/X on July 29.)

A lot of the discussion about the depreciation of/undervaluation of China’s real effective exchange rate takes for granted that this is about the currency. After all, it’s called the “exchange rate”!

But it’s actually not. Mostly, it’s about prices. A rant:

China’s real effective exchange rate (REER) is a composite of two things, each themselves a composite: 1) an index of the renminbi’s exchange rate against a group of its trading partners, weighted by trade, 2) an index of China’s prices relative to its trading partners.

The standard reference for the REER is the calculation by the BIS. Its numbers show the pure currency part, the nominal effective exchange rate, is now right back at the recent peak of March 2022. The CFETS index, the trade-weighted exchange rate targeted by the PBOC, is not too different.

The REER, however, is down 14% from its peak in 2022. And that is purely, 100%, because of relative prices. Almost all of the move came in 2022-23, when China had a massive deflationary episode while the rest of the world had continued inflation.

So when people say, China’s real effective exchange rate has depreciated or is undervalued, the thing in the real world they are referring to is that prices in China (the CPI for this calculation) have stayed basically flat while rising elsewhere in the world.

Why has this happened? Ultimately, the domestic price level in an economy is under the control of the monetary authorities. This means that it is China’s own choice (mostly through inaction) to tolerate this persistent deflation/disinflation.

Many of China’s trading partners are understandably unhappy with this situation. China’s macroeconomic choices have had the effect of supercharging its export competitiveness: everything in China substantially cheaper relative to the rest of the world.

But there is debate over how to get the desired appreciation of the REER. Some people say that this issue could be solved if you could force China to simply revalue its nominal exchange rate, i.e. raise it enough to get rid of that 14% gap.

The problem with this idea is that a) you can’t force China to do something it doesn’t want to do, b) it’s not clear that a change in the nominal exchange rate would, in isolation, actually work to push up the domestic price level.

My view is that an appreciation of China’s REER is very much to be desired, but the real (haha) way to get there is through higher inflation domestically, and that means different macro policies.

Plus, if the markets recognized that China was actually doing more inflationary macro policy, the nominal exchange rate would appreciate, so you would get a double whammy of a rising price level and a rising exchange rate that would wipe out the REER depreciation.

The ultimate question, which no one has the answer to, is how to convince the Chinese authorities that different macro policies are required. I think complaining about trade surpluses and excess capacity has not worked, and is not going to work.

The key issue, in my view, is the domestic labor market, the true casualty of this tolerance of deflation. Policy decisions have condemned a generation of Chinese youth to widespread under- and un-employment. This is a political pressure point that might actually produce change.

If China would just pursue a conventional macro policy of moderate inflation and full employment, its own citizens would benefit massively and its trading partners would be much happier. There would be peace on earth and goodwill toward men! 🙂

Reading China’s five-year plans, one of the most immediately striking things is just how many things they attempt to plan. These are not just documents about where to build airports or highways (although such proposals do go into the plan). Recent iterations also cover many things that are not obviously amenable to top-down direction, such as what sorts of technological breakthroughs will happen in the future, and how social mores and customs will develop. In China’s Leninist political system, government attempts to specify the contours of “national economic and social development” are the norm rather than the exception.

In this context, restraint can be more notable than ambition. It’s not that surprising when plans declare extraordinary goals; it is more surprising when plans step back from attempted intervention. The 15th Five-Year Plan for the years 2026-30, now being drafted according to a set of “recommendations” approved by the Communist Party leadership in October, will certainly not be a laissez-faire document. Most media coverage has emphasized how the recommendations show China prioritizing industrial policy as it attempts to gain an edge in its geopolitical competition with the US. Here is a relevant section from the official English translation:

A modernized industrial system provides the material and technological foundations for Chinese modernization. We should keep our focus on the real economy, continue to pursue smart, green, and integrated development, and work faster to boost China’s strength in manufacturing, product quality, aerospace, transportation, and cyberspace. The share of manufacturing in the national economy should be kept at an appropriate level, and a modernized industrial system should be developed with advanced manufacturing as the backbone.

The obsession with advanced manufacturing as the avatar of the “real economy” is familiar from years of similar documents from China’s planning apparatus. But if we read this passage in the spirit of looking for what is missing, rather than what is included, one thing stands out: the language on the share of manufacturing. The 14th Five-Year Plan, when it was published in 2021, included a new goal to “maintain the basic stability of the manufacturing share” of GDP.

Previous plans had included goals to raise the service sector’s share of GDP, which at the time was seen as an indicator of the structural transformation and modernization of the economy. The 14th plan dropped those goals, and replaced them with the goal for stability in the manufacturing share of GDP. Rather than accepting that China’s economy would eventually look like everyone else’s–with the manufacturing share of GDP gradually falling over time–planners wanted China to stand out globally by avoiding the trend of de-industrialization.

It’s one thing to declare such goals, another to achieve them. The structural composition of an economy is not particularly easy to control: there is no single dial to turn that will boost manufacturing or scale back services. And in the event, the 14th plan’s declaration did not stop the manufacturing share of GDP from declining: it has gone from 26.6% in 2021, the first year of the plan, to 24.9% in 2024, and looks like it will tick down a bit more to around 24.8% by the end of 2025, when the plan period concludes.

This is not a large decline, and what counts “basic stability” is, after all, up to the government to decide. The Ministry of Industry and Information Technology, the agency responsible for this target, has already publicly declared that the manufacturing share is in fact basically stable. But the direction is clear: the manufacturing share of GDP has been in mostly uninterrupted decline since 2021. It spiked up during the post-Covid export boom of 2021, but has since resumed the downward trend.

The change in the wording for the next plan must therefore be read as an implicit admission of defeat: it was not actually possible to keep the manufacturing share of GDP unchanged, even with very generous subsidies for manufacturing investment and output. The next plan’s goal has thus been scaled back, from stopping the decline in the manufacturing share to just making sure the manufacturing share is “appropriate,” whatever that means. That this wording accepts some continued future decline in the manufacturing share of GDP was made clear in an explanatory article published in Seeking Truth, an official journal of the Communist Party leadership:

Traditional economics holds that a decline in the share of manufacturing and a rise in the share of services as the size of a country’s economy increases is a universal economic phenomenon. … The share of manufacturing in China displays an inverted U-shaped curve of “first rising and then falling,” which basically accords with the theory and practical regularities of industrialization. However, what we must be alert to is that this share must not decline too rapidly, nor fall to an excessively low level, as this could lead to serious consequences.

