"Will job losses lead to social unrest?" My Take

Today, there was a note written by Wang Tao at UBS with the above title. The question is a crucial one today because bullish sentiment for China depends almost entirely on the Chinese government's ability to prevent large scale social instability. The conclusion of the piece was that "large-scale unrest that threatens general social stability and overall investor confidence is unlikely." The arguments sound quite reasonable:

1. Job losses will only be about 15 million, or 3.5% of non-agricultural employment.

2. Migrant workers, who are hit the hardest, can't organize effectively anyway.

3. China weathered the last wave of unemployment, which saw unemployment level at around 35 million, with little difficulties in the late 90s. And, China had less money back then.

I have no issue with argument #3. Indeed, though ugly, China weathered the unemployment wave in the late 90s with only scattered protests in a smattering of cities in the northeast and the interior. All in all, not so bad (well, it was bad for the unemployed!).

However, I find the 15 million figure highly unlikely, even for now, much less for 2009. A Ministry of Labor and Social Security official revealed recently that some 10 million migrant workers have already been laid off and returned to the countryside. It seems extremely optimistic to say that total unemployed migrant workers in 2009 will be 15 million. Let's start the calculation again.

First, total migrant labor population in 2007 was about 140 million. I think global recession and partial rebalancing will impact at least 20% of the migrant labor population. This makes about 28 million migrant workers unemployed in 2009. On top of that, education officials revealed recently that only 70% of the 6 million or so college graduates found jobs in 2008. This leaves unemployed college graduate population at 1.8 million. In 2009, I expect things to be much worse, so at least 50% of college graduates will be unemployed, creating another 3 million or so college educated unemployed. For the Chinese government, this is the most dangerous ingredient since it was collective action by college students that led to systemic political trauma in 1989. To prevent recurrence, the central government will likely roll out a scheme to pay for the living expenses of almost all unemployed college graduates. I think my students at Northwestern would have wished for that policy too!

So, we are up to 33 million unemployed already, nearly the highs in the late 90s. If we add a couple million more from layoffs from urban private and FIEs, we would have an unemployed force of at least 35 million. I would not be surprise if at the end of 2009, we see an unemployed labor force of 50 million. To be sure, if we believe in the other arguments that Ms. Wang puts forth, we would not see a major problem. At the extreme, the central government would simply roll out subsidies that pay for the living expenses of every unemployed person. Even if the unemployed force reaches 50 million, the Chinese government would only have to pay (50 million*100dollar*12 months) 60 billion USD (408 billion RMB). That is a substantial sum, but China can surely handle it for two to three years, suffering perhaps slightly lower credit ratings. However, the notion that migrant workers have less ability to act collectively is unfounded based on everything that we know about unrests in China. All of the rebellions in Chinese history were led and carried out by peasants, including the one that put the current regime in power. Besides 1989, the largest domestic disturbance took place in rural Renshou County in the mid 90s, which saw the deployment of tens of thousands of troops. Furthermore, unlike the layoffs in the 90s, which mostly affected middle-age or elderly SOE workers, the current wave of layoffs affects a young and vibrant cohort most capable of carrying violent collective action against the state. Without any systematic triggers, we at least will see a spike in localized riots which necessitate the mobilization of People's Armed Police (PAP) units all over China. The central government would also be compelled to (and they are doing so already) roll out generous unemployment benefits for migrant workers and college graduates (to the tune of 300-400 billion RMB). If a systematic trigger occurs and instability spreads to a sizable city, we will see the large scale mobilization of both PAP and army units and possibly substantial bloodshed. In most scenarios, the CCP regime would still survive a large scale, cross regional rebellion. However, "overall investor confidence" will be lost.

What is the "systematic trigger" which I refer to? I don't know exactly what it would be. However, if we look back in history, it can be a wide range of events, including the death of a popular leader, a serious natural disaster, the spread of a deathly infectious disease, a small student demonstration turned violent, religious groups....etc. The point is that there is an interactive effect which is quite harmless when unemployment is low (as shown last year). However, when unemployment is high, a number of different shocks can interact with unemployment to create something explosive. I think the UBS note, besides getting some numbers and history wrong, vastly underestimates the possibility of this interactive effect.