The evidence of China’s compliance with North Korea sanctions is still mixed.

This week, there has been much talk and excitement about a new study, by the new blog Beyond Parallel, analyzing satellite imagery of six select sites along the Chinese-North Korean border, and finding evidence of a recent decline in bilateral trade. From this, the study concludes that China may be (as Josh Rogin paraphrases it for The Washington Post) “Beijing has been quietly punishing Kim by cutting off the flow of funds to his regime.” Here are the study’s two main findings:

First, the satellite images indicate a substantive reduction of economic activity on the Sino-North Korean border measured by the fewer trucks, trains, and boats in the February 2016 image compared to a similar timeframe in 2015. [….] In the aftermath of North Korea’s January 2016 nuclear test, this observed downturn in activity was comprehensive across customs areas, railway, and road traffic.

Second, the images also suggest that independent Chinese actions were taken to reduce trade in this region after the nuclear test and prior to China’s signing on to UN Security Council Resolution 2270. These findings run contrary to some estimates that Sino-North Korean trade (particularly Chinese exports) increased in the first quarter of 2016, and might confirm large anomalies in trade data as reported by China’s customs statistics, KOTRA (Korea Trade-Investment Promotion Agency), and other organizations.

The study is interesting and data-driven, and every North Korea-watcher should celebrate the launch of any new information source that promises this kind and quality of analysis. What’s more, by analyzing the volume of traffic at multiple sites over several months, the study is less vulnerable to the regime’s manipulations than the satellite theater that is almost the only good reason to read 38north (the contributions of J.R. Mailey and Andrea Berger being two other notable exceptions).

It’s especially tempting to feel triumphal about Victor Cha’s conclusions that China has taken “unilateral measures to drastically curtail trade interaction along their border,” and that China is “squeezing [North Korea] more than we were led to expect.” Still, the evidence and my objectivity restrain me to say, “Not so fast.”

First, there is also substantial evidence that China is still violating key provisions of the sanctions to prop North Korea up. North Korea’s most important export by reported volume is coal, followed by other minerals, and as NK News’s invaluable Leo Byrne has noted, the trade in sanctioned minerals continues. To some extent, North Korea has shifted its coal exports to other avenues, including Alibaba.com. At the land border, trucks loaded with titanium are still crossing into China. Worse, North Korean ships that have been specifically designated by the U.N. are still operating, and in some cases, are coming very close to Chinese ports they aren’t even supposed to approach. Then, their transponders go dark. This suggests that those ships are either landing in Chinese ports or off-loading their cargo onto smaller vessels without landing. Both alternatives violate UNSCR 2270.

Second, the kinds of commerce that benefit the regime most (as opposed to market trade that benefits the North Korean people) aren’t easy to measure with satellites. North Korea’s other lucrative exports include gold, weapons and weapons and technology, and labor. Its most essential imports include bulk cash, wire transfers, gold (again), and luxury goods that come in on Air Koryo. It probably also earns significant revenue through tourism. These are not things that can be measured by counting railcars.

Third, the study focuses on overland trade but tells us little about maritime trade. If the authors of the study want to improve the utility of this project — and I emphasize that it’s potentially a very valuable one — it should also examine maritime traffic to and from the key North Korean ports of Nampo and Sinuiju. Maritime trade is more likely to be under the control of, and to the immediate benefit of, the regime. It should specifically look for trade in bulk cargo like coal, imports and exports of fuel, and the movement of designated ships (it’s possible to match IMO numbers from transponders with satellite images).

Fourth, there may be other explanations for Beyond Parallel’s observations. I’ve long felt that Korea-watchers were far too trusting of officially reported statistics on China’s trade with North Korea, and the case of China’s fuel exports to North Korea illustrates just how easily China can manipulate those statistics. But to the extent we believe those stats, they do show a significant decline in North Korea’s exports over the last six months. The problem with attributing this to sanctions is that this decline extends a trend that we began to observe earlier, particularly in the mining industry. In fact, at Benjamin Katzleff Silberstein has pointed out on several occasions, this decline in trade volume has a closer correlation to the decline China’s economy than it has to sanctions. An interesting question is whether China’s own internal market controls, including its restrictions to prevent capital flight, may be playing a role, but that question is beyond the depth of my knowledge of economics (anyone? Bueller?).

Other potential causes of a decline in bilateral trade include regime-driven trade and travel restrictions leading up to the party congress in May, and problems with North Korea’s infrastructure, such as the partial collapse and subsequent repair of the Sino-Korean Friendship Bridge last October. (The effects of this may or may not have ended before Beyond Parallel’s study began.) That would also help explain why the study found that trade began to decline before U.N. sanctions were increased in March.

Finally, we should not hope for China to enforce sanctions in unilateral ways that depart from the strict letter of the U.N. Security Council resolutions, whether by under-enforcing or over-enforcing sanctions. The main reason sanctions haven’t worked thus far has been — and continues to be — China’s under-enforcement of sanctions. That is why Congress decided that secondary sanctions were necessary to force China to comply, by dividing the interests of China’s fundamentally hostile government from those of its more pliable banks and industries, which need access to American markets. But what we don’t always realize is that sanctions over-enforcement is an equal danger. This veers off onto a long tangent, so I’ll save it for tomorrow’s post.