Among the main reasons that the manufacturing share of GDP tends to decline and the services share to increase in higher-income economies are that incremental household spending shifts from goods to services as incomes rise, and that changes in relative prices tend to make services (which tend to have slower productivity growth than manufacturing) a larger share of nominal GDP over time.

Both of these trends are a consequence of economic success, in terms of higher output and productivity: at least some of what is decried as “de-industrialization” is good actually. It is also good for China’s planners to recognize that reality and not distort the economy in pursuit of an impossible outcome. Obviously, a loss of export competitiveness would be a less positive reason for the manufacturing share to decline, but it makes more sense to focus on that issue directly than indirectly through an ambiguous indicator like the manufacturing share of GDP. Again, it’s not just significant what plans attempt to control—it is also significant what they do not attempt to control.

Will this change make a difference? Arguably, no. Since it’s not really possible for the government to directly control the relative shares of GDP of different sectors, it might not make much practical difference whether it has such stated goals. And there is still plenty of planning language about supporting the manufacturing sector. But it’s also arguable that changing the rhetoric will have effects. Technically, everything in the five-year plan has the force of law, and government officials are obligated to pursue the goals. The goal of maintaining the manufacturing share of GDP may have served as part of the justification for lots of specific policies that subsidized manufacturing output and investment.

There’s even a case to be made that those policies actually depressed the manufacturing share of GDP: by boosting output and capacity above where they would have been otherwise, subsidies probably contributed to the sustained decline in prices for China’s manufactured goods over the past few years, weighing on their share of nominal output. If subsidizing manufacturing output and investment is no longer so politically urgent, because there is no structural imperative to meet, then perhaps there will be room for those subsidies to retreat at the margin. That would certainly not be the outcome most outside observers expect from a five-year plan generally described as “doubling down” on industrial policy.

I recently wrote a piece at my job, and somewhat unusually for us, it’s come out from behind the paywall so that everyone can read it. Here’s the opening:

There are two statements about China’s economy that receive broad agreement today. One is that China faces a long-term structural slowdown in its growth rate, due to a laundry list of factors including the law of large numbers, changing demographics and less catch-up potential as an upper-middle-income country. The other is that China’s economy has recently been underperforming, suffering from extended deflation and weak employment, and would benefit from more aggressive cyclical stimulus. It’s not contradictory to believe both of these things, but there is some tension between them.  

The way people resolve that tension is, at least implicitly, by taking a view on the output gap–the difference between actual economic growth and its underlying potential. If you believe more strongly in the structural slowdown story, then you think the output gap is unlikely to be large, as the economy’s potential growth is steadily declining. If you believe more strongly in the current underperformance, then you think the output gap is large, and needs to be addressed. What’s curious is that while China’s government professes to believe that the potential growth rate of the economy is high (thanks to all of its wise industrial policies), it doesn’t act that way.

Whatever their expressed views might be, China’s policymakers are not acting as if they believe there is currently a big output gap. While fiscal stimulus has stepped up in 2025, the support from monetary- and property-policy measures has been less than expected. Official rhetoric is increasingly treating deflation not as a sign of deficient aggregate demand, but as a structural problem best dealt with through regulatory changes and industrial policy. Such interventions have stepped up since the July meeting of the Central Commission on Financial and Economic Affairs, chaired by top leader Xi Jinping, said the government would “regulate and manage disorderly low-price competition among enterprises in accordance with laws and regulations.”

Despite some public differences, it seems that both Chinese government officials and the staff of multilateral institutions hold fundamentally similar views: they believe more strongly in the structural slowdown of the Chinese economy than in the contemporary evidence of a large output gap. Financial-market participants, by contrast, tend to hold more strongly to the view that more aggressive stimulus is necessary.

I’m with the markets here: I believe in the output gap. The structural slowdown of the Chinese economy is an incredibly consensus view, and I believe in it too. But for a variety of reasons I think it makes more sense to stay agnostic about forecasts of potential GDP growth, and focus more on the signals the current performance of the economy is sending.

You can read the whole thing here.

In June I went to India for the first time. It was a quick trip, just a week, but I still got a pretty intense introduction to current debates on the Indian economy at the India Policy Forum in Delhi. As a China specialist with only a newspaper-reading level of familiarity with India, I was quite intimidated by the prospect of joining a roomful of people who have spent their entire lives working on tough economic questions of India. It turned out, though, that if you’re going into the equivalent of a high-level graduate seminar on Indian economics as an outsider, having a background in China is not the worst preparation. Almost every paper and every talk about India’s economic problems seemed to be motivated by some implicit comparison with China.

At first I thought I was over-interpreting things because I am biased to look at things from a China perspective, so I asked around to check my perceptions. The answer was: Yes, all this is in fact about China. It’s because the Indian elite has assumed for decades that India is destined to be the next global economic superpower after China, so the overriding question for them is why that hasn’t happened and what needs to be done to make it happen. The comparison with China seems to revolve around a number of generally accepted stylized facts, which form the basis for identifying policy issues and posing research questions. To me (as, again, an outsider to these debates), the key ones seemed to be around manufacturing, investment, human capital and state capacity.

One of the most obvious economic contrasts between India and China is that China has followed the East Asian model of export-led manufacturing, like Japan, South Korea and Taiwan (albeit at much greater scale), and India has not. All of those Asian economies historically had high manufacturing shares of GDP, and while in all of them the manufacturing share is lower today, it is still higher than in India. In recent years India’s manufacturing share of GDP has been around 13%, about half of the level in China, and historically it has never exceeded 18%. One of the central debates among Indian economists is how much of a problem the relative weakness in manufacturing is, and what should or could be done about it.

Historically, it is definitely true that manufacturing has been a key motor for structural transformation away from traditional agriculture to modern production based on wage labor and economies of scale. And the reason many Indian economists worry about the low manufacturing share is that it seems to be a symptom of slow structural transformation: the agriculture share of GDP and employment has remained largely static in recent years. So while the modern sector in India’s economy is clearly growing, it doesn’t seem to be pulling workers out of the traditional sector, limiting the income gains. That concern is behind proposals to encourage the growth of labor-intensive manufacturing.

