On farm productivity

Posted in development, economics, macro by jrahman on January 24, 2013

Many years ago, when Dr QIB Chowdhury lived in Bangabhaban and Tarique Rahman was not widely reviled, Shafiq Rehman hosted a TV programme (I forget in which channel) where politicians from across the aisle participated.  One item involved asking BNP leaders who they thought was a successful minister in the previous Awami League government.  Matia Chowdhury — the agriculture minister in that government — was chosen pretty unanimously.  Ms Chowhdury is agriculture minister in the current government too.  Interestingly, according to the latest Prothom Alo – ORG Quest poll, she is not viewed as a success any more.  Over two-fifths surveyed believe under her watch, farmers are not doing well.  Over half consider her to be a failure as a minister.

I was intrigued by the poll results.  Is there any way to check her performance in the data?  Well, as with garments sector, one good way to check how a sector is doing is to look at its productivity.  When productivity rises, profits and wages can both rise.  The standard measure of productivity in agriculture sector is farm yield (output per unit of land).  The thin green and gold lines in this chart shows annual growth in yield (kg per hectare) in rice and all crops respectively.  As you can see, these annual growth series are very volatile. This is because there is a lot of variation in agricultural output on an annual basis. To make any inference, one needs to smooth the data. There are more fancy techniques, but for a blog post annual average over the previous five years should suffice — these are the thicker lines (green for rice, gold for all crops).


Suppose Matia Chowdhury took some concrete steps to give lasting benefit the farm sector in, say, 1996-97.  How would that show up in the data? If as a result of her actions, productivity growth accelerated for a while, then we would expect to see the thick lines step up towards the end of the 1990s, and stay elevated into the 2000s.  And in fact, that’s exactly what we see.  Farm yields grew pretty steadily in the 1970s and 1980s.  But by the mid-1990s, farm yields basically stopped growing.  And then yield growth rebounded strongly in the late 1990s before moderating to historical pace in more recent years.

The data is consistent with the following story:

  • whatever ‘green revolution’ policies were taken by the Mujib, Zia and Ershad governments, by the time Matia Chowdhury became the minister, they had run their course;
  • whatever policies she enacted revived farm yield;
  • her successors maintained these policies, but absent new steps, yield growth has eased.

Of course, there is a natural limit to farm productivity.  Bangladesh is a very densely populated country.  Our farms are small.  With so many rivers and canals, perhaps the delta is not the most conducive place for large scale mechanised farming.

Are we hitting the limit?

One way to answer that would be to check our farm productivity compares internationally.  Our rice-producing, densely populated neighbours in South East Asia are natural comparators.  And as this chart of crop yield shows, our farmers are more productive than Thai and Filipino ones, but not as much as the Vietnamese and Indonesian ones.


This wasn’t always the case.  In the 1960s, Bangladeshi farms were similarly productive as their Indonesian peers.  But Indonesians lept ahead in the 1970s and 1980s, and have maintained their lead since.  Vietnamese farms have steadily become more productive since the 1980s, and became more productive than Indonesians by the end of the 1990s.

Can Bangladeshi farms match Vietnamese or Indonesian levels of productivity?

I have no idea.  I am not familiar with the research in this area.  But digging a bit deeper, I found something interesting.

This chart shows fertiliser consumption per hectare of arable land during the 2000s.  Vietnamese farmers use significantly more fertiliser than others in the region.  Now, I know nothing about farming, so happy to be educated.  But my guess would be that fertiliser use, at least in some context, is positively correlated with farm yield.


Notice the rise in fertiliser use in Bangladesh at the last data point?  That figure is for 2009, after Ms Chowdhury’s return to the job.  If fertiliser use is a good predictor of higher productivity, and if the 2009 was beginning of a new trend and not a once off, then farmers may well have better times ahead.

Data source: World Bank World Development Indicator for crop yield and fertiliser, CEIC Asia for rice yield.



8 Responses

Subscribe to comments with RSS.

  1. Diganta said, on January 25, 2013 at 12:20 am

    Could you provide me the data source?

    • jrahman said, on January 25, 2013 at 3:31 am

      Updated into the post.

  2. Diganta said, on January 25, 2013 at 4:32 am

    One unanswered question from your graph is that why Thailand is still a rice exporter (was largest till last year) despite the fact that their per hectare productivity is not up by much and they are not willing to “invest” more fertilizer into their agriculture. I remember my father once explained that Thailand, which thinks beyond sustaining the population like most of South Asia, is more focused on quality of rice rather than the quantity.

    • jrahman said, on January 27, 2013 at 8:47 am

      I don’t know why Thailand is a rice exporter, but I suspect it has a comparative advantage in rice production, even if others have higher productivity, and thus absolute advantage.

  3. shafiq said, on January 27, 2013 at 8:03 am

    Very informative. The first graph doesn’t really cover BNP regimes 1991-96 and 2002-2006 with glory. BNP should ponder upon why Awami Leagues rural support remain so resilient.

    • jrahman said, on January 27, 2013 at 8:50 am

      The difference between the two 1990s government is indeed quite starke.

  4. Why did poverty fall? | Mukti said, on March 10, 2013 at 9:20 am

    […] point) can be attributed to rising returns from farm – underscoring the importance of farm productivity.  Rising returns from non-farm assets contributed to another 3 percentage points.  Poverty also […]

  5. […] green revolution is a relatively straight forward story, as is its impact, and remittance we have discussed above. […]

Comments are closed.

%d bloggers like this: