Melamine, China, and Expecting to Be Poor

China has sentenced Geng Jinping and Zhang Yujun to death and Tian Wenhua to life for their roles in the melamine-tainted milk deaths of six babies and the sickening of at least 300,000 others. Geng used melamine, a toxic industrial chemical, to fool regulators by increasing the protein content of watered-down milk. Zhang ran a workshop producing melamine-tainted powder sold as a protein enrichment. Tian was a dairy boss who for months delayed notifying regulators that her company's products, including baby formula, contained melamine.

Reporting for the AP, Anita Chang writes that deficient official oversight contributes to chronic food quality and safety problems in China. (Various versions of Chang's story were picked up. One is here http://agweekly.com/articles/2009/01/22/commodities/dairy/dairy28.txt, another here http://news.yahoo.com/s/ap/20090122/ap_on_re_as/as_china_tainted_milk/print). Chang contrasts the Chinese system in which milk comes from a "patchwork of producers" with the United States dairy industry in which "dairies run farms with thousands of cows and are better able to control quality." This comparison implies that Chinese food quality and safety could be improved with better official oversight, and that centralizing producers would make official oversight easier. Whether or not centralization is necessarily a good idea will be the main point of discussion herein.

In a very basic analysis of the situation, there are two ways to improve official oversight. The first is to change the production system to better fit the existing oversight system. The second is to change the oversight system to better fit the existing production system. Chang implies that, in China, the system of production, the patchwork of producers, should be changed to better fit the existing oversight system. These two systems can be thought of as either centralized or distributed. Chang characterizes the existing Chinese dairy production system as decentralized and the oversight system as centralized and implies that the production system should be centralized.

Two consequences of greater centralization will be discussed. The first are the political impacts of increasing the size and importance of primary producers. The second is the impact on the supply stream. Centralization has political consequences in terms of the power relationships between industry and government. Centralization makes producers more politically powerful by increasing their importance to the food production system and their wealth. Increased importance leads to an obligation of government to avoid infringing on the operations of firms deemed "too big to fail", embodied in the notion that "owing the bank $100 is your problem; owing the bank $100,000,000 is the bank's problem." If the closure of a dairy producer would disrupt the supply chain and consumption of dairy products, that becomes the government's problem, and that producer has a strong bargaining position. Reducing the number of producers also promotes monopoly power. The increased wealth from the monopoly power can be used for lobbying, soliciting government for preferential treatment. In these ways, centralized production can actually weaken the efficacy of government.

The second issue of greater centralization addressed here is the impact on the supply stream. Centralized production means that a relatively small number of producers are responsible for the bulk of production. This amplifies the supply-stream impacts of a problem at one producer. If one producer has a problem with quality or safety, its impact will be felt by many consumers due to the large amount of product coming from that producer. In a distributed production system, each producer contributes a small part of the supply stream, so amplification would not have as large an impact. Amplification is a significant drawback to centralized production and has shown up in some recent US food scares in centralized industries, e.g. salmonella on tomatoes and E. coli on spinach, in which problems at a small number of large producers resulted in widespread contamination.

But this begs the question of why the melamine contamination became such a large problem in China. After all, Chang says China's dairies are a "patchwork of producers", presumably distributed. If a distributed production system dampens the impact of amplification, how then did the melamine contamination in China come to be a national problem and sicken hundreds of thousands of babies? The answer to this question requires examining where melamine entered the Chinese dairy production process.

Investigations into the incident found that middlemen between milk producers and dairy companies were primarily responsible for the melamine contamination. These middlemen watered down raw milk and then added melamine to artificially increase its protein content. A low protein content would have tipped regulators that the middlemen were watering down the raw milk. Melamine was not being added by producers of the raw milk, presumably what Chang would refer to as the producers in the "patchwork of producers", but instead by consolidators of the raw milk. The point at which melamine was added was therefore a point of centralization in the production process and served to amplify the effect of the contamination. Chang's representation of the Chinese dairy production system as a "patchwork of producers" is somewhat misleading. The "patchwork of producers" of raw milk was not where the melamine problem occurred.

If this is true, how would increasing the size of individual producers, as Chang suggests is what safeguards the US system, guard against an incident like the melamine contamination? Here a subtle distinction in Chang's article must be noted: Chang attributes US quality and safety to the dairies themselves rather than to official oversight. Again, Chang writes that in the US "dairies run farms with thousands of cows and are better able to control quality." Here Chang minimizes the importance of oversight in the US dairy system, saying instead that the dairies themselves are controlling quality. But, as found in the Chinese melamine investigation, it was not the dairies that were responsible for the contamination. It was people and firms located at points of centralization between primary producers—the dairies—and end-product producers. The producers of raw milk had no incentive to add melamine to their product, because the only reason to do so was to hide from regulators that milk had been thinned. So how, then, would increasing the size of the dairies in China prevent another melamine incident or one like it?

Increasing the size of individual producers could reduce the middlemen involved in the production process and eliminate a point of entry for contamination. In the Chinese example, middlemen consolidated raw milk from multiple small producers in order to gather a large enough volume to sell on to other firms like baby formula producers, firms that need large amounts of milk on a daily basis. Middlemen had an incentive to use melamine that primary producers did not. If raw milk producers were capable of providing large amounts of milk to firms directly, the need for middlemen may be reduced. Official oversight of the producers could then be feasible. But notice how this does not make official oversight any different. Rather, it is the reduction in middlemen, the reduction of the number of people and firms involved in the production process, that would make official oversight possible. This is an example of changing the production system to fit the existing oversight system.

In China, centralization and control are central components of government, while in the US government is more distributed across states and also fragmented internally into branches. Chang's argument that centralized production would enable better official oversight in China is valid when it is evaluated from a perspective resting on the assumption that the production system should be changed to increase the efficiency of the existing oversight system. If Chang's argument is approached with different assumptions, such as that centralized production should be avoided or that government itself should be changed rather than only the production system, it appears weaker.

China's mode of government and of development is oriented toward centralization, be it with population growth—a great migration from rural to urban areas—or power production—the construction of the Three Gorges Dam that will provide about 3% of the country's electricity. China will choose to centralize production to fit better with centralized government before it will decentralize government to fit a distributed production system.

In other contexts, decentralized government regulation of distributed production may be more politically acceptable. The advantages of a distributed production system arise from the system's inherent redundancy, yet this redundancy is antithetic to a mode of government regulation that rewards economies of scale. Economies of scale refer to the advantages that producers can gain when they increase in size. The most important competitive advantage gained from economies of scale is the ability to sell products more cheaply than smaller competitors. Economies of scale only work in certain regulatory and business frameworks. In capitalism, the system employed in both the US and China, economies of scale are very rewarding, and most firms strive to achieve them. However, economies of scale can also lead to monopolies when one firm becomes so large and its pricing so powerful that it drives all competitors out of business. This situation develops in capitalism absent of regulation and led to a crisis in the US in the early 20th century. This crisis resulted in the creation of anti-trust (anti-monopoly) laws that limited the extent to which firms could harness economies of scale to shut down all competitors.

In the past half-century, China has undergone massive centralization. Millions of people have been removed from agricultural livelihoods and centralized into cities; the Yangtze River has been diverted for the past decade as the Three Gorges Dam—the world's largest—is constructed; and censors block information from flowing into and out of the country. Centralizing the country's dairy industry fits into this framework and will almost certainly be pursued. The government likely will not miss the opportunity provided by the melamine scandal to eliminate an entire level of middlemen.

However, there is danger in this path. Centralization increases the efficiency of production in part by reducing the number of employees. This reduces jobs. China has a large population that has been removed from labor-intensive agricultural work and must now be supported by industrial and service jobs, and the existence of these jobs requires continued, constant growth in consumption of goods and services. The era of centralization in the US in the early 20th century led to a tremendous over-capacity to produce goods, which in turn contributed to the Great Depression when there simply was not enough demand to satisfy the productive power of world industry. Reviving demand required more than 10 years as well as a World War. Also of concern is the timing of China's growth with the world economic crisis. The US unemployment rate is now expected to approach if not exceed 10%. The population of China is more than 1 billion people. Ten percent unemployment in a country that large could mean 100 million people unemployed. The vast majority of these people would be living in newly-created cities and would have memories of agricultural life that had been taken from them, agricultural life where their purpose made sense and where they could provide for their basic needs with their own labor. This is not to paint a romanticized vision of Chinese agricultural peasantry. That life likely was and still is hard work that destroys the body. But it is not hard to imagine dissatisfaction growing in these people with the promises of their government for a better life in the city falling flat.

The melamine contamination is an interesting starting point for thought and discussion on the impact of centralization and on the correct balance between centralization and distribution in production systems, as well as the role of government in all of this. Contrasting the Chinese and US cases can provide insight, but is also rife with pitfalls as overgeneralization and conjecture abound. However, the present economic crisis may mark the endpoint of one of the great periods of centralization. Entire generations of people around the globe only know economies that grow at extraordinary rates, only know societal change that advances with great speed, and fully expect their own lives to mirror the change and increase in affluence that older generations went through. The realization that these expectations will be unfulfilled may lead to social problems so bad that they make an economic crisis ten times worse than the present one seem preferable.