India’s government has also set explicit targets for the manufacturing share of GDP: the “Make In India” initiative back in 2014 called for pushing up the manufacturing share to 25% by 2022, which obviously didn’t happen. NITI Aayog, the public-sector think tank that has replaced the old Planning Commission, is still talking about a 25% target, this time for 2047, the centenary of India’s independence. It is worth noting that China has, in fact, never set a quantitative target for the manufacturing share of GDP. Currently the government just talks about “maintaining basic stability” in the manufacturing share, which hasn’t stopped it from declining.

It’s also interesting that a recent comparative study of China, India, Indonesia, Mexico, and South Africa emphasizes that transitions out of agriculture played a limited role in raising household incomes out of poverty; income gains within sectors often played a larger role than reallocation among sectors. Which is not to say that more structural transformation of the economy would not be helpful for India, but perhaps it is not right to get hung up on the relative shares of different sectors. After all, there’s no economic theory that can tell India what its optimal share in GDP of manufacturing should be. You can’t treat shares of GDP like temperature readings, and use naive cross-country comparisons to say, oh, this level is obviously too low, there’s a deficiency, or this level is obviously too high, there’s an excess.

A closely related contrast between India and China is its relatively lower investment share of GDP. One assertion I heard, which received broad agreement, was that no country has been able to successfully develop without sustaining an investment share of GDP of over 30% for decades. This is based on the East Asian development experience: Japan, South Korea and Taiwan, as well as Malaysia and Singapore, all sustained 30%-plus investment rates for long periods of time, alongside their higher manufacturing shares of GDP. China of course, is the investment champion, for better and for worse, keeping its investment share of GDP consistently over 30% since 1992, and over 40% for the past 15 years. Again, China is implicitly (and often explicitly) the point of departure for the Indian debate. A World Bank report this year endorsed a “reform” agenda to push India’s investment share of GDP to 40% by 2035.

India’s investment share of GDP is actually not that low: it’s been around 30% in recent years, and was over 30% for roughly the period of 2005-13. True, this does not quite meet the (perhaps arbitrary) historical threshold for an East Asian-style investment boom. And some Indian economists express concern that the recent pickup in investment is aided by large public-sector investment in infrastructure, which may not be sustainable, and that private-sector corporate investment has been static or falling as a share of GDP. While corporate profits have been strong in recent years, they report that the weakness of private-sector investment seems to be concentrated in manufacturing: companies just do not see good prospects for adding manufacturing capacity. If India does not have a manufacturing boom, it probably will not have a broader investment boom.

An alternative view expressed by some other economists is that India is already growing rapidly with a 30% investment rate, the current returns on services investment are high, and it’s not obvious that it needs to push a lot more investment. I found myself sympathetic to this side of the debate. China today, after all, is hardly a model to follow: it has experienced massive excess capacity and declines in returns on capital since the 2008 global financial crisis, largely due to forced public-sector investment.

Still, China’s current excess investment is a product of both its peculiar quasi-socialist institutions and particular historical circumstances, neither of which have very close analogues in contemporary India. If India can find productive uses for a few more percentage points of GDP worth of investment, that would probably be welcome, and would hardly put it on the path to Chinese-style excess. But equally, it’s hard to take a simple cross-country comparison as sufficient justification for massive macroeconomic interventions to push up the investment share.

A more supply-side perspective on India’s relative dearth of manufacturing and investment focuses its failure to establish the necessary social preconditions for an East Asian-style growth takeoff. As Wang Feng’s recent book China’s Age Of Abundance emphasized, China was able to seize the opportunity presented by labor-intensive export manufacturing in the 1980s because it had made substantial improvements in health, education and gender equality in prior decades. China’s workforce already had the necessary human capital to transition effectively to jobs in the modern economy, when reforms started to make those jobs available.

In terms of some basic measures of human capital, India has yet to catch up to where China was decades ago. According to the World Bank, India had an average adult literacy rate of just 77% in 2023–roughly China’s level in 1990. This average gap is aggravated by the scandalously low female literacy rate in India, just 70%. (Other measures like average years of schooling also show a gap). One reason why manufacturing has not taken off in India may well be that India does not have a mass workforce equipped with the basic skills to perform manufacturing jobs.

The low average endowment of human capital in India is striking given the high achievements of the most educated: India obviously does have a workforce equipped to perform high-skilled jobs in information technology and services. A recent paper by Nitin Kumar Bharti and Li Yang collects a remarkable amount of historical data to show substantial differences in educational priorities between India and China. After independence, India invested more in university education but neglected primary and secondary education, producing a relatively small cohort of skilled graduates amid a large uneducated and illiterate population. China, by contrast, initially focused on achieving broad-based basic education before emphasizing the expansion of university education after the 1980s.

Bharti and Li document that India has since the 1990s tried to expand enrollment in primary and secondary education, which is reflected in rising literacy rates. But the quality of basic public education still seems poor, and more education spending is not always producing better results. There are lots of shocking stories about widespread teacher absenteeism and incompetence. It is generally reported that higher-income Indians will do whatever they can to stay out of the public education and healthcare systems. This points to an even more fundamental contrast between India and China, which is in state capacity: the government’s ability to get things done.

In particular, India’s government seems to struggle to deliver basic public services and provide a baseline level of human welfare. Another fact about India that is, deservedly, often repeated is its alarmingly high proportion of underweight and stunted children. The last survey data collected by the World Health Organization show that 36% of young children in India were stunted (far too short for their age) in 2020. That’s similar to the proportion in China in the mid-1980s; but China at the time had per-capita GDP of only a third of India’s today. An interesting case study of Karnataka, one of India’s most developed states and home to the IT hub of Bangalore, reinforces the point. While Karnataka has had one of the best growth stories in India, it has still a higher proportion of children who are stunted, underweight or not enrolled in school than some poorer states. The significant increase in economic resources available to the state government does not seem to have translated into better outcomes for much of its population.

One set of these comparisons and stylized facts, those on manufacturing and investment, are frequently used to argue for more aggressive efforts to transform India’s economy and put it on something more like a Chinese (or East Asian) trajectory. But the other set of comparisons, on human capital and state capacity, implicitly counsel caution. If India does not have the necessary preconditions for a manufacturing takeoff, companies probably would not respond strongly to government favoritism for manufacturing (and indeed most people seem to think the response to the Make In India initiative has been underwhelming). In any case, a government that struggles to deliver basic public services is unlikely to be able to execute interventionist industrial policies effectively. Even China has plenty of waste and corruption scandals alongside the success stories.

