“Famine in North Korea”: An Interactive Review (3 of 3)

[OFK:  In this post, Marcus Noland and Stephan Haggard respond to  Part 1 and Part 2 of my  review of their book, “Famine in North Korea:  Markets, Aid, and Reform.”]

Josh Stanton has written by far the most thoughtful and cogent analysis of Famine in North Korea that we have seen to date. Stanton’s review is generous, but also raises important questions about virtually all elements of our analysis. In the interest of furthering both scholarly and policy debate, we address the questions that Stanton raises in some detail and roughly in the order that they are raised in the book itself.

Causes and Culpability

In thinking about the causes of the famine, we were interested in the first instance in the interpretation offered by the regime itself, namely that the floods of 1995 were responsible for a catastrophic decline in the aggregate supply of food, both through the loss of output and through the destruction of some stored stocks of grain. This claim is risible: the famine was already underway in the summer of 1995; indeed, North Korea had negotiated bilateral assistance agreements with Japan and South Korea and the first shipment of aid was already en route when the floods hit.

Those familiar with Amartya Sen’s pioneering work on famine know that this interpretation falls within a more general class of explanations that focus on the decline in the overall supply of food: people starve because there is simply not enough to go around. Sen, by contrast, found that a number of famines he studied occurred not because of a decline in aggregate supply, but because food was badly distributed. In particular, Sen focused on cases where the poor starved because they were incapable of commanding the resources to buy adequate food in the face of spiraling prices. He characterized such settings as examples of “entitlement failures” in which distribution, rather than aggregate supply was to blame.

Curiously, Stanton makes the case in favor of a distributional failure more categorically than we do. Our argument is somewhat more complicated and consists of two parts. First, we demonstrate that there was indeed a secular decline in the aggregate supply of food, but this decline began well before the floods of 1995. To blame the floods for the famine is a little like blaming the last batter who strikes out for the loss of a baseball game. Just as the lack of runs cannot be blamed on the last batter alone, so it seems misleading to us to blame the loss of output in 1995 as the underlying cause of the famine when output had been declining for five years. We do not claim that the floods were unimportant; to the contrary, they were devastating. But their impact must be placed in context.
We attribute the declining availability of food in the first half of the 1990s to a combination of external shocks (the collapse of the Soviet Union) and the particular agricultural strategy the country had pursued in preceding decades. We also note–and this is a crucial part of the analysis–that the regime could have compensated for these shortcomings if it had made a number of external adjustments, such as expanding exports, maintaining a capacity to borrow, or seeking aid earlier. We do not therefore deny the importance of declining supply, but underline that shortages were not the result of natural events alone or even primarily, but had important policy causes.

By the early 1990s, the supply situation was therefore already precarious. If famine were to be avoided in the absence of the external adjustments just noted, there was very little margin for error: food would have to be distributed with almost uncanny care. Suk Lee has argued that the regime did in fact try to spread supplies evenly across the country, and that things might have even been worse in the absence of these efforts. He also notes quite rightly that the breakdown of the logistics infrastructure of the country–for example, the absence of fuel to move grain across provinces–played a role and we agree.

We are very skeptical that distribution was neutral, however. There seems to be ample evidence from data provided by the North Korean authorities themselves that Pyongyang was privileged in the allocation of food, for example. There were two distinct forces that gave rise to this unequal distribution of supply.  First, we argue that as shortages deepened, local officials faced incentives to hoard, and the inter-provincial system of transfers broke down, leaving the historically food deficit northeast exposed.  On top of this decentralized system-fraying there is convincing evidence that the regime also sought to discourage aid to the hardest hit parts of the country. We doubt that they did so to punish deliberately those regions, which in addition to the politically suspect, included large concentrations of the regime’s urban proletariat base as well. Rather, they did so to shore up supplies in even more favored population centers (thus the charge of “triage”). For these reasons, we argue that issues of distribution were central to the observed outcomes.

Stanton raises the question of culpability, and suggests that we are uncertain or divided between us on it. This is not the case, although our thinking on this question has since been advanced by reading the excellent work of David Marcus on famine crimes (The American Journal of International Law, Vol. 97, No. 2. (Apr., 2003), pp. 245-281). We do not argue that the regime set out to starve the people who ultimately died as Stalin did in the Ukraine (with the important exception perhaps of those incarcerated in the country’s vast gulag). But we do believe that the government both showed favoritism in the distribution of food–mostly across regions–and failed to anticipate the ultimate effects of the stubborn pursuit of self-sufficiency as a means of attaining food security. The intent to kill is not required to show criminal negligence, nor to rightly categorize the North Korean famine as a crime against humanity.  All that is required is that the authorities could have plausibly known the effects of their actions–and inaction. As Stanton very rightly notes, the failure to fully appreciate the costs of self-reliance is what makes the work of analysts such as Christine Ahn and John Feffer so disturbing.

The Death Toll

One of the most curious responses we have had to our book is that the death count we provide is too low. This response is particularly odd when no real defense is given of higher estimates beyond the fact that someone else has uttered them. In particular, the number of 2.5 or 3 million excess deaths has become a kind of focal point: anything lower is portrayed as a kind of betrayal. Yet there are quite obvious reasons why that number is almost certainly too high, and Stanton actually reviews them quite clearly.

We know from a variety of different types of evidence that the toll of the famine was particularly high along the East Coast and in the Hamgyung provinces in particular. If we take death rate estimates from refugees–a particularly distressed population from these particularly distressed provinces–and project them across the entire country (minus some share of the population which was shielded from the famine), we can get to the 3 million number. But that is quite clearly an implausible extrapolation, particularly in light of the evidence that we present that other provinces almost certainly did not fare as badly. To draw a contemporary analogy, it would be as if we took the deaths from civil conflict in Darfur as a share of the population of those provinces and then constructed an estimate of total civil war deaths in the country by projecting that share onto the entire population of the Sudan.

While recognizing that there are a variety of estimates of the famine toll, we find the estimates from two independently conducted studies to be the most convincing: one by US Census Bureau demographers Daniel Goodkind and Lorraine West (Goodkind, Daniel and Lorraine West. 2001. “The North Korean Famine and Its Demographic Impact. Population and Development Review 27, no. 2: 219″“38.) and the other by economist Suk Lee (Lee, Suk. 2003. Food Shortages and Economic Institutions in the Democratic People’s Republic of Korea. Ph.D. diss., Department of Economics, University of Warwick, Coventry, U.K.).

As Stanton correctly observes, both studies are based on commonly used age- and gender-specific population models. These models work off some simple relationships among population size, birth rates, and death rates. As illustrations, if we know population size, and the birth and death rates at a particular starting point, we can project forward population growth; if we know population magnitudes at starting and ending points, we can infer the death rate (assuming the birth rate has not changed). If we know the population size at the starting and end points and the birth and death rates, we can measure “excess deaths” if the terminal population falls short of the projection. Goodkind and West used information from the 1998 WFP nutritional study and the Great Leap Forward to calibrate two variants of their model; Lee relies entirely on DPRK official statistics. Stanton raises a fair point about the application of the Chinese experience to the North Korea case, suggesting that it might bias downward the estimated toll, though the use of the 1998 WFP nutrition survey–the one which reported a staggering scale of malnutrition that was not reproduced in the prior or subsequent surveys–would appear to cut the other way. Similarly one can reject Lee’s finding precisely on the grounds that the underlying official statistics are themselves unreliable.  But it seems to us that as imperfect as they may be, two independently conducted studies employing widely accepted methodologies arriving at a common conclusion are more persuasive than simple extrapolations. It may well be the case when future scholars have access to the DPRK’s internal documents that our preferred estimate will be revealed to have been far too low. But for now, the needless deaths of 3-5 percent of the population are more than enough to make the North Korean famine one of the worst of the 20th century.
On Aid

Only in the discussion of aid and diversion did we feel that Stanton missed a key point we were trying to make in Famine. A central question we ask in the chapter on diversion is “what happens to diverted food aid?” Stanton–and he is not alone–seems to assume that it disappears from the total food supply. This would be true if all food aid were consumed or hoarded in stocks by the military.  That at least some was stockpiled in this way cannot be ruled out. But we argue that a substantial share of diverted food in fact finds its way into the market where it has the effect of increasing supply and lowering prices from what they would otherwise have been. This diversion has important distributional consequences: intended beneficiaries of food aid do not get it all, or get it only by paying for it. But the additional supplies nonetheless expand total consumption. It is therefore not necessarily a mystery that a high level of diversion could co-exist with an improvement in the overall supply picture and an avoidance of the recurrence of famine. Yet as we also argued, diversion almost certainly implies increasing inequality in the distribution of food, as traders and politically-connected intermediaries and cronies benefit at the expense of the vulnerable population.

Stanton raises some intriguing questions with respect to the relation between aid and markets.  Clearly the availability of aid affects prices in the markets, as have other shocks such as the closure of the Chinese border during the SARS scare.  However markets in North Korea remain fragmented. On a given day prices for even relatively homogeneous commodities differ across locations–with the price of food consistently higher in the northeast than elsewhere.  Trying to better calibrate such relationships is an issue of ongoing research.

Engagement

On no question have we been more taken to task than our position with respect to engagement. On one side, critics like Stanton say that we are idealists falling into the same old trap of “hoping engagement will induce Kim Jong Il to make modest and gradual reforms. He appears to support a more radical course of action, such as cutting aid dramatically, on the grounds that we don’t know how much aid is getting through in any case (see the discussion on diversion above; again, the point made there is germane to Stanton’s case) and that we simply encourage Kim Jong Il’s bad behavior.

On the other side–and   particularly in South Korea–we have repeatedly been criticized by defenders of the Sunshine Policy as hawks and apologists of the Bush administration. Critics have said that to discuss human rights or the famine is likely to be counterproductive and only undermine prospects for a negotiated settlement to the nuclear issue. Our doubts about the efficacy of aid in achieving political or economic objectives are taken as a frontal assault on engagement. 

In fact, we believe our position is quite coherent. First, we wrote the book because we believe that the outside world should not be silent on these issues. We find it disturbing, even shameful, that there is such a limited public discussion in South Korea of the human rights record of the regime.

We are also highly skeptical of the argument that anything more than short-term political concessions are likely to be achieved by offering aid as a quid-pro-quo. The record of engagement as a strategy of quid-pro-quos is much more dubious than is typically thought, and may even confirm to the regime that they can sell their bad behavior. And in more recent work on North Korea’s external economic relations, we explicitly make the point that non-commercial interactions with North Korea do not have the transformative effect on North Korea that defenders of engagement claim. Rather, those effects are more likely to come from more strictly commercial relations, such as those that increasingly characterize North Korea’s economic integration with China (see Stephan Haggard and Marcus Noland, “North Korea’s External Economic Relations,” Peterson Institute Working Paper 07-7 available at http://www.iie.com/publications/interstitial.cfm?ResearchID=794.

Yet we also have pretty unambiguous evidence of the effects of attempting to isolate North Korea or engaging in domestic posturing about regime change. This approach suffers from a fatal coordination problem: China and South Korea will not go along, and they have demonstrated the capability of offsetting pressure that the US and others may bring to bear. In practice, the strangulation approach had a three year run, and failed utterly: a crisis over a suspected enrichment program–unambiguously a violation of North Korea’s commitments–ended   up in North Korea’s testing of a nuclear device and a serious threat to the integrity of the NPT. Advocates of isolation and regime change have to come to grips with the fundamental geopolitical constraints under which US policy toward North Korea operates. And while we yield to no one with respect to our antipathy toward the Kim Jong-il regime, we see no other realistic option than to negotiate with it.

We also believe that the various shortcomings of the international aid program do not constitute the basis for a termination of all aid to the country, particularly if supply conditions are precarious. As Stanton rightly quotes us, to take this position–supported by some refugees–is to gamble that a policy that has not succeeded in the past will miraculously have effect, while risking the lives of those deprived of food and assistance in the meantime.

Rather, we support the use of aid, as one of a number of other policy instruments, to encourage the transformation of the regime into one that is less threatening externally, and less repressive internally. Food security for the North Korean people ultimately will only be achieved through a functioning economy which means deepening the institutionalization of the market economy and opening to the world. In the short-term, North Korea may continue to rely on aid, but our objective should be to wean them off the international dole.  When humanitarian concerns are not pressing, then aid should be reduced and North Korea should be induced to engage in reforms that will allow it to purchase food on its own. When humanitarian issues are pressing, the outside community should coordinate to the best of its ability and aggressively raise issues of transparency, access, and monitoring; we have long argued that the nature of the bilateral food aid programs of China and South Korea undermines these objectives. Negotiations on economic issues should not begin and end with a discussion of aid and what North Korea will get: rather, they should also discuss questions of reform and opening that will make aid effective, facilitate commercial exchange and socialize North Koreans to engage in the outside world.

This policy does not make for easy soundbites. Advocating them does not necessarily mean that we believe that they will work; dealing with North Korea always involves difficult choices among unappealing options and thus a high degree of frustration. But the other options–containment, isolation, a cutoff of assistance–have had little effect and may have even strengthened the regime.