Given China’s limited land and an increasingly rich urban population that seems to like to eat, few things are more important to the country than productivity in the agricultural sector. The productivity miracle that has taken place in China’s factories is well documented – advanced machinery has been imported; designs have been begged, borrowed and otherwise obtained; millions of graduates have entered the factory workforce; and local governments have thrown their weight behind building industrial parks. As a result, productivity in the factories (producing more with the same inputs) has risen. And finally, in the last five years or so, real wages have risen as some of those returns have been paid to workers. Naturally, the first thing these folk spend their earnings on is better food – more meat, more oils, and more fruits and vegetables. Back in the fields, can the farmer parents of the new factory workers keep up? If they cannot, is China doomed to ever-higher food-price inflation?
There are various ways of increasing farm output; some are more easily achieved than others. On the one hand, the basic inputs can be increased in the following ways (i.e., extensive growth):
? Increase the amount of land under cultivation. However,

given competing demand for land from residential developers and industrial parks, this is more or less impossible (see On the Ground, 14 December 2010,‘China – Chen Xiwen on the future of land’). Most ‘reclaimed’ land these days tends to be of poor quality.
? Increase the number of people working the fields. This is also extremely difficult given the higher wages on offer in urban factories and services industries.
? Increase other capital inputs, including machinery, fertilisers and infrastructure.
On the other hand, one can attempt to increase productivity (intensive growth), for example in the following ways:
? Use fertiliser and pesticides more efficiently.
? Use machinery to replace labour.
? Use higher-yielding crops and disease-resistant animal breeds.
? Combine land plots to increase economies of scale and facilitate modern farming practices.
? Use better irrigation systems.
Of course, if you buy a tractor and then let it rust, or if you throw excessive amounts of fertiliser at the soil, this will not increase productivity. But if you do any of the above things intelligently, you should end up with much more output for a little more input. Because it is hard to work out exactly where the productivity is coming from, economists tend to talk about total factor productivity (TFP), which is made up of two broad factors: technical change (which pushes out the production frontier) and efficiency(which measures how far a farmer is from the production frontier). In today’s note, we lay out what the experts know about agricultural productivity in China’s fields. In short, the story is surprisingly positive – there has been, and apparently continues to be, productivity growth. Our concern, though, is about the future.
Productivity growth
A glance at China’s agricultural sector suggests reasons to be optimistic and pessimistic about productivity. On the positive side:
? A lot of government money has gone into R&D – mostly on grain until the 1980s, and on horticulture and livestock since then. China was using semidwarf varieties of rice before the Green Revolution elsewhere, and was the first country to use hybrid rice.
? Farmers are freer these days to decide what they grow – which in theory should mean that they use their fields more efficiently. The share of fields dedicated to grain fell from 76% in 1980 to 51% in 2005, and is almost certainly well below 50% now. Fruits and vegetables account for more than 5% of all farmland, compared to 2% in Europe and the US.
? There are also signs that farmers are upgrading their machinery. Sales of power tillers and belt tractor are falling, while sales of bigger geared tractors are rising (though tillers still d o m i n a t e sales), as Chart 1 shows. By the end of 2009, China hosted a population of 3.3mn of geared tractors, and is making about 300,000 new ones a year (of which some 14% are exported). The government is paying farmers to upgrade their equipment – in 2010, the Ministry of Finance put aside some CNY 15.5bn (USD 2.2bn) for subsidies, up from CNY 4bn in 2008. Buy a big tractor and you should get 30% of the cost back. One manufacturer told us that most tractors are bought and then hired out to neighbours.
? ‘Extension agents’ have disappeared. These agents, who share the gains made in the lab with farmers, were common in the Maoist era but have now all but disappeared. Without them, new seed varieties and better fertilisers and irrigation techniques are not transmitted to farmers. Instead, salespeople for fertiliser, seed and pesticide companies travel around the country marketing their own wares.