Yet it is also worth emphasizing that India does not come off the worse in all comparisons with China. It’s an important fact that, since the pandemic, India has overtaken China to become the world’s fastest-growing large economy, averaging real GDP growth of over 7%. This is all the more remarkable given that India clearly is not having an East-Asian-style boom in manufacturing and investment. Therefore, it must be having some other kind of boom: one led by services. The IT outsourcing wave, which got started in 1985 when Texas Instruments opened a research center in Bangalore, seems to have stepped up to a new level in the last few years, possibly boosted by the post-pandemic changes in working patterns. Foreign investment clearly plays a big role in this process, and I saw huge swathes of new office parks going up outside Delhi, adorned with multinationals’ names.

Given that, by general consensus, India has not solved its outstanding problems in agriculture and manufacturing, the fact that the services boom is strong enough to power 7%+ aggregate growth is pretty impressive. It’s hugely important that India now has a self-reinforcing growth cycle in foreign and private-sector investment and exports. It’s almost reminiscent of what happened in China in the 2000s after its WTO entry, even if India’s cycle is mostly in services, which have fewer spillovers to the rest of the economy than manufacturing. If there is an alternative school of thought to the one focused on trying to “be like China” in terms of macro aggregates, it is that India should focus on building on what is already working. That means not just facilitating the tech boom, but also making more sectors outside IT services attractive to investment from both domestic and foreign businesses, and trying to steadily improve state capacity and public services.

Incremental reforms might be more politically realistic than trying to drive a “big push” in manufacturing investment. Most of the commentary I heard in Delhi was pessimistic about the outlook for big policy changes, since the current fast GDP growth sends a signal that things are going well, and makes disruptive and politically costly reforms seem less necessary. But it is also true that evidence-based comparisons with China should be able to motivate lots of different ideas about what India can do: China itself has gone through lots of changes, and the growth model has been quite different at different points in time. There are a lot of people these days who think “being like China” means “doing manufacturing-heavy industrial policy,” but this is a perspective distorted by the particular set of government priorities since 2015 or so.

China’s turn to aggressive industrial policy came only after it had already become a quite successful export manufacturer. And that turn was explicitly justified on grounds of national security rather than promoting growth. It was precisely because China was already quite developed that it could afford to put lots of money into import substitution and speculative technological bets. The industrial-policy apparatus of today’s China is a completely inappropriate model for India, which faces a very different set of problems with a different set of capabilities. A focus on removing impediments to a private-sector-led expansion would in fact make India more like China–just the China of a different, less statist era.

The reason fiscal policy is interesting is that it is the concrete expression of a country’s political priorities: how governments spend money tells you how they work and what their priorities are. By the same token, it is impossible to really interpret fiscal policy without some understanding of the political structure in which the government operates. There are a lot of major differences in political structure–to put it mildly–between the country I grew up in, the US, and the country I have spent my professional life in, China. One of the ones whose importance took me a while to figure out is that China is a unitary state while the US is a federal state.

This issue springs to mind every time I read assertions like this about China: “The ratio of central government debt or sovereign debt to GDP is a mere 21 percent, the lowest among the world’s major economies.” (I’m quoting from the most recent example to have arrived in my inbox, but this argument is so widespread that I don’t need to pick on anyone in particular.) It is indeed true that, if you look at statistics on government debt in China, the majority of it is assigned to local governments rather than central governments (see chart). Many people therefore contend that while local governments have strained balance sheets and limited capacity to borrow further, the central government does not.

Such a distinction would make sense if China were a federal state: if the central and local governments were independent entities with clearly defined constitutional and legal roles and separate finances. But China is not a federal state, and local governments are not separate from the central government. There is only one government throughout China; local governments are merely the authorized agents of this state. There is no constitutional or legal support for the idea that China’s local governments have any independent fiscal power and could ever be considered as having balance sheets that are separate from the central government.

It’s true that China’s government is often deliberately obscure about its true organization and structure. But some things are out in the open. Let’s turn to the Constitution of the People’s Republic of China:

Article 105. Local people’s governments at various levels are the executive bodies of local organs of State power at various levels and are the local organs of State administration at the various levels.

Article 110. Local people’s governments at various levels throughout the country are all organs of State administration under the unified leadership of the State Council and are all subordinate to the State Council.

That’s pretty clear. There is only one government in China, and the State Council, meaning essentially the executive, controls that government. As a matter of constitutional theory, local governments have power and authority only because the State Council gives it to them. The fiscal implications of the unitary state are also clearly spelled out in the Law on the Administration of Tax Collection:

Article 5. The competent department for taxation under the State Council shall be in charge of the administration of tax collection throughout the country. The national tax bureaus and the local tax bureaus in various places shall administer tax collection respectively within the limits set by the State Council. 

The local people’s governments at various levels shall strengthen their leadership over or coordination of the administration of tax collection within their respective administrative regions, and support the tax authorities in performing out their duties in accordance with law, calculating the amounts of taxes to be paid according to the statutory tax rates and collecting taxes in accordance with law. 

This is also quite clear. The authority to raise tax revenue lies solely with the central government. Again as a matter of constitutional theory, local governments’ job is to help implement the tax policies decided by the central government. They do not have any authority to raise taxes on their own, and have no independent sources of revenue.

If you look at local government budgets in China, there are two sources of funds: revenue “at the local government level,” historically 55-60% of total revenue, and transfers from the central government, historically 40-45% of the total. In reality, though, these are the same thing: money from the central government. The central government allows local governments to retain a share of the taxes that are collected at the local level. The remainder that is handed over to the central government, and the taxes collected at the central level, are then redistributed back to local governments as transfers. Ultimately all tax revenue is controlled by the central government, which decides how much money local governments get.

