268 – Conservation agriculture in developing countries

Is “Conservation Agriculture” really a win-win for farmers and the environment? It reduces the losses of soil and nutrients from agricultural land, but are farmers who adopt it better off or worse off? In asking this, my focus is on developing countries, where soil erosion remains a serious problem.

Soil erosion reduces farmers’ livelihoods, but also causes off-farm damage, particularly to rivers, lakes and dams, and “Conservation Agriculture” has been promoted as a solution. A particular Conservation Agriculture (CA) package of three farming practices has been widely promoted in developing countries as a win-win option for farmers and the environment. It consists of zero tillage, retention of crop residues for soil cover (mulching), and rotation of cereal crops with legumes, which fix nitrogen and so increase soil fertility.

Something close to this package has been widely adopted in North America, South America and Australia, but its adoption by poor farmers in Africa and South Asia has generally been disappointing, despite years of active promotion by international organisations.

Does this mean that poor farmers can’t recognise a good thing when they see it, or is the problem that benefits of Conservation Agriculture to these farmers are not sufficient to outweigh the costs? Given the massive differences between small subsistence farms in southern Africa and large commercial farms in North America, we should be open to the possibility that a practice could be very attractive to one group of farmers but not the other.

Photo: Marc Corbeels

Photo: Marc Corbeels

In 2012 I was invited to review the evidence on the economics of Conservation Agriculture in Africa and South Asia, and to present my findings as part of a workshop on Conservation Agriculture held in Lincoln Nebraska that year.

Part of the motivation for the workshop was to try to address issues of controversy about Conservation Agriculture in developing countries. Some people had been critical of what they saw as promotion of CA as a silver bullet solution without sufficient consideration of circumstances where it did and didn’t work for farmers. Others hit back at these criticisms and defended CA as the best available option.

The workshop marshaled the evidence on various facets of CA. The results have recently been published in a special issue of the journal Agriculture, Ecosystems and Environment. All the papers can be accessed for free here.

Looking at past literature on the economics of CA for poor farmers, my co-authors and I found that there wasn’t all that much literature, but what there was tended to indicate that CA (or its components) should be economically attractive to small farmers. However, looking critically, little of this evidence was based on sophisticated analysis. Many of the economic models used seemed too simplistic to be reliable. 

So we set out to build a model of our own and apply it to a case study in northern Zimbabwe. Marc Corbeels from CIRAD joined the team, providing his data and extensive experience for the case study.

A key finding is that the farm-level economics of CA are highly case-specific. In some situations it is competitive with traditional agriculture, but in others it falls far behind. Organisations promoting CA need to be quite careful and discriminating about where it is actually a viable strategy if they are not to waste their own and farmers’ resources.

We didn’t find any scenarios where CA was substantially better than traditional agriculture. Where it was better, it was only slightly better, but where it was worse, it was sometimes dramatically worse. This would be important to farmers who are worried about risk.

CA tends to be more attractive for larger, wealthier farmers and to be much less profitable for the smallest, poorest farmers. No wonder many of them have been reluctant to adopt it.

There are several elements that can reduce the benefits of CA for small farmers. One is that zero tillage can increase the number of weeds in the crop. This either reduces crop yield or requires additional effort for weed control. A second factor is that some of the benefits take about a decade to kick in (increased yields from zero tillage combined with mulching). The poorest farmers might not be able to afford to look that far ahead when they make their farming decisions. Thirdly, crop residues may not be available for mulching. They may be harvested to feed to livestock, or in some cases farmers cannot keep other people’s cattle off their crop residues even if they want to. (The community doesn’t allow it, and there are no fences anyway.) Fourthly, legume crops may or may not be profitable enough to be worth including in the rotation. It depends on their yields and sale prices.

Some extension programs have emphasised the importance of farmers adopting all elements of the CA package. This is quite naive, as full adoption of farming packages rarely happens anywhere. Farmers almost always pick and choose the elements that they think will work for them, and leave the rest. We found that the economics tend to favour this approach in the case of CA – partial adoption tended to be more attractive than full adoption.

So, overall, this is one of those cases where the farm-level economics pose a barrier to changes that people would like to see, at least in some cases. There are good and bad ways to respond to this information. The bad ways include ignoring the information and continuing to promote CA in an untargeted way, or thinking that the adverse economics could be overcome by better or more intensive efforts to promote CA. The good ways include using the information to target CA promotion to situations where it is likely to be adoptable, and redoubling efforts to develop more appropriate soil conservation practices for situations where it is not. In some cases it might be feasible to subsidise CA to increase its uptake, but this won’t overcome adverse economics unless the subsidies can be maintained indefinitely. If they are to be maintained indefinitely, we would want to have confidence that the system will result in large enough reductions in the off-farm costs of soil erosion. 

Further reading

Pannell, D.J., Llewellyn, R.S. and Corbeels, M. (2014). The farm-level economics of conservation agriculture for resource-poor farmers, Agriculture, Ecosystems and Environment 187(1), 52-64. Journal web site (access to the paper is free) ◊ On-line video presentation ◊ IDEAS page for this paper


  • stirfry
    27 May, 2014 - 8:55 am | link

    Conservation agriculture is inherently chemical dependant. Selling CA is really selling chemical. GM will be next on the list. They will have to buy their chemical and their seed. Good luck putting a spray rig behind that poor donkey.

    • 27 May, 2014 - 11:46 am | link

      Yes, herbicides are important to the success of conservation agriculture. One of the key reasons why CA works better for larger farmers is that they can afford to buy herbicides and spray equipment. It doesn’t follow that this is a bad thing. Like most things, there are benefits and costs of CA, and the thing is to evaluate whether the benefits outweigh the costs, including any concerns about the use of herbicides. There certainly are health risks for farmers if they don’t use suitable protective gear, which they often don’t in developing countries. There could be potentially be environmental impacts as well in some cases. Both these negatives can be addressed, at least to some degree. And there is no doubt that there can be benefits of CA for farmers and the environment in the right circumstances.

      • mr mehrvar
        21 July, 2018 - 11:17 am | link

        The worst enemy is to defend badly! I mean if you define CA as a collection of badly manners, you can not call it CA because you are designing it incompletely and firstly you are cheating yourself and secondly the next generations.

  • Marta Monjardino
    27 May, 2014 - 10:51 am | link

    David, a very interesting and useful analysis, which seems to bring much sense into the whole debate. However, I wonder how the analysis would benefit from an added component: fertilizer. Since lack of organic resources can be a major constraint to adoption of conservation agriculture, presumably appropriate use of inorganic fertiliser (if available) would greatly increase crop productivity, hence crop yield and residues. Interesting to see how this would trade off against potential higher weed burden, labour costs, etc.

    • 27 May, 2014 - 11:55 am | link

      Thanks Marta. We did include fertilizer in there for scenarios where we judged the farmer could afford it. But we didn’t look at fertilizer as a substitute for mulching. Some of the nitrogen benefits from mulching legumes could indeed be supplied by fertilizer if it was available and affordable. But mulch provides other things (e.g. protection against soil erosion) that fertilizer wouldn’t give you.

  • Les
    29 May, 2014 - 7:55 am | link

    Do I sense the assumption that economics inevitably drives human (farming) behaviour?

    Recall reading an interesting book recently (“Changing Land Management – Adoption of New Practices by Rural Landholders”) that made a coherent case for consideration of a much broader suite of drivers of change on the farm…

    • 29 May, 2014 - 9:02 am | link

      Thanks Les

      That’s a pretty good book! http://www.publish.csiro.au/nid/21/pid/6483.htm

      Thanks for giving me the chance to clarify this issue.

      Obviously, I never assume that economics is the only factor driving farmer behaviour – I’ve written extensively about the many other factors, as you point out. But it’s certainly one of the key drivers and is well-worth understanding.

      And when I say economics, I mean it in a broad sense. Included in the economic model we developed for this study are considerations of: the high uncertainty associated with new practices during the learning phase; the relative riskiness of new and old practices; the attitudes of farmers to risk and uncertainty; use of farm labour; the ability of the farmers to access markets and purchase required inputs; the time lags between adopting the new package and generating the benefits; the farmer’s attitudes to time lags and their planning horizons; farmers’ skills and capacity; and cultural norms that constrain some farmers from controlling what happens to their own crop residues. Some of these might be seen as social issues, but they also play roles in the overall assessment of benefits and costs. I described most of the existing economics studies of CA as simplistic because they don’t represent these sorts of things.

      Then there are other factors that you wouldn’t build into an economic model but still might influence behaviour: social, cultural and attitudinal factors that might enhance or inhibit adoption. Farmers might have objectives that cannot be captured in an economic model (although many can, either directly or by proxy).

      So, an economic model, done well, can give good indications of likely adoption behaviour, but should be treated as indicative rather than definitive.

      An exception to that would be cases where the new practice being considered is highly inferior (from the farmer’s perspective) to their current practice. If a good economic model indicates that farmers would be made significantly worse off by adoption, then I’d be prepared to stick my neck out and say that adoption will not happen on any significant scale, even if there are other social factors that seem to favour adoption.

      In my experience, any error in (tentative) predictions of adoption based on an economic model is likely to be on the up side. Actual adoption is usually lower than economic models imply it should be. This just strengthens the conclusion of this study.

