Analysts are trying urgently to evaluate the potential impact of a full-fledged trade war on the Chinese economy. This typically involves estimating how much various tariff scenarios will reduce China’s GDP growth, with current estimates ranging from the minimal, 0.1 or 0.2 percentage points, to the substantial — as much as 2 percentage points.
This is probably the right way to evaluate the impact of external shocks on other countries. For China, however, it’s wholly inappropriate. The fact is that Chinese GDP will be unaffected by a trade war with the U.S., no matter how severe, because the government will do whatever it takes to meet its growth targets.
That doesn’t mean higher tariffs won’t damage the Chinese economy. The pain will instead show up instead mainly in the form of a more rapid rise in debt. The harsher the war’s impact, in other words, the more debt China will need to achieve its growth target.
To understand why, one needs to appreciate the difference between GDP growth as a system output and as a system input. In most economies, GDP is a measure of output generated by the economy. At the end of every period, during which the economy does what it does subject to standard economic constraints, government statisticians measure the relevant changes in activity and this is reported as the amount by which GDP expanded or contracted.
This isn’t what happens in China, as even Chinese leaders will occasionally admit. In China, the government sets the GDP growth rate early in the year at a level thought adequate to accommodate its social and political objectives, among which is to keep unemployment low. The political nature of the target modifies the standard economic constraints, encouraging local governments to generate whatever additional economic activity is required so that, along with the economic activity of the private and real-estate sectors, the target is reached (within a few tenths of a percentage point).
Two factors unique to China are critical for this system to work. First, till now local governments haven’t been subject to hard budget constraints. They can engage in near-unlimited amounts of non-productive economic activity unconstrained by worries about remaining solvent.
Second, and necessary for the first, local governments control most credit creation within the banking system. Because such loans are directly or indirectly guaranteed, banks don’t have to write down loans when the projects they fund cannot service the debt. This allows the banks to extend as much new credit as local governments need to meet their targets.
If, as is widely acknowledged even by the government, the amount of debt written down in every period is less than the amount of non-economic loans extended during that period, recorded GDP growth will rise, regardless of whether the underlying economy can or can’t generate enough activity on its own.
As long as China has debt capacity, and the government is willing to use it, China can achieve any GDP growth target it wants. Thus, while GDP numbers may tell us something about the government’s social and political priorities, they’re a poor measure of the underlying performance of the economy. They tell us even less about the impact of trade war on the Chinese economy.
This doesn’t mean there’s no way to evaluate the adverse impact of a trade conflict. As any systems analyst knows, while an input can never tell you anything about the performance of a system, an output can. What analysts need to examine is the appropriate output data, such as the amount of new credit needed to generate additional economic activity. This will show how much unproductive activity is required to reach the growth target.
There’s one silver lining. Chinese leaders seem determined to get debt under control, which will ultimately require either giving up the growth target or lowering it considerably. A trade war, then, could be used politically, to overcome internal resistance to allowing GDP to slow. Pain is inevitable. The only question is whether China seeks to hide it, or use it to its long-term advantage.