Body
The Bank of Korea has released its annual North Korean GDP estimates, and the BOK has concluded that the North Korean economy contracted by 1.1 percent. I will briefly discuss why these figures should be taken with large grains of salt, and then apply my own subjective smell test.
On methodology: to compute GDP, which is a value-added concept, one needs information on the levels of all economic activities, information on the prices of inputs and outputs, and a recipe book that tells you how many widgets and hours of labor are needed to make, say, a car.
In the case of North Korea, the existence and veracity of all this information is highly problematic. I want to make clear my comments are not a criticism of the BOK per se—theirs is a thankless task—but that the task of computing North Korean GDP is a hazardous exercise.
On activity levels, my impression is that some things (the number of boxcars leaving the coal mine, for example) are amenable to counting via satellite photography. Similarly, there are ways to get a handle on crop production. But figuring out manufacturing is more challenging in part due to the issues of product variety and quality. Services are the most challenging—no boxcars to count, and more generally, it is hard to measure output apart from measuring inputs (we face this problem even in the West), and potentially huge differences in quality: think quality differences across American high schools.
Like I said, take these figures with a grain of salt.
Then, there is the issue of collecting price data to figure out how much the inputs cost to generate a particular value of output.
Then, there is the recipe book used to aggregate all these figures. The BOK says it uses South Korean prices and value-added ratios. Those are surely wrong for the task at hand, but it is not obvious what alternative would be any better.
Then, my understanding is that once these calculations are made they are subject to inter-ministerial bargaining, i.e. they are politicized.
Like I said, take these figures with a grain of salt.
So, what does the Noland informal smell test reveal? Agriculture is down (-0.8%)and that is consistent with analysis of the FAO—the reversion to the new normal after several good years of harvests.
Mining is down (-2.6%), and it is conceivable that production is down as the report says. But it is almost surely the case that the value of output is down due to the ongoing fall in world mineral prices.
The report has manufacturing (-3.4%) contracting by an even larger margin than agriculture or mining, however, with both light manufacturing and heavy industry down. The result for light manufacturing is surprising: apparel exports to China have been rising and KIC was still in operation. Put an asterisk by that one.
Electricity production was down a whopping 12.7% due to diminished hydroelectric output because of the drought. I have no special insight into hydropower, so I will take them at their word, but it is curious that electrical power production could fall by 12.7% and overall GDP only fall by 1.1%.
Construction and services were the only bright spots in the BOK’s telling. And here is where I think that the subjective assessments of observers and the BOK probably diverge most sharply. We know that a process of relaxation or decriminalization, or whatever you want to term it, is ongoing and this had led to a steady increase in private sector activity for lack of a better term. This growing non-state economic activity appears to be concentrated in services. And we know services are hard to measure. My guess is that the BOK has underestimated services activities, which even by their own analyses now account for nearly one-third of the economy, but this is just my guess.
So, did the North Korean economy contract last year? I suppose it is plausible, but I would not be surprised that if we could properly calculate GDP we would find that it actually registered modest growth. The intriguing thing is that I suspect that the North Koreans themselves do not even know.