  • Brett Robinson
    30 May, 2014 - 5:03 pm | link

    Thanks for an excellent article. It is refreshing to see extra analysis of the sweeping generalisations that are heard about tillage and fertiliser. Here in Qld, farm size is an obvious determinant of the use or otherwise of zero till equipment and methods. This is despite the substantial benefits of RT and ZT for large farms.

    I wonder if it isn’t coincidental that legumes in cropping give us headaches in the Australian subtropics as well as in tropical Africa and SE Asia? In Qld, grain legume yields are low, they don’t fix much N over and above the export in grain, and critically – their residues don’t persist (just what you don’t want in a ZT system). Collectively, they are economically and environmentally high-risk.

    On the concept of recipe-based solutions to problems, the always-brilliant Greg McKeon once said to me: “Brett, it’s not just horses for courses. It’s ALWAYS horses for courses”.

    Thanks again for raising a whole lot of interesting questions.

    • 30 May, 2014 - 5:14 pm | link

      Thanks Brett. Yes, “It’s ALWAYS horses for course” is a pretty good rule of thumb. Very occasionally a new technology comes along that absolutely everybody adopts (e.g. GM cotton), but most often it works in some situations but not others.

  • Matt Giraudo
    26 June, 2014 - 9:59 pm | link

    Conservation farming, or perhaps more correctly, reduced tillage has through its adoption reduced some degradation issues whilst exacerbating others. Conservation agriculture in WA has generally led to reduced erosion and improved soil health. However in my experience, landholders in WA who still employ more traditional tillage techniques generally remain profitable and can maintain good soil health.

    Reduced tillage has also led to:
    – Increase reliance on chemicals for weed control
    – More rapid onset of herbicide resistance
    – More rapid soil acidification – through less soil disturbance
    – More rapid development of soil-water repellence – through reduced soil disturbance
    – Increase farm debt – through increase investment in machinery
    – Tendency toward continuous cropping resulting in generally less resilient farm enterprises.

    It’s worth noting that the key physical improvements to soil health that accrue from reduced tillage are in part soil type and landscape dependent. However, as you point out, the key issue is that reduced tillage systems also tend to require greater input costs and increased investment in and reliance on technology. CA is therefore typically more technically demanding and capital intensive than more traditional farming systems. Clearly in some cases this will be counter-productive to the existing farming enterprise.

    A proper comparative analysis would be interesting – as I’m not sure anyone has actually considered the overall benefits and costs of reduced tillage systems on the farming enterprise. It is simply assumed to have been effective, with the “costs” to the farming system not being properly accounted for.

  • Abena
    6 September, 2016 - 8:18 pm | link

    Very interesting piece.
    For African countries especially a lot of conservation Agriculture packages have been introduced and indeed it has taken a lot of time for farmers to adopt some or any of these practices due to the fact sometimes, due to the long -term benefits of these practices and lack of understanding in how these packages work.

    But currently farmers are beginning to adopt some of these farming practices to improve nutrient quality in the agricultural soil.

  • Abdirizak Mohamed Ali
    27 November, 2016 - 2:32 pm | link

    Thanks mr David pannell for your analysis of this topic coservation agriculturein developing countries.
    Indeed as a somali our farmers they do not understand what the conservation agricultureis , in our country doesnot exist what is called conservation agriculture
    Somali is acountry which has ahuge potential of agricultural / arable land,approximately somalia has 8.2 million Ha of agricultural land which is suitable for farming.
    One of the major problems which affecting to this resource include land degradation& soil erosion ,So in order to prevent land degradation&soil erosion we have to use modern farming technologies such as conservation agriculture.
    So our farmers need to understand this concept and then apply.
    Thanks more & more mr.David pannell.

  • ahmet hasim keskin
    11 December, 2016 - 5:40 pm | link

    Hi David

    Wheat price is very important for my country food sector. Because bread main part of our meals. So high level wheat price effects all community. İ wonder we can learn wheat price around the world. İt is will ensure ve real cost management for next times our agricultural policy.

  • Howard
    20 February, 2017 - 6:33 pm | link

    This is Howard Brown;

    Mr. Pannell: your lectures and writings are not just interesting but extremely informative. Obviously, you have a wide scope of knowledge in your field and what is apparent as I continue my study is that you thoroughly enjoy what you do. That’s very commendable.

    The lively discussion on Conservation Agriculture ignite my interest to make one observation and to ask a question on the same.

    The observation is: before these organisations promote these innovative (please don’t get me wrong, I am for innovation) and ‘apparently’ attractive methods of doing things; I would have thought that if they had done more researches and more case studies (like what you did), the promotion would have brought about a more sustainable result. One wonders about their whole motives behind these promotions’.

    As you rightly said, CA tends to be more attractive for larger, wealthier farmers and to be much less profitable for the smallest, poorest farmers. No wonder many of them have been reluctant to adopt it. I agree with you.

    Due mainly because of affordability and if I may add, practicