HISTORIC PERFORMANCE OF AGRICULTURE: Despite a precipitous fall in the relative contribution to the GDP of Bangladesh, agriculture has performed remarkably well so far. There was a time when agriculture, broadly defined to include crops and horticulture; livestock and poultry; fisheries and forestry, accounted for the bulk of the economy in all respects, viz., GDP, employment, food security, foreign exchange earnings, and supply of industrial raw materials. In the early nineteen seventies the share of agriculture in the total GDP was more or less 60 percent and went down to 48percent by the early nineteen eighties. By the turn of the century, it had been halved to 24 percent or so. In recent years it has fallen further to slightly above 16 percent only, with the crop sub-sector contributing about 12-13 points of that. 
Employment-wise however, the fall has been much more modest and agriculture still accounts for more or less one half of the total employed labour force. Also, despite such a fall in relative importance in the GDP, domestic production of food, particularly of rice has increased almost three times from nearly 12 mn metric tons (of husked paddy) in the late 1970s to 34-35 mn mt by now. And this has happened despite the overall fall in arable land under cultivation. At least in terms of availability of rice from domestic production, Bangladeshi farmers have managed to by and large feed the population which has more than doubled from the 75 million during liberation to above 160 million at present.
The record of growth appears to be not insubstantial. Based on official estimates of value added by the sector and sub-sectors, agriculture as a whole had been rising at an annual rate of just above 4 percent over 2005-06 to 2013-14. Among the sub-sectors fisheries had grown at the highest rate of 5.8 precnt closely followed by forestry at 5.3 percent. The contribution of livestock to the GDP had grown at the rate of 3.7 precnt while that of crops grew at the slowest rate of only 3.6 percent.  These apparently are not low growth rates, but may not be sufficient compared to the potential demand for their output. On the other hand these rates appear to be a little higher than what had been achieved earlier over a preceding period, say 1997-98 to 2007-08. For example agriculture grew at a rate of 3.4 percent and crops at 3.1 percent during that period. 
Despite such achievements, agriculture operates almost upon a razor's edge on several counts. But before that, one must understand why agriculture remains so important despite the low contribution to GDP at present and its basic characteristics.

ROLE OF AGRICULTURE IN DEVELOPMENT PROCESS
Historically in almost all major economies, agriculture plays or had played several key roles in the process of economic development. Most importantly, it provides food and employs the people - thus providing livelihood and income to most. It often supplies key raw materials for industrial processing, and is also a ready market for industrially processed goods such as necessary agricultural inputs (chemical, fertiliser etc.) and as well as domestically produced consumer goods. It earns foreign exchange through export of agricultural or agro-processed commodities, and supplies investible resources through various agricultural taxations as well as manipulation of terms of trade between agricultural and industrial sectors (i.e. relative cheapening of agricultural commodities as opposed to industrial goods) through pricing, taxation, and subsidy policies. There may yet be one other contribution of agriculture during the initial industrialisation process which is that the ready availability of the basic staples becomes the main wage good without which there would be inflationary pressure and the industrial wage would be higher and the domestic industrial production may not be competitive with others. Thus, growth in production of food crops indirectly helps industrialisation at least in the initial stages. In Bangladesh, agriculture has played all these roles in various combinations at different times particularly since the British period. Among all such roles of agriculture, that of ensuring food security will probably always remain paramount and essential for physical survival. That may appear to be now under challenge. To understand this, one needs to first take a look at the characteristics of Bangladesh agriculture.

Bangladesh agriculture may be characterised by 
     Major influence of climatic and physical environmental factors including natural hazards such as floods, drought, cyclonic storms and the like;
    Preponderance of landless, marginal and small farmers operating no more than 2.5 acres or just about a hectare particularly in crop cultivation but increasing emergence of commercialised agricultural farming in case of livestock (dairy and poultry farming), fisheries and forestry (nurseries);
    Pre-eminence of crop cultivation, particularly heavy dependence on rice with increasing yet limited diversification to non-rice crops as well as non-crop agriculture such as livestock and poultry and fisheries;
    Increasing dependence on technology which is becoming more resource (such as water) and energy intensive both directly (electricity and diesel for pumping water, power tillers and husking machines and in rice mills) and indirectly (e.g., use of urea produced with natural gas as basic raw material); 
    Instability in growth of agriculture in general, more particularly in crop production and rice output with seasonal differentiation.
These characteristics are not new and over time have been sought to be counterbalanced through development and extension of new technologies, spread of irrigation and specific pricing and subsidy policies. 
I would like to focus particularly on instability, the last of the mentioned characteristics. The rates of growth as described earlier will not make it clear. For that one needs to look at the year to year fluctuations in the rate of growth for recent years.
Takin agriculture as a whole and only the crops and the fisheries as sub-sectors - crop cultivation has shown substantial fluctuation from year to year. GDP of crops has grown in some years only as less than 1 percent to as high as 7-8 percent in some years, while agriculture which broadly mimics crops in pattern has grown at rates between 2-3 to 6 percent. But of even more interest is the observation that while both crops and fisheries showed good promise to begin with, over the years there had been perceptible decline in their performance and subsequently in agriculture as a whole.
Except for a short-lived spurt over 2009-10 or thereabout, both crops and agriculture as a whole had been a story of continuous decline in year-to-year rate of growth. Fisheries also experienced sharp falls from earlier years but later held on. These fluctuations are of particular interest because over the years since 2009 or so, there had been no major natural hazards such as big floods, cyclonic storms or major wide-ranging droughts. Furthermore there had been some good policies such as absolute and relative decline in prices of non-urea fertiliser and relative rise in urea prices very possibly leading to a better balanced use of fertiliser that is conducive to soil health.
The fluctuations in agriculture and its subs-sectors - particularly crops - have been mirrored somewhat by that in the output of main staple rice. Rice output, it is well known now, depends by and large on the dry-period irrigated boro and rain-fed aman. In 2012-13, boro accounted for roughly 55 and aman 38 percent of total output of nearly 34 million mt. But both exhibit substantial year to year fluctuations, aman more so in downswings and boro counterbalancing those by upswings. Aman's fluctuations are caused mainly by natural factors related to climate-related events such as drought, floods and cyclonic storms along the coast. Note further that over the years while boro expanded and replaced aush substantially, aman acreage more or less remained unchanged because the seasonal and other environmental niche it occupies is well-suited to its cultivation. In any case, despite it being subjected to various natural hazards, aman output is an important key to food security on two accounts: First because it accounts for a major share of total output and that if it fails, even after a good boro harvest there is practically no major supply of rice from domestic production for a long period of 8-9 months. The same applies for failure of boro which, however, is less vulnerable to natural hazards and more amenable to suitable and timely policy interventions. Indeed, historically in greater Bengal, famines have always been preceded by major failures in aman output.     

