A joint food security assessment from UN agencies working in the DPRK sparked headlines in May after concluding that 10.1 million North Koreans were facing a deep “hunger crisis.”

Based on findings from the UN’s Food and Agriculture Organization (FAO) and World Food Programme (WFP), May’s ‘Joint Rapid Food Security Assessment‘ estimated that aggregate North Korean cereal production for the 2018-2019 period would be “about 4.9 million mt” – 12 percent below the 2017 year-average.

After the FAO/WFP in March published DPRK Ministry of Agriculture data showing only 4.95 million tonnes of overall food production in 2018 – down over a million from 2016 – the combination of indicators has re-ignited conversation about how grave the situation may currently be in North Korea.

But though the situation seems grave on the surface, when zooming out of the latest FAO/WFP warnings and viewing the data in context of other indicators, several interesting issues appear to emerge.

Firstly, though the FAO/WFP describe the current situation as the worst in ten years, when looking at the situation from 1990 to present it appears that the situation appears to be relatively good.

Secondly, data shows that it might not cost the DPRK that much to solve the shortfall by itself and that, for some reason, the FAO/WFP consistently over-estimated the total of what North Korean farmers would likely to produce each year.

Thirdly, while signs of a pending food shortage crisis appear to loom in the FAO/WFP assessments, market rice and corn price data tracked by the Daily NK, in contrast, shows food prices at all-time seasonal lows.

Fourthly, though fertilizers and cereals are often imported by North Korea to improve the overall situation, the volumes brought it don’t seem to be linked to the overall amount of food harvested each year by the North.

Fifth, while the FAO/WFP blames the poor performance of the DRPK agricultural sector in its annual reports on drought, flooding, or a combination of the two weather problems, the volume of North Korean state media reporting on these problems seems to have no connection to overall food harvest.

Finally, though the DPRK Ministry of Agriculture used to provide data to the FAO/WFP on the nationwide volume of water in irrigation reservoirs, these figures appear to have no relation to either annual crop production or to state media reporting on drought and flooding.

To better understand the latest FAO/WFP data in context, NK Pro invited leading DPRK economists and data specialists to share their analysis and observations about what, exactly, might be going on in North Korea at this time:

  • Benjamin SilbersteinAssociate Scholar at the Foreign Policy Research Institute, and co-editor of North Korean Economy Watch.
  • Go Myong-Hyun – Research fellow at the Asan Institute for Policy Studies
  • Peter Ward – An NK Pro writer and researcher focusing on the North Korean economy
  • Stephan Haggard –A regular NK News contributor and director of the Korea-Pacific Program, and distinguished professor of political science at UC San Diego

Note: All charts below have been assembled using data mined from historic FAO/WFP reports and from third party sources. Underlying data available on request.


NK Pro: With the big picture in mind, how bad is the current “crisis”?

Benjamin Silberstein: I wouldn’t even call it a crisis as of yet. So far, all we have are foreboding warnings. They should be taken seriously but there’s little to suggest that a true crisis is unfolding as of now.

In context, the numbers are not great, but also not catastrophic. The numbers for cereal production used in this graph are somewhat odd and don’t correspond with the WFP numbers for the past few years, for whatever reason. I’ve checked them against some other assessments and I can’t figure out why they’re so different. In any case, numbers for the past few years in the FAO database are much lower than what WFP has reported in their food and crop assessments.

I say that to note that North Korean harvests have been growing for the past few years. In that context, this year looks bad. But compared to the 1990s and early 2000s, it’s fantastic.

The question then becomes: what do we define as “good” and “bad” in the North Korean context? Is “not famine” to be classified as “good”? I don’t think so. A downturn in the harvest is bad because margins are so incredibly thin in North Korea, especially for some communities. Not all can shop on the markets.

So I would say, this year is bad, but in context, not catastrophic.

Stephen Haggard: The first thing we need to admit is that all of the data is vulnerable to error, and the messages from it are contradictory. In particular, the price data is confusing. Shortfalls of the stated magnitude should be accompanied by sharper price effects, and particularly against a backdrop where world benchmark prices are increasing.

The fact that prices do not seem to be rising suggests some stabilization mechanism, such as release of stocks on the part of the government. On the other hand, the margins for error are clearly small, with PDS rations falling and overall output under some measures falling below prior FAO estimates of overall food needs.

Peter Ward: Given the relative stability of food prices and the exchange rate, it doesn’t seem that things are all that bad. The FAO-WFP data gives us only a limited picture with regard to the actual supply of cereals because it doesn’t include any private production anymore. The Multi-Indicator Cluster Survey (MICS) carried out with UNICEF funding by the North Korean Central Statistics Bureau indicates that over 50% of households (including nearly 30% of urban residents) had private agricultural lands.

This was considered a statistically representative survey from which to base conclusions regarding nutrition, yet the WFP was unable to make or obtain estimates about private production. This is rather problematic.

Go Myong-Hyun: The current dip in production seems to be a temporary instance in the long term trend of production recovery. Four million tons seems to be the critical lower limit. We are yet to get there.

NK Pro: Relatively speaking, how affordable are these figures? What’s stopping DPRK paying?

Go Myong-Hyun: North Korea has experienced a current account deficit for much of its economic history, so there is an inherent incentive for them to minimize the trade deficit by achieving self-sufficiency in crop production.

Having said that, the almost total ban on exports imposed by UN sanctions has exacerbated the deficit even more. It is also compounded by the fact that the crop price increased in 2018.

Stephan Haggard: It is plausible that the North Korean regime has miscalculated its access to foreign exchange. As I showed in a recent post with Marcus Noland and Leo Byrne, there has clearly been a falloff in export earnings to China as a result of the maintenance of core sanctions; the fact that the North Koreans sought aggressive sanctions relief in Hanoi demonstrates the pressure. The calculations of the cost of import requirements exceed total North Korea exports to China by a significant margin.

Peter Ward: It is unclear what the actual requirements are because we do not know enough about the actual state of domestic production. However, if we assume that the WFP/FAO import requirement estimates are accurate, North Korea would likely have to spend several hundred million dollars at a minimum to fulfill the requirements. The wholesale spot price of corn is around $160, rice $400, soybeans $400, wheat $210 etc. and they would want to mix the cereals to maximize the nutritional benefits of variety.

Do they have the cash on hand to do so? I do not know. I hope they can afford to import some of what they need, and that the international community provides them with meaningful and targeted assistance to help resolve the nutritional issues they face.

Benjamin Silberstein: I can’t say for sure what’s preventing the DPRK from paying. But it’s a matter of priorities. Hundreds of millions are being spent on other things: weapons, luxury goods, boastful infrastructure projects that will realistically render very little revenue in the long run, or even at all.

So it’s a matter of political priority. People experiencing food shortages are just not as important as the core classes, and whatever needs Kim Jong-un has decided should take precedence.

I actually suspect, though, that part of the problem has to do with logistical capacity. Food security isn’t just about what’s available — it’s also about who gets it. It may be that the DPRK has a real need for assistance from the international community to deliver supplies to those who really need them, but I’m not sure.

NK Pro: What do we learn about official NK data points given the almost consistent over-estimation each year by the FAO and swinging nature of PDS data?

Go Myong-Hyun: The consistent overestimation indicates there is a systematic bias in estimation. It could be due to the difference in methodology between FAO and DPRK, or some other factor. I don’t think it is a significant issue.

Peter Ward: The data is inconsistently collected, and that the FAO should provide more raw data and be open to having detailed methodological discussions about how the data is collected, how the estimates are made, and when their estimates are inaccurate.

The PDS data is especially curious and rather difficult to make sense of given what we know about the functioning of the PDS from other sources. It would be nice if the FAO would release their food balance data for the post-2013 period, and perhaps write up a detailed note on how the data was collected, how estimates were made, and whether any obvious errors were made and then corrected.

The major issue with the current system is that the raw data remains inaccessible, and the FAO offers only very limited information about how estimates are made and conclusions drawn.

Donors, the media, and the public at large are supposed to take the FAO’s word for it, by the looks of things, or email them. It would really help their case if they would make raw data and the methodological assumptions undergirding it freely available up to 2018. If not raw data, then at least the collated figures post-2013 for food balances, so that we do not have to sift through uneven and inconsistent reports penned annually.

Benjamin Silberstein: Honestly, I don’t think there’s necessarily that much politically motivated manipulation going on, as much as lack of capacity in data gathering. We don’t know exactly how the PDS value is calculated, for example. I’d have to imagine it to be literally impossible given that there’s little consistency in PDS-distribution at all, according to most people from North Korea.

Stephen Haggard: I would not make much of the discrepancies between the FAO and the Ministry of Agriculture estimates. First they are relatively small and in any case have converged (in part because of FAO dependence on Ministry of Agriculture information). Two things are important in the attached graph. First PDS rations have fallen 25% from the 2013-4 peak, but more importantly both estimates for 2018 fall below FAO estimates of total food requirements, and by more than half a million metric tons.

NK Pro: What explains the lowering prices at the markets but worsening food shortages (according to DPRK official data) tracked by FAO/WFP?

Stephen Haggard: To me this is the biggest puzzle, and again in the context of rising world market prices. It suggests to me that the government could be releasing stocks, something they may not want to advertise.

Benjamin Silberstein: It seems that market prices have fairly little to do with overall supply. This may sound strange but I think it’s best explained by all the unknowns about the market. This year, prices have dropped as the economy overall has gotten at least somewhat (we don’t know how much) depressed by sanctions. As people can’t pay the same prices they used to, suppliers have to lower their prices to sell anything at all.

There’s just so much we don’t know about the markets — who are the consumers, how are grains supplied (from what sorts of fields and farms), so I’m weary of drawing strong conclusions from market prices about the reality on the ground.

Peter Ward: There are several potential explanations I see. Perhaps one set of data is wrong, or perhaps both contain bias. But if we assume that they both correspond to the realities on the ground to a meaningful extent, then how is it that prices could be falling even as supplies are also falling?

Perhaps sanctions could be to blame: workers in export industries who held food prices up are no longer getting paid or paid far less, and therefore aggregate demand is down further than aggregate supply. If demand has declined more than supply, then we might see price falls like this.

Go Myong-Hyun: Imports/smuggling from China that is not captured by the official statistics.

NK Pro: Do you see any correlations? What do you make of the trends visible here?

Peter Ward: One of North Korea’s main issues with fertilizer is that it doesn’t have the electricity to produce it in sufficient quantities, so it has to import. Some years when the electricity supply is better, it may produce more. What’s more, if the weather is bad, fertilizer may not help to increase the harvest that much. So fertilizer imports are to some extent an indicator of electricity supply, but not necessarily a lead or lagging indicator of food supply and production.

Food import reporting may disguise the presence of subsidies from China. It is important to examine the actual price paid for each cereal. In 2011 and 2012, the Chinese may have been providing North Korea with grain at heavily subsidized prices, and then stopped after 2013 or 2014. Hence, the higher import numbers may reflect Chinese support, rather than pressing North Korean need for more food due to harvest shortfalls.

Benjamin Silberstein: It looks like fertilizer and cereal imports declined as harvests increased. It may be that the government decided to draw down on these, seeing that yields were increasing. But again, it’s difficult to draw hard conclusions based on these figures.

Stephen Haggard: I believe the government has signaled that it was accelerating fertilizer imports in order to increase productivity, including particularly in the summer crops that are to be harvested shortly.

Go Myong-Hyun: From the look of it there seems to be an inverse relation between crop production and fertilizer imports, lagging by a year or two. The relation between crop imports is less clear, but the trend is similar.

NK Pro: Do you see any correlations? If not, what does this suggest about the impact of the weather?

Stephen Haggard: Correlations between output and the weather should be clear; correlations between output and what the media says represent a completely different question. At any given point, the government may or may not decide to play up natural disasters. I wouldn’t make as much out of this as something that correlated actual weather patterns with output.

Weather is bad, no doubt about it; but precisely because of such vulnerability North Korea should be more deeply engaged with the world economy, exporting things like light manufactures and raw materials and importing food. The reason it can’t do that is now related to the sanctions. But the sanctions are tied back to the nuclear program. The easy way out of these dilemmas is to reach an agreement and shift economic development strategy. The offers made at Hanoi were not serious on either side; let’s hope that changes.

Peter Ward: There does not appear to be any correlation between the two, and this indicates that weather as reported in the official and the actual harvest may not be related.

Benjamin Silberstein: No impact whatsoever. It seems that decisions on emphasizing, or de-emphasizing, poor weather conditions in state media are based solely on political considerations, and little on the state of things in the food economy.

Go Myong-Hyun: Not sure there is any relation. Maybe a better search term would be a compounded one= drought + flooding.

NK Pro: Does there appear to be any connection with the volume of water in reservoirs and the official crop production capacity?

Peter Ward: Reservoir levels may have an effect on the following year’s harvest, hence the drop in 2014 may explain some of the fall in the harvest in 2015, while the rise in 2015 could explain some of the rise in crop production in 2016. It’s unclear though from such a small sample, and it’s unlikely to explain all variation in crop production.

Go Myong-Hyun: As far as I know, North Korean agriculture depends more on irrigation than water reservoirs, unlike in South Korea.

Benjamin Silberstein: No, I see no such connection.

NK Pro: Does there appear to be any connection with the volume of water in reservoirs and state media reporting of drought and flood?

Peter Ward: There does not, but here it would be worth examining the context in which the words are being used for the relevant years, and also looking at how quickly irrigation reservoirs are likely to be depleted.

Go Myong-Hyun: I don’t see any correlation.

Benjamin Silberstein: Again, no connection as far as I can tell.


NK Pro: Overall, FAO/WFP have for long been reliant on official data provided to them by DPRK Ministry of Agriculture. Given the contextual presentation of the data, what would your conclusions be about the validity/ reliability of such data?

Benjamin Silberstein: Given what we know about data gathering conditions on the ground, we should most likely look at the numbers as a compromise of sorts between the DPRK government and the WFP. The WFP does corroborate much of the data with satellite imagery, so it’s not that all the data is constructed from political motivation. It’s the best we have and since we know a fair bit about the conditions under which the data is gathered, we can make a fair assessment of how good it is.

My overall impression is that the data is indicative, but that the numbers themselves matter fairly little. Corroborating reports from independent sources such as Daily NK is extremely helpful in this regard.

Go Myong-Hyun: We always have to take DPRK data with some skepticism. But were the regime intent on falsifying data they could claim a much higher level of crop production than officially reported: after all, the regime claims that they have reached self-sufficiency in food production.

So I think there is a kernel of truth in the official statistics – a pretty sizable one, actually. I think the official numbers are indicative of the real production level.

But the official stats are openly misleading about one very important source of crops in North Korea: private production from individual patches.

I think their share in overall production is way higher than before, but FAO/WFP figures don’t reflect this. If the official and private production figures are added up then the overall food situation is probably not too dire.

Peter Ward: I think the WFP-FAO should be more transparent, this would very much help their cause. They should try and maintain continuous time series of key variables, available to download and up-to-date for the current year.

Their data collection methods should also be clearly explained in a separate methodology note. The data coverage of certain variables is patchy, and they have seemingly avoided pointing out inconsistencies or implausible numbers.

The fact that a UNICEF-sponsored survey in 2017 found that over 50% of households have private agricultural lands but the WFP-FAO has not estimated private production for several years is a major issue. I am sure they provide reports and data in good faith, I am grateful that they do the work they do, and I think they need far more funding. But the inconsistencies in the data need to be better addressed.

Edited by Oliver Hotham