? Land agglomeration is proceeding at a slow pace (see On the Ground, 25 February and 10 March 2010, ‘China– Chongqing’s experimental land reform’, Parts 1 & 2, and On the Ground, 31 May 31 and 2 June 2010, ‘China –Four hero farmers’, Parts 1 & 2). With most land still being farmed in tiny parcels, it is not possible to use big tractors efficiently or to enjoy other economies of scale. Google Earth is a great way of showing the problem. Figure 1 shows a northern patch of Chongming Island, just north of Shanghai. On the left are some consolidated fields – note the long strips of land and the lack of farmhouses. On the right is the traditional mode of farming, which is still by far the dominant format around Shanghai.

What does the data say?
To cut through the anecdotal mists, Jin Songqing and colleagues analysed data on inputs and production for 23 commodities during the 1980s, 1990s and early 2000s (‘Productivity, efficiency and technical change: measuring the performance of China’s transforming agriculture’, Journal of Productivity Analysis, August 2009, by Jin Songqing, Ma Hengyun, Huang Jikun, Hu Ruifa and Scott Rozelle). They showed that China’s overall TFP growth was strong for vegetables and meats and not bad for grains, as Chart 2 shows. For staples, TFP was around 2% a year in the 1990s, and for horticulture and livestock, it was 3-5%.
The numbers also showed broadly where TFP growth was coming from. Most of it was from technological improvements. These mostly took the form of new grain varieties (both produced through domestic R&D and imported), plus new genetic material for hogs, dairy cattle and other livestock, which was often imported. In another paper, the same authors estimate that the production frontier for rice rose 2.3% y/y during the 1980-95 period, 2.5% for maize and 1.3% for wheat(‘The creation and spread of technology and TFP in China’s agriculture’, 2001, Jin Songqin, Huang Jikun, Hu Ruifa and Scott Rozelle). They show that since the early 1980s, some 400 different rice varieties have been available to China’s farmers, 25 per province. For maize, farmers completely change the varieties they grow every three years on average, rivalling turnover rates anywhere else in the world.(Another expert with whom we spoke was more sceptical, arguing that with 10,000 different seed suppliers, farmers have a very hard time finding good seed, or even the same seed they used last year. In other words, high turnover rates might not reflect farmer choice.)
There is also evidence of increasing specialisation. Scott Rozelle, who has been involved with much of the best work on productivity, reports the result of a 2004 survey in which he was involved (‘Rising demand, trade prospects, and the rise of China’s horticultural industry’, with Daniel Sumner, Mechel Paggi and Huang Jikun). 30% of rural communities reported that they were specialising in one or a few crops in 2004, up from 21% in 1995. And, in contrast to other countries, the authors found that poor farmers in areas far from cities were also specialising (a trend that was no doubt facilitated by better rural roads).
In contrast, efficiency lagged far behind – it fell for 14 of the 23 commodities studied. In other words, farmers were falling further behind the production frontier. The authors suggest a couple of possible reasons for this. First, the land newly brought under the plough may have been less fertile. Alternatively, they argue, the collapse of the extension system might have made it difficult to get the technology out into the fields. To address the problem, Jin and colleagues call for government investment to restore a working extension system.
China’s productivity growth is leading others’
Despite these problems, China’s performance relative to its peers is impressive. Keith Fuglie uses data up to 2006 from the Food and Agricultural Organization (FAO) to construct TFP growth estimates for the world (‘Is a slowdown in agricultural productivity growth contributing to the rise of commodity prices?’, Agricultural Economics, 39, 2008). Global agricultural output growth has been maintained at about 2% a year since 1960s. Global agricultural TFP growth was 0.9% during the 1970-89 period, he estimates, and 1.6% from 1990-2006. We show his numbers for China in Chart 3; it has been outperforming since the 1980s.