CLIMATE CHANGE - THE NEW CHALLENGE
Against such a backdrop, a new challenge, climate change, has arisen. While this is not the place to go into any detail of what constitutes climate change and what may be the impact on Bangladesh agriculture, let it be stated that this is going to increase instability of agriculture in general in all its sub-sectors and particularly for crops and rice production due to the extremes of weather and the related natural hazards that are expected over the next several decades. Indeed, agriculture as we know now may cease to exist in the extreme case-scenario.  We already know that even at present conditions aman is very vulnerable to climate and climate-related events and disruption of aman will thus jeopardise food availability. In the future, climate related natural hazards may become more frequent and more severe, adversely affecting aman output. On the other hand, uncertainties of rainfall will also mean uncertainties regarding irrigation during boro season. Furthermore, if the regional warming becomes quite severe, boro will be adversely affected also due to temperature extremes. If either aman or boro or both substantially fail in a year, that will bring huge food insecurity, economic, social and ultimately political instability. There are two questions that may arise here. Firstly ,what needs to be done to avert such a menacing possibility and what has already been done and secondly, are we prepared to face such uncertainties as a nation? 
To both the questions, the answer is that there is some progress in thinking and action but not enough. Instead of going into details, I would refer the interested reader to the relevant literature. But I would like to discuss the second issue briefly. There appears to be only a limited awareness of the urgency of the problem as manifested most glaringly in the policies of the country. This is possibly the main reason for the inadequate planning and investments for growth in agricultural and food production under climate variability and change. 
In general, the broad policies, particularly the recent ones are sensitive, sometimes in a holistic manner, to climate change impacts and issues (e.g. Bangladesh Climate Change Strategy and Action Plan; Bangladesh Country Investment Plan for Food Security and Sixth Five year Plan). Unfortunately, such awareness has not been reflected in sectoral policies except the food policies. The National Agricultural Policy, 2013 (pertaining to crops) mentions climate change only cursorily - not as an issue of imminent importance. Two major policies related to livestock and fisheries are silent on the issue. 
On the other hand, policies supportive of agriculture or affected by agriculture show limited attention or none at all to climate change, climate variability impacts or how inter-linkages will be impacted by climate change. Furthermore, while there is some awareness regarding food security under climate change, the concern remains confined by and large to staples such as rice. As in the case of non-crop agriculture, non-rice crops have also not been given attention regarding climate change impacts. 
One particular issue that has been more or less completely bypassed by the policy makers in Bangladesh is the issue of mitigation or reduction of greenhouse gas emission. The various energy policies have been completely silent on this. Only the BCCSAP has given a clear message couching it as low emission development and including it in the Sixth Plan. While the guidelines on research and small scale irrigation policy highlight the need for ensuring water use efficiency and thus indirectly lowering the demand for energy and fossil-fuels and ultimately lowering greenhouse gas emission, no direct mention of mitigation has been made. Nor hace the issues of water efficiency and related methane emission reduction been highlighted. The issue of nitrous oxide emission due to use of nitrogenous fertiliser has been completely lost sight of while at the same time lamenting heavier than necessary use of urea has been brought up in some of these policies.
One of the reasons behind the lack of or limited awareness may be the so called Rules of Business of the government which specify the mandates of each ministry. Under such Rules of Business, climate change is a matter under the purview of the Ministry of Environment and Forests. With greater global and national understanding of the climate change processes and their impact on the development process, the mandate should have been broadened to include other arms of the government. This has been lacking and may have resulted in the mismatch between the broader frameworks and policies, and the sectoral (i.e. ministry-specific) ones. 

.................................................................
The writer is a Professorial Fellow at the Bangladesh Institute of Development Studies. He may be reached at asaduzzaman.m@gmail.com