Yesterday I blogged about agricultural productivity in Thailand. I included a chart which compared rice yields in Thailand with yields in China, Japan and Korea. Some readers have pointed out, both in comments on the post and by email, that the comparison between Thailand and Japan/Korea may not be particularly informative given quite different technical, social, ecological and agronomic conditions. Chris Baker asks ”what happens if you put Malaysia, Philippines, and Indonesia on the chart – countries with climatic, social, and technological conditions relatively similar to Thailand?” Here is the answer:
The key points from the data are:
- Of the five countries shown, Thailand has the lowest yields. Surprisingly, the IRRI data suggest that yields in Laos have been higher than Thai yields since the early 1990s. Is this plausible?
- In the 1960s, Thailand and Indonesia had very similar yields. But Indonesian yields increased rapidly in the 1970s and 1980s. Since 1990, Thai yields have, on average, been 44% lower than Indonesian yields.
- In the 1960s, the Philippines had lower yields than Thailand, but the situation was reversed in the mid-1970s. Since 1990, Thai yields have, on average, been 20% lower than yields in the Philippines.
- Malaysia’s productivity trajectory is closest to that of Thailand, with modest productivity growth from a somewhat higher base. Since 1990, Thai yields have, on average, been 20% lower than yields in Malaysia.
- (Not shown on the graph – yields in Vietnam were similiar to those in Thailand until the mid-1970s; since then they have grown to a level slightly higher than Indonesia.)
The comparisons within Southeast Asia are not as stark as the comparisons between Thailand and north Asia – but Thailand’s lackluster performance is evident.
Chris Baker has provided some data (from the Thai Office of Agricultural Economics) that are slightly different to the IRRI data, and show a healthy jump to over 3 tonnes per hectare in 2008. One reader has suggested, via email, that recent increases in Thailand could be the result of rice price guarantee schemes that have encouraged farmers to spend more on fertiliser for rice production.
Thanks very much to those of you who have made very helpful contributions on this topic. I look forward to more discussion of this important issue.
Andrew, thanks. Much more interesting chart. The Thailand performance is pitiful. What is striking is the complete stagnation over the 20 years 1971-1991 while both Laos and Indonesia increased yields by 50 percent. Since 1991, Thailand at least shows some growth.
I wonder why the IRRI and OAE figures are different? Where else would IRRI get data from? I confess I was given the OAE figures and I haven’t checked them, but the donor was reliable.
Remember this time last year, during the rice price spike, Dhanin CP was saying: let us manage the irrigated paddy area under contract farming, and we’ll double yields. Looks feasible. The question is why is that not attainable under owner cultivation?
How does Ban Thiam fit with this picture. Above or below the national average yield? Same trajectory?
Are there no data for Vietnam? Surely it should be on the chart.
Quality comment or not?
0 
0
It is quite understandable that Lao’s (or other SEA countries’) rice yields be much greater than Thailand’s. Have a look at the proportion of rice production in irrigated vs upland conditions.
http://www.irri.org/gis/ricedist/maps/rice_asia_2005.png
http://www.irri.org/science/ricestat/data/may2008/WRS2008-Table30.pdf
Quality comment or not?
0 
0
One possible explanation is that not all kinds of terrains are suitable for rice growing. Indonesia has volcanic soils which are well known to be quite rich. In Thailand, the best rice growing plots are in the central plains, which have already been maxed out with rice growing decades ago and are progressively turned into suburbs areas. Rice growing on more marginal areas with lower yield may then weigh down the national average, especially if they can’t reliably grow more than 1 crop per year.
Also, the strains of rice may be a factor. Do you have information on yield of jasmine rice, or other varietals? If there’s a high premium for one varietal, that will offset the lower yield. Thailand has always had a surplus of rice, so what matters is less how much rice is harvest but how much monetary return the farmers get for what they put in.
Quality comment or not?
0 
0
From the information that I got from one of my uncle in Phan, Chaing Rai, Rice productivity in the area is a lot higher than the data shown in the post. Rice productivity of marginal land in that area is between 3.5-4 tone/hectare/crop whereas in good farmland farmers can get up to 6-7 tone/hectare/crop. Most famers in that area grow sticky rice, Gor Khor 6.
Quality comment or not?
0 
0
OAE data, like the Dept of Agricultural Extension ones come from interviews with village headmen or other “key” farmers who are supposed to know everything about the area cultivated, costs of production, yields, etc. in their village/tambon/amphoe area. In theory, officials of the OAE or DOAE are supposed to conduct serious stratified-sampling surveys, but I was told there’s quite a bit of “chair surveys” (sorry, I forgot the nice Thai expression for that), meaning lazy officials do not always do the interviews. As a result, the DOAE data showed exactly the same yields for all tambon and amphoe in my region of study and for several years in a row.
In other words, for infra-national geographical units, we’ve got to work with these data, but cannot blindly trust them.
The most comprehensive source of national-level data remain the FAO ones (same as IRRI).
To understand the low yields in Th, one must take into account the fact that 40% of Thailand’s is cultivated – a huge proportion in regional and global terms. That 40% is not solely concentrated in fertile, well-irrigated and drained areas and does include a lot of less well endowned regions. This fact has of course a direct result, diluting high yields with lower ones (as noted by Bystander). Another implication is that the 40% is to a significant extent occupied by farmers who – for perfectly understandable reasons – consciously decide not to aim at the highest possible yields, and rather aim at 1) minimizing risks of losses and debts in commercial rice cultivation and 2) concentrate their meagre capital resources into more economically rewarding and safer investments, either in other more lucrative crops, or more likely in non-agricultural endeavours (including education for their children).
I strongly believe the best way to augment yields in Thailand is not to aim at large capitalist farms – it’s been proven again and again that owner-occupiers in medium-size farms get the highest returns/ha and yields. Rather, making on-farm income stable and meaningful (i.e comparable to wage labor) is probably the only way yields on these types of farm will significantly improve. To do so, the key is probably in implementing a good social safety net and a crop insurance system for farmers (or any other policy change which would meaningfully protect all farmers from price and weather fluctuations).
Quality comment or not?
0 
0
Thanks for the comments. Just a couple of quick responses (more next week).
On Vietnam – I will add it to the graph next week and re-post it. I ran out of time yesterday.
On Ban Tiam (the village where I have been working in northern Thailand). Rice yields there are well above the national figure – between 4 and 5 tonnes per hectare. I don’t have any figures on trends in yields over the past 10 or 20 years but the commonly expressed view is that yields have significantly increased as a result of improved varieties.
Quality comment or not?
0 
0
Jean-Philippe, how do FAO/IRRI get their data?
Quality comment or not?
0 
0
To be honest, I am not sure how the FAO get / produce all their data, so if somebody else knows better I’ll be happy to have more info.
What I know is that the FAO (or its major statistical partners; World Bank, Eurostats, etc.), send out questionnaires to the relevant statistical institutions in each country. For Thailand, it means that several organisations and ministries are involved because the stats collected pertain to agriculture, forestry, fisheries, etc. For agricultural production data (and not agricultural population), the data sent to FAO comes from the Office of Agricultural Economics.
Once the “raw” data has been collected, it is evaluated by country experts, adjusted to fit international definitions and standards (in forestry, they would for example add rubber plantations to the forest cover), and, when the data is believed to be erroneous or hasn’t been transmitted, it is corrected or replaced by expert guestimates.
Over the last few years, the FAO has made great efforts to improve the transparency of their data collection and adjustment procedures. I’ve seen quite detailed reports on the production of forestry data. For agricultural stats in Thailand, the relevant info is available from
1) “Metadata for National Agricultural Statistics”, http://faorap-apcas.org/rdes/PPT/thailand_metadata.pdf (which I just found and haven’t read yet)
and
2) FAO (2002) Proceedings of the national seminar on the system of food and agricultural statistics in Thailand. Bangkok: FAO RAP. available at http://faorap-apcas.org/thailand.html
One last thing, I believe the OAE data and enumerators are more reliable and serious than those of the DOAE. The problem with OAE – I believe – is their low sampling size in each agro-ecological regions. There’s huge variability within agro-ecological zones and I’m not sure they were or are currently able to get representative samples. As for the NSO agricultural census, they are not that reliable, in part because use untrained personnel (teachers mostly). Have a look at the previously mentioned sources for more info.
Again, I might be wrong, I looked in details to agri stats, but that was in 2002 for my masters .
hope it helps
Quality comment or not?
0 
0
Jean-philippe makes a very important point about the accuracy of data and the prevelance of armchair surveys. The difficulty in obtaining accurate quantitative data is somewhat underemphasised in many reports precisely because the conclusions to be drawn from the data take precedence over all but the most formulaic methodological considerations. Armchair surveying is an aspect of a series of cultural obstacles faced by the data collector. In many parts of Laos, and possibly also in Thailand, the chief is seen as the source of authoritative information in a village community where statements about amounts of anything are heavily influenced by preconceptions about what correct answers might be, by poor education, and by very different understandings of the purpose of counting production. It is a bit like asking people what their incomes are. If we have no way of observing and recording the money they recieve then we are dependent upon whatever they wish to disclose, and in most cases this happens to be much less than the reality. I’m sure attempts to amass data on rice yields are subject to similar cultural distortions.
Quality comment or not?
0 
0
As I live and work with small-scale rice farmers in northeastern Thailand (Esan), the topic of yield is commonly discussed. Though I’m young and somewhat inexperienced, I hope my observations can make a contribution to the discussion here.
Esan is known for it’s poor soil quality to begin with, but I believe farmers’ dependence on nitrogen fertilizers (as opposed to organic soil improvement) has been a major contributor to the very gradual increase seen on the chart. Farmers have increased their chemical fertilizer use exponentially since it was first introduced during the Green Revolution. While fertilizer use seems to have reached a plateau, with exhausted soils and inefficient water management in paddies (as well as many seasons with real drought conditions, like in 2007 and 2008), yields are bound to remain low or even decline.
Further, as Jean-Philippe mentioned, many farmers simply don’t put the time and energy into their rice crop in order to reap high yields. Many decide to invest in their children’s education or buy goods for use around the household (some more beneficial than others). Fertilizers are also very expensive to begin with. Many farmers are only in the village during the planting and harvesting seasons, working in other labor-intensive jobs during the rest of the year. Rice prices remain relatively low, so there’s little incentive to focus on rice when crops like cassava, sugarcane and rubber are frequently promoted and advertised as more lucrative.
That said, there is a small movement of organic farmers working to improve soil, preserve and develop indigenous seed varieties and support local food systems – the Alternative Agriculture Network – Esan. We have been using System of Rice Intensification (SRI) techniques with huge success to preserve and expand indigenous rice varieites. Many of these varieties yield as high or higher than Jasmine, and importantly, are suited to the ecological diversity in a given farmers’ paddies. Some of our members in Ubon Ratchatani province have also been able to achieve 1.5 to 2 ton per rai yields, by planting using SRI techniques and growing a “floating rice” variety, well-suited to flooding in the area.
For more information about our network, please visit aanesan.wordpress.com
Quality comment or not?
0 
0
I’d like to add on the adoption of modern rice varieties that Thailand’s clearly an outlier. MV were planted at the end of the 1990s over only 18% of rice area. Comparative figures are more than 90% in Vietnam and Philippines and more than 50% in all other SEA countries, except Malaysia and Cambodia for which I couldn’t find recent data. In Laos, the rate of adoption went from less than 1% to more than 50% in about a decade.
On Thailand and its choice of taste (higher quality and price) over quantity, see Rerkasem B. (2007) Having Your Rice and Eating It too: A View of Thailand’s Green Revolution. ScienceAsia, 33(s1): 75-80.
For a major source of stats on MV, see http://beta.irri.org/solutions/index.php?option=com_content&task=view&id=250 . Data for Laos can be easily found on google scholar.
Quality comment or not?
0 
0