Alejandro Nin-Pratt and colleagues examine China and India in more detail, again using the FAO database (‘Comparison of agricultural productivity growth in China and India’, Alejandro Nin-Pratt, Yu Bingxin and Fan Shenggen, October 2009). In the 1960s, China and India had a similar input mix. But during the 1980s, China increased output at twice the pace of India, while its inputs rose only half as fast. In other words, China was more productive by far. We show some of the differences in Chart 4. In 2002-06, China’s crop and livestock yields were more than twice those of India, and it had one-third fewer labourers working per 1,000 hectares.(Output per farm worker grew 4% in the 1980s-90s in China, but barely 1% in India.) One secret behind this success is China’s far greater use of fertiliser, which has grown 3.5 times over the last 20 years. Another reason is China’s ability to get surplus workers off the land and into industry, thus naturally raising percapita agricultural output.
And now what?
In short, then, the story is a remarkably positive one. Past performance, though, is no guarantee of the future. China’s farms face a number of challenges:
? It is widely suspected that the amount of agricultural land continues to decline, despite official statistics to the contrary. New land put under the plough (mostly grassland and forest) is not as fertile as the arable land which is being absorbed by the cities.
? Water is a big issue in the northern plains of China, where aquifers are in decline.
? Farmers already use a large amount of fertiliser (and probably too much nitrogen). There is probably more room for potash use and broader improvements associated with more efficient application, but most of the gains associated with fertiliser have likely already been achieved.
? Agricultural land is being consolidated, but the pace is slow, and consolidation is harder to achieve with grains than with commercial crops such as fruits and vegetables (we will explore the reasons for this in a future note).
? There are few signs that money is going into employing and training extension agents, so efficiency will continue to lag.
That said, there are signs that output can continue to grow:
? A huge number of farmers’ collectives have been set up in the past few years. These help farmers specialise and improve the quality of their inputs and outputs. One possible problem, though, is that collectives are often set up by local officials, who may abuse their powers to forcibly consolidate land.? Use of machinery is on the rise –and government subsidies are promoting it.? Government R&D spending is still significant. For instance, much effort is being put into developing ‘local’transgenic seed varieties that would boost output, and these may be available in scale within the next decade.
To conclude: China will likely need to import more grain Much of the first 20, if not 30, years of reform brought big, positive structural change to the countryside. The breakdown of the commune system and the creation of the household responsibility system in the 1980s incentivised families to grow and sell their own food. This sparked a boom in productivity and output. In the 1990s, productivity rose as tens of millions of rural residents left the countryside, thus raising per-capita output. In the 2000s, the migration out of the countryside continued – and the elimination of agricultural taxes lightened the burden on those who remained. But in the 2010s, there seem to be only old people and children left in the countryside, and the land tenure system does not look likely to change anytime soon. This means that fruit, vegetable and meat farming should be able to modernise and achieve higher productivity (even if labour input costs present challenges), since these sectors do not need much land and – given urban demand – can afford labour. The creation of the commercial pig industry in the last three years is proof that economies of scale are possible.
However, we worry about grain. We are concerned that it will be hurt by the inability to combine land quickly enough, by an increasingly old labour force, and by ever-worsening water scarcity in northern China. ‘Accelerated’ urbanisation will take more land away. Transgenic varieties will help to raise yields, but it appears ever more likely that the grain autonomy policy will need to be relaxed.
It is very hard for prices of manufactured goods to rise in China. Intense competition forces manufacturers to achieve efficiency gains through mechanisation, economies of scale and leaner management practices. But at the same time, one only needs a small plot of land and a bank loan to build a factory and dramatically increase output of, say, televisions. In other words, when it comes to manufactured goods, productivity and its inputs can be increased.
The same is not true for grain. Land is an essential input, and one can neither create land, print it, nor borrow its design from a competitor. So, to some extent at least, China’s demand for grain – assuming it continues to grow – will need to be met via imports. Corn appears to be the most obvious candidate; Brazil, Argentina and the US will be the suppliers. In other words, China will need to import these countries’ land, water and higher agricultural productivity. This is not a bad bargain, since North and South Americans will import manufactured goods from China in return – but it will be hard for some in Beijing to swallow.
(Author: Economist, Standard Chartered Bank)