Ziplining beats parsing questionable figures At a closed-door conference earlier this year in Seoul, a group of South Korean and American economists gathered to analyze the North Korean economy. Someone observed that because of the unexpected change in the South Korean election calendar, the 2017 Bank of Korea analysis of the North Korean economy would be released under a progressive government. Estimating macroeconomic changes in North Korea is a dirty job, but someone has to do it, and the analysts at BOK deserve our thanks. Nevertheless, it is widely suspected that their work is subject to politicization, and when two of the Americans—expecting a big jump in estimated growth—attempted to solicit bets on the 2016 estimated growth rate they could find no takers. So when the BOK
Marcus Noland considers the following as important:
This could be interesting, too:
The Sound of Economics writes Apps without borders? How COVID-19 apps show the limits of the EU digital single market
Petri, Plummer writes New East Asian trade blocs create tough choices for China
Demirgüç-Kunt, Lokshin, Torre writes Early lockdowns save more lives and livelihoods
At a closed-door conference earlier this year in Seoul, a group of South Korean and American economists gathered to analyze the North Korean economy. Someone observed that because of the unexpected change in the South Korean election calendar, the 2017 Bank of Korea analysis of the North Korean economy would be released under a progressive government. Estimating macroeconomic changes in North Korea is a dirty job, but someone has to do it, and the analysts at BOK deserve our thanks. Nevertheless, it is widely suspected that their work is subject to politicization, and when two of the Americans—expecting a big jump in estimated growth—attempted to solicit bets on the 2016 estimated growth rate they could find no takers.
So when the BOK released its analysis two months ago (sorry for the lag—I was vacationing with my kids in Hawaii), the estimated jump in the overall growth rate to 3.9 percent was not entirely a surprise. Anecdotal accounts paint a picture of improvement, at least in Pyongyang, and past progressive governments have been suspected of juicing the numbers to paint a rosy picture. (To be fair, the conservatives have been thought to shave the figures the opposite way.)
Apart from politicization, estimating North Korean GDP is a challenging task technically. The BOK analysts gather quantity data using a variety of means from various government agencies, the National Intelligence Service figuring prominently, and then convert those quantities to values using the value-added weights of the South Korean input-output table. Quantity output in some sectors may be relatively easy and accurately estimated using satellite photography and other techniques (think the rice crop or coal production) but the quantity output in other sectors, particularly services, may be harder to judge. To make matters worse, quality differences which may be critical in advanced manufacturing and services, aren’t captured by the quantity measures. The operational implication is that the BOK’s estimates for services and advanced manufacturing may underestimate year-to-year change: risk aversion pushes the estimates toward marginal output changes relative to the previous year, even if the change in true, quality-adjusted output has been more substantial.
If there is a place where the BOK is systematically underestimating North Korean growth, it is in the service sector.
In short, the basic inputs to the calculation are measured with error, with the possibility of a systematic tendency to underestimate change. Those errors are compounded by the aggregation process. There is no reason to believe that the South Korean parameters are appropriate for the North Korean economy, but in the absence of North Korean prices and value-added ratios, the use of the South Korean figures is an understandable, if less than ideal, response. This flawed estimate reputedly goes through an inter-agency consultation process which is apparently where the politicization occurs.
So, having now reviewed how this particular sausage got made, how does it taste? Some of the results are consistent with my priors, some are not, and in one case this year’s analysis seems to have corrected a really strange result that BOK put out last year.
As mentioned earlier, overall GDP in 2016 is thought to have risen 3.9 percent. (Since the calculation method does not use North Korean prices, the BOK does not put out estimates of the North Korean inflation rates. All these growth rates are “real.”)
Agriculture, forestry, and fishery increased 2.5 percent in 2016, after shrinking slightly in 2015. Agriculture is strongly affected by weather conditions and the availability of inputs and could be expected to bounce around year to year. Fishery output may be adversely affected moving forward by the sales to foreign countries of fishing rights in North Korean waters, though the revenues from those sales will help the state treasury.
Mining increased 8.4 percent after shrinking in 2015. It is an open question how UN sanctions will affect the mining sector going forward.
In manufacturing, on the BOK’s numbers, heavy industry experienced a stunning turnaround growing at 6.7 percent in 2016 after contracting 4.6 percent in 2015. This result appears particularly odd when placed against the estimate that light industry (which includes textiles, garments, and shoes) grew at only 1.1 percent (after shrinking 0.8 percent in 2015). Apparel exports were one of the notable bright spots outside of the extractive sector, and one might have expected a higher rate of growth. One might reconcile this oddity by noting that a small growth rate on a large base may be larger than large growth on a small base, except that the BOK estimates that heavy industry is roughly twice the size of light industry. These results are a bit of a head-scratcher.
Electricity, gas, and water was up a whopping 22.3 percent on the back of a rebound in hydropower and thermal generation after contracting 12.7 percent last year. It’s really hard to know what to make of these results. One normally thinks of energy consumption moving roughly parallel with economic growth, and as I observed last year, it’s hard to believe the economy grew at all if electricity, gas, and water output declined 12.7 percent. This year it’s just the opposite: the BOK estimates electricity, gas, and water output grew at close to six times the rate of the overall economy. It’s very hard for me to make any sense of the figures for this sector over the past two years. It’s possible that hydropower and thermal power generation are substituting for coal- or oil-fired electrical generation, which might be desirable from the standpoint of the North Korean economy, but would not lead to a giant surge in electrical output (unless there was some means of storing huge amounts of power). Again, I am left scratching my head.
As expected, the output of services showed little change, up 0.6 percent after rising 0.8 percent in 2016. The slow steady estimated growth is not unexpected for reasons that I have already elaborated. But I suspect that BOK is capturing neither the true quantity of output nor its rising quality, and the growth estimate is downwardly biased. If there is a place where the BOK is systematically underestimating North Korean growth, it is in the service sector.
See you next year.