What’s surprising is that, in a unitary state, there could even be such a thing as local government debt. And indeed before 2010 there was not. The local government bonds that have been sold since then are a curious thing: the central government approves their issuance, so the local governments do not have any independent borrowing authority. And the central government controls how much revenue local governments have, so whether local governments can repay the debt still ultimately depends on the central government. This is a weird arrangement. The central government could just issue the same amount of money in treasury bonds and redistribute it to local governments. But for internal political reasons that I honestly struggle to understand, it is considered desirable that some of this debt be “in the name” of local governments. Even though local governments do not have any authority to raise revenues to repay the debt!

Because of this unitary structure, it makes no sense to split either the income statement or the balance sheet of China’s government between central and local entities. There is only one state in China and its finances are unified. I should point out that an excellent recent quantitative overview of China’s government balance sheet by the IMF, “Fiscal Policy and the Government Balance Sheet in China,” does not fall victim to the fallacy I criticize; the authors follow best practices by presenting their assessment on a “general government” basis, i.e. the combination of central and local governments. That’s the right way!

What all this means is that you cannot wave away the debts of local governments by saying they are local, and it’s only central government debt that matters. The central and local governments are part of a single unitary state, and the central government is in charge of figuring out how to repay all of its debt. This is precisely why it has been such a disaster for the central government to allow local governments to engage in all that uncontrolled off-balance-sheet borrowing. If China was a federal state, local fiscal incontinence wouldn’t matter too much ultimately. When local authorities borrowed beyond their means, they eventually wouldn’t pay it back, and it would just be an issue between them and their creditors without national implications. But because China is a unitary state, every time local authorities abused their power by creating unauthorized liabilities, they worsened the fiscal situation for the entire state.

Moral of the story: a unitary state needs to act like one, not pretend to be a federal state. A combination of centralized revenue-raising authority and decentralized liability-creating authority is the worst of both worlds, and the sooner China gets away from it the better.

P.S. My thanks to the excellent NPC Observer website for making it easy to track down the relevant laws.

The usual goal of industrial policy is to, by supporting or protecting a particular industry, allow it to more quickly achieve the economies of scale necessary to be competitive. China has plenty of experience with this type of industrial policy: witness how it has scaled up in the manufacturing of solar power, lithium batteries, and electric vehicles to a size that dominates global markets. But the official rhetoric on industrial policy is now turning to a different and less well-trodden path: pursuing economies of scope as well as scale. Or, to put it in less technical terms: China’s government wants to preserve and improve competitiveness in a wide range of different industries, not just specialize in the most profitable ones. It sounds like a simple change, but it’s a significant one for China’s trading partners.

As is usual in China today, the signal of this change in priorities has been delivered by the man at the top, Xi Jinping. At the May meeting of the Central Commission on Financial and Economic Affairs, one of many steering groups he chairs, Xi laid out his vision of a “modernized industrial system.” He explained that such a modernized system has three key characteristics: it is “complete” ( 完整, also translated as “comprehensive”), “advanced” (先进) and “secure” (安全). The most novel of these objectives is “completeness,” and Xi briefly explained what that objective means in practice: “We must keep promoting the transformation and upgrading of traditional industries, and not take them as ‘low-end industries’ to be simply eliminated.”

On its surface, this could sound like one of the invocations of the importance of blue-collar manufacturing jobs now common in other countries on both the populist right and left. In fact, this statement is an intervention in a specific Chinese policy debate that clearly indicates a reversal of the previous direction. China’s turn toward high-technology-focused industrial policy well predates Xi, and actually began in the prior administration of General Secretary Hu Jintao and Premier Wen Jiabao (see Barry Naughton’s excellent short history of industrial policy for more).

In the mid- to late 2000s, “upgrading the industrial structure” was a regular buzzword, and official support for “emerging” and “strategic” industries ramped up. At the same time, though, the government tried to restrict resources going to less-desirable industries, usually defined as those that are highly polluting, energy-intensive, or in excess capacity. These efforts reached a peak in 2008; here is some representative language in Wen’s government work report from that year:

It is essential to appropriately control the scale of fixed asset investment and improve the investment structure. We will maintain strict control over the availability of land, credit and market access, and pay particular attention to strengthening and standardizing supervision of new projects to ensure they meet all the conditions for launching. Haphazard investment and unneeded development projects in energy intensive and highly polluting industries and industries with excess production capacity will be resolutely stopped, and market access will be tightened and capital requirements will be increased for industries whose development is discouraged. Work on illegal projects will be resolutely stopped.

In this vision of industrial upgrading, there is both positive and negative discrimination: the government pursues policies that favor high-end industries and disfavor low-end industries. Such a vision is based on ideas of national specialization and integrated global trade. It’s influenced by the “flying geese” model originally articulated by a Japanese scholar and popular among development economists in the 1980s and 1990s. The idea is that, as nations advance up the technological ladder, they leave behind production of low-value or commoditized products, which then creates opportunities for lower-income countries to industrialize by making those goods. Nations at different income levels specialize in making different kinds of products, and by trading with each other everyone becomes better off.

Xi’s vision of “completeness” or “comprehensiveness” does away with this. Rather than allowing China’s low-end industries to shift to other, lower-income countries, and then importing those products, the idea is to maintain the ability to produce the full range of goods within China. Low-end industries are not abandoned but become targets for technological upgrading in order to preserve their competitiveness. As the Chinese economist Xu Zhaoyuan explained in an exegesis of Xi’s remarks:

We cannot allow traditional industries to transfer abroad too quickly. This requires continuous strengthening of policy support for the upgrading and transformation of traditional industries, improving product innovation and efficiency to enhance their competitiveness, while also continuously reducing the cost burden of the real economy, especially reducing various transaction costs. 

The background assumption is clearly no longer that nations can specialize to take advantage of an open global trading system, but rather that they need to minimize external dependencies and vulnerability to trade disruption. The Chinese economist Yu Yongding called it part of China’s response to the US decoupling campaign:

Re-emphasizing the importance of comprehensiveness is a reaction to the new geopolitical reality. While China cannot and should not produce everything – autarky is impossible for a modern economy – it should be able to quickly launch or increase production of critical goods, as needed.

China does indeed have a very broad range of manufacturing competence: according to Yu, “China ranked among the world’s top three exporters (by volume) in 2,400 of 4,000 categories of intermediate goods traded globally between 2017 and 2020.” More simply, there are very few goods that China does not make. I looked at China’s exports broken down by 4-digit HS code; out of 1,241 categories, there were zero exports in fewer than 50, a share that has remained largely constant over the past decade.

In a sense, then, achieving a “complete” industrial industrial system would just mean maintaining the status quo. On the other hand, as China is probably only a few years away from qualifying as a “high-income” country on the World Bank’s definition, one might expect that rising incomes would have at least some impact on China’s cost structures and competitiveness in making in different products.

As is often the case with these high-level slogans, it’s not totally clear what the practical implications of Xi’s policy shift are going to be. It would not be that unreasonable for Xi to say, “I don’t think we should have government policies that actively try to shut down particular industries, because those industries employ Chinese people and earn money and there’s just no good reason to get rid of them.” It would be somewhat less reasonable, certainly from the perspective of China’s trading partners, for Xi to say “Instead of just subsidizing the high-tech industries of the future, I think we should subsidize every single industry that China has so that China can have a comparative advantage in making everything.”

At the least, the rhetoric of “completeness” does not offer a lot of hope that other countries are going to be able to benefit from China’s growth by selling it stuff. The “flying geese” theory is now criticized, not totally unfairly in my view, for contributing to the hollowing-out of industry in the US and other high-income countries. But it did offer a basis for mutually beneficial trade with the developing world: as high-income countries lost competitive advantage in some industries, the low-income countries gained it, and could use those industries to raise their own incomes.

China these days is trying to knit together a coalition of other developing countries also opposed to US dominance of the global system. But its official economic theory does not offer much of a basis for what it likes to call “win-win” ties with other developing countries. China wants to keep producing itself all the stuff that poorer developing nations in Africa, Asia and elsewhere might want to sell it. China acknowledges a need to import necessary raw materials, and that’s about it. Is this maximum mercantilism really an attractive vision for an alternative global economic order?

As China’s data continue to disappoint, there is a persistent theme in much of the outside commentary on its economic woes: that China is for some reason failing to take the “obvious” step of sending stimulus checks to households. The implicit argument is that the US handed out massive subsidies directly to households, got a great post-pandemic recovery and everything turned out fine. China did not deliver subsidies to households, and that’s why everything is very much not fine.

Why is China, still, not taking this course in spite of the positive example of the US? Of course, an obvious answer is that the people in charge don’t think the US example was that positive, and anyway aren’t particularly prone to think of the US as a model to emulate. The merits of the pandemic fiscal-policy response are still pretty contested in the US. Plenty of people think of the stimulus as a fiscally irresponsible gamble that ultimately had pretty disruptive macroeconomic effects, because of the huge interest-rate increases that were required to bring inflation under control; my perception is that many people in China share these views. Still, the case for stimulus is probably getting stronger rather than weaker at the moment as China falls further away from potential growth and full employment.

My suspicion is that part of Chinese policymakers’ reluctance to use direct transfers to households as a short-term stimulus stems from a fear of setting a fiscally destabilizing precedent. If debt-financed transfers failed to generate a sustainable recovery, the money would be wasted. But if transfers succeeded in generating a good burst of growth, that could have even bigger longer-term effects. It would mean that, the next time China falls short of potential growth and full employment, the political pressure to roll out household transfers again would be overwhelming. What started as a one-off policy response could become entrenched as the expected response to any growth slowdown, and would add to government deficits and debt over many years rather than just one.

Why should this hypothetical possibility be a serious concern for Chinese policymakers? Because it is exactly what happened after the 2008 financial crisis. An unprecedented global shock led to an unprecedented policy response, as the central government encouraged local authorities to use off-balance-sheet borrowing to fund a wave of public works projects. If that had been a one-off measure, it would have been a powerful example of effective and unorthodox policymaking. Instead, the infrastructure stimulus institutionalized fiscal irresponsibility on a massive scale: a decade and a half on, the off-balance-sheet local borrowing is even bigger relative to the economy than it was in 2008, according to IMF estimates. It would be hard to find a better example of Milton Friedman’s quip that “nothing is so permanent as a temporary government program.”

China’s political system therefore does not have a good track record of being able to take away the punch bowl in good economic times in order to be able to share out more punch in the bad times. The dubious legacy of the 2008 stimulus means that China now needs fiscal consolidation to get long-term debt dynamics under control–at exactly the moment that it once again faces a shortage of aggregate demand.

Striking the right balance between the structural and cyclical issues is indeed quite difficult. Maybe the right answer is in fact that the cycle needs more attention at the moment, because a failure to address the loss of growth momentum would allow hysteresis to set in and create even more long-term costs for the economy. But it is perhaps understandable that the people inside China’s system are reluctant to experiment with new forms of fiscal stimulus before they have gotten the old ones under control.

What accounts for the extraordinary rise of China’s private sector after the economic reforms that followed Mao’s death? The 1980s were a pivotal decade for China in many ways, as the rapid growth of the private sector transformed the structure of a still officially socialist economy. In 1987, Deng Xiaoping famously said that the explosion of private-sector activity in the countryside in particular came as a surprise to him:

In the rural reform our greatest success — and it is one we had by no means anticipated — has been the emergence of a large number of enterprises run by villages and townships. They were like a new force that just came into being spontaneously. …If the Central Committee made any contribution in this respect, it was only by laying down the correct policy of invigorating the domestic economy. The fact that this policy has had such a favourable result shows that we made a good decision. But this result was not anything that I or any of the other comrades had foreseen; it just came out of the blue.

Of course, the township and village enterprises–which is to say, private companies avant la lettre–did not actually come out of nowhere. They drew on China’s long traditions of commercial enterprise, and people’s experience with living and operating in a market economy before it was suppressed in the 1950s. In an interesting new paper, “Markets under Mao: Measuring Underground Activity in the Early PRC” (the link is currently open-access), Adam Frost and Zeren Li offer quantitative evidence to suggest that hidden market activities continued at scale even through the height of the Maoist period:

There was already substantial market-based activity prior to the launch of economic reforms. Even after the “socialist transformation” of the Chinese economy was ostensibly complete, Chinese citizens continued participating in “underground market activity,” i.e. private acts of exchange that occurred outside of systems of planned allocation and distribution and which were intentionally concealed from the state. A broad host of actors, ranging from rural people who “abandoned farming to take up commerce” to merchants who specialized in the illicit wholesale trade of ration certificates, devised novel strategies to evade state control and engaged in consensual private transactions. While these individuals often filled critical voids in the economy, they were collectively maligned as “speculators and profiteers” and, for three decades, were the recurring targets of mass campaigns. Yet, even at the height of the Cultural Revolution when anti-capitalist sentiments reached their zenith, “speculation and profiteering” were never wholly suppressed.

The authors use local administrative documents recovered from flea markets to compile data on 2,690 cases of “speculation and profiteering” that authorities prosecuted in two areas, the rural county of Chun’an in Zhejiang and the city of Zhenjiang in Jiangsu. These records often list what items the “speculators” were accused of selling, and at a what price. The value of the transactions was usually pretty substantial:

The mean case value (i.e. the estimated total value of activity described in each case) is about 334 yuan for Chun’an and 362 yuan for Zhenjiang. To put these figures into perspective, in 1955 the national average income was 102 yuan for an urban worker and 94 yuan for a rural farmer, and income levels remained stagnant for most of the 1960s and 1970s. In other words, the mean case involved activity that represented three years’ worth of consumption for the mean worker. Given that prosecuted individuals probably succeeded in concealing some portion of their gains, this figure is likely only a fraction of the true quantities of goods, cash, and ration certificates involved in each case.

The authors use various assumptions to estimate the total size of underground market activity from these figures. The results depend so much on assumptions that the exact figures are not particularly meaningful, but the range of estimates is consistent with contemporary estimates of the “shadow economy” of illegal transactions in most countries–perhaps suggesting that Maoist China was more economically normal than its ideology would indicate. This finding is also interesting:

We observe that the average spread between the purchase and resale price of items in underground market transactions was relatively low and remained so throughout the entire period of observation. There was, on average, no more than a 19% mark-up on items that were bought and resold in underground markets. These figures suggest that the maximum perceived risk of capture was low, even during the height of the Cultural Revolution.

The low black-market premium also suggests these kind of underground transactions were widespread enough that the scarcity value was not extreme, and that underground traders faced enough competition that they could not charge exorbitant prices. The repeated campaigns against “speculation and profiteering” that generated these administrative documents therefore do not seem to have been particularly successful. When the formal prohibitions on these private transactions were lifted in the post-1978 reforms, many Chinese people had plenty of experience with trading and business that could be quickly put to use.

Frost and Li’s account lines up with my own thinking about the reasons for the early success of China’s economic reforms–although when I pondered this question before, I focused more on the fact that China did not actually spend that much a time as a full-fledged socialist economy (see my previous post, “How long was China Communist?“). The time from the forced nationalization of private firms in the mid-1950s to the re-legalization of urban and rural private sectors in the early 1980s was barely three decades. By contrast, socialist prohibitions on private enterprise in the Soviet Union lasted two full generations, long enough to wipe out any previously existing base of skills and knowledge formed in the private commercial economy (which was not that developed in tsarist Russia anyway).

These two explanations seem fully complementary to me: China’s socialist prohibitions on private transactions were neither complete enough to truly stamp out market-based activities, nor did they last long enough to for the population of people with experience running businesses in a market economy to die off. Once the state began to tolerate markets again, the hidden traditions of private enterprise could come out into the open.

I got involved in a Twitter discussion with Brad Setser and others over the nature and causes of China’s high share of global manufacturing. This prompted me to go through some tedious statistical work to establish some basic facts for my own satisfaction. The results are now more or less final, so I am going to outline them here.

We know that China has a high share of manufacturing in its GDP, with the sector’s value-added accounting for about 28% of total value-added at last count. This is higher even than other manufacturing champions like South Korea (25%), Germany (21%) and Japan (20%), let alone the relatively de-industrialized economies like the US (11%), UK (10%), Brazil (12%) and South Africa (13%). Since China is such a large economy, accounting for about 19% of global GDP, its manufacturing sector is also very large relative to the world economy. As of 2021, China accounts for 31% of the world total of manufacturing value-added, according to the UN national accounts database.

Why is China’s manufacturing sector so large? In part, China is making goods for its own use, so its large manufacturing sector reflects the growth in China’s own demand. In part, China is making goods for use by others, so its large manufacturing sector also reflects its success as an exporter. We can start answering the question by quantifying the relative contribution of those two factors.

I did this by using the OECD Trade in Value Added (TiVA) database. Among other things, the database breaks down China’s manufacturing exports by whether the value-added originates domestically or abroad (in the form of imported goods and services used to produce exports). Although there is some change over time, about 80% of the value of manufacturing exports ends up contributing to domestic value-added. Once we know the amount of domestic manufacturing value-added generated by external demand, we know that the rest must be generated by domestic demand.

Doing this simple calculation shows that in recent years, about 40-45% of China’s manufacturing output has come from exports, while 55-60% has come from domestic demand. This pattern was established in 2009 by China’s massive property-and-infrastructure stimulus in response to the 2008 global financial crisis. Since then, the level of investment activity in the economy has stayed very elevated. So we can say that China’s manufacturing sector is indeed mainly oriented to domestic demand, but it’s definitely true that the contribution from exports is quite large. A 55-45 split in an economy of China’s size is a pretty significant reliance on external demand. And that reliance has increased more recently. The current edition of the OECD TiVA database ends in 2018; extending the estimates to 2022 shows that the export contribution has probably picked up quite a bit due to the pandemic export boom.

Nonetheless, China’s manufacturing share of GDP has declined since around 2010, meaning that manufacturing value-added has grown more slowly than the rest of the economy. The value-added breakdown shows that most of that slowdown has come from domestic demand, probably investment. What’s surprising is not so much that China’s investment boom has cooled off from the stimulus-driven peaks after the financial crisis, but that the slowdown has been so gradual. From about 2015-19, a slowdown in exports also contributed to the declining manufacturing share, but the pandemic export boom boosted the external demand contribution again. In a counterfactual world without the pandemic export boom, China’s manufacturing share of GDP would most likely be noticeably lower today.

The breakdown between exports and domestic demand can also be used to shed light on China’s share of global manufacturing (using world manufacturing value-added as the denominator rather than China’s own GDP). This shows a steadily rising trend, meaning that while China’s manufacturing growth did slow down relative to the rest of China’s economy, it continued to be faster than manufacturing growth in the rest of the world. But the drivers of the increase shift over time in ways that reveal the changing patterns of growth.

From 2000-2008, China’s share of global manufacturing rose mostly, though not entirely, because of growth in exports: this was the export boom caused by the mass relocation of manufacturing capacity to China after its WTO entry. Export value-added rose to 8.4% from 2.7% of the global total, while domestic value-added rose to 6.1% from 4.3%. From 2008-2019, export value-added rose further to 11.5%, while domestic value-added rose much more, to 16%. Again, this is the post-financial crisis investment boom. Over 2020-21, export value-added rose to 13.9% while domestic value-added rose to 17.4% (the UN database that supplies the global total of manufacturing value-added hasn’t yet updated to 2022).

Whether China can sustain its pandemic-era gains in exports is obviously an important global macro question. Some of that boost was due to surges in demand in the US and elsewhere that are now retreating. But some of it was due to supply-side developments, like China’s emergence as a major vehicle exporter, that could be more durable. Success on the export front would certainly help support China’s share of global manufacturing and its manufacturing share of GDP. But the crucial factor is really whether China can sustain the super-elevated levels of investment that have driven domestic demand for manufactured goods. Given the unwinding of the property boom and the complete buildout of many forms of infrastructure, this seems increasingly unlikely. Broadly, the fading of the post-crisis investment boom is why I think China’s manufacturing share is probably going to decline again (see my earlier post, “Re-de-industrialization“).

Technical note. Making these calculations using the OECD TiVA database was pretty straightforward. Extending them into more recent years using China official data was a bit tricky. The total for China manufacturing value-added in the TiVA database was basically the same as in the NBS national accounts. However, the value of manufacturing exports is not the same; the TiVA database is built on top of international input-output tables that try to make different countries’ trade figures consistent. Usually, the value of China’s manufacturing exports in TiVA is around 80% of the value of manufacturing exports reported by China Customs. I’m not sure what the reason for this is, but it seems to suggest the headline value of manufacturing exports is overstated. Using the Customs value of manufacturing exports generated a residual for domestic manufacturing demand that was implausibly small, so I adjusted it to be consistent with the TiVA data by using the ratio between the Customs figure and the TiVA figure.

Shortly after being reappointed as the governor of China’s central bank, Yi Gang gave a speech in Beijing with the rather dull title of “Building a Modern Central Banking System to Contribute to Chinese Modernization.” As is obligatory with the speeches of most Chinese officials, it opens and concludes with references to the important decisions of the Communist Party and the major slogans of the moment.

The key term in the title of the speech is one of these coinages from Xi Jinping: the more precise translation would be Chinese-style modernization (中国式现代化), as it refers not simply to the modernization of China but modernization done China’s way. It’s one of the umbrella terms for his overarching project of achieving national greatness.

The bulk of the speech consists of Yi summarizing the central bank’s major policy priorities and justifying how it has pursued them. The temptation for foreign readers is to skip over the political sloganeering at the beginning and end and focus on the technical content in the middle.

Indeed, this was the approach that Yi himself took when he gave a more elaborate version of the same presentation, in English and with charts and references to the economics literature, to an audience in the US. Yi favorably compared the gradualist methods of the People’s Bank of China, which moves interest rates rarely and in small intervals, to the dramatic recent swings in Federal Reserve policy (see the writeup in The Economist for more on this angle).

But abstracting Yi’s speech from the Chinese political context in this way risks blinding us to the some of the significance of what he says. How Yi positions the central bank’s technical policy priorities relative to the overarching goals of the Communist Party leadership is a very important part of his speech. What I think he is doing is nothing less than trying to ensure that modern central banking and financial regulation can be implemented in the Chinese political context.

To maintain political buy-in from the leadership for his preferred policies, Yi needs to convince them that these policies do in fact serve the overarching political goals set by Xi. He starts off his speech by citing the official mandates of the People’s Bank of China, as written in Chinese law. Based on the text of the law, he asserts that

Preserving currency stability and financial stability are the two central tasks of the PBOC. On this point, everyone has increasingly reached a consensus in recent years. By properly accomplishing these two tasks, we can promote full employment and economic growth, which better serves Chinese-style modernization.

That is pretty direct and clear: the traditional goals of the PBOC and the consensus of technical experts are not in conflict with the current political agenda, but actually support it. Yi further elaborates that the concept of currency stability includes stability of consumer prices and stability of the exchange rate, both of which are not only beneficial to the population at large but help the nation achieve long-term goals:

Stable consumer prices and exchange rates serve to safeguard the pocketbooks of consumers, so that the money in their hands won’t depreciate. Fundamentally, this is a people-centered effort to safeguard the interests of the broadest possible majority of the people. …

Price stability and the basic stability of the renminbi exchange rate at an adaptive and equilibrium level provide strong support for us to realize the strategic goal of Chinese-style modernization. 

At the end of the speech, Yi acknowledges that the new agenda laid out by Xi at the Party Congress and other high-level meetings requires at the Party Congress “means new requirements” for the work of central bankers. But fundamentally what he is saying in this speech is that these new requirements can be met by the same old policies. The key ones he highlights in the speech are: a less controlled and more market-determined exchange rate, strict regulation oriented towards preventing financial risk, and a conservative monetary policy biased against dramatic moves.

This emphasis on stability and continuity is probably why Yi’s speech got little attention in the press. He did not propose anything new, and largely emphasized recent accomplishments. But that is the point he wants to make: these policies have been successful in making China stronger, so they should continue. You don’t have to agree that his argument is correct on the merits to understand that he wishes to preserve the PBOC’s autonomy and policy preferences in an era of heightened politicization and grand campaigns.

It’s not hard to imagine how some of Xi Jinping’s priorities could potentially have big implication for macroeconomic policies. He has a vision of a whole-of-society effort to refashion China’s industrial base to make it less vulnerable to external shocks. His “dual circulation” concept calls for reducing China’s economic dependence on the rest of the world while increasing the rest of the world’s economic dependence on China.

To pursue these ideals, some officials could perhaps argue for an aggressively undervalued exchange rate, that keeps exports competitive and deters domestic spending on imports, or a much looser monetary policy, to ensure industrial upgrading has plenty of resources. That stuff hasn’t happened, perhaps because Xi favors strict supervision and control in finance as in other areas. But it’s still an open question how existing macroeconomic policies will adapt to the political requirements of Xi’s so-called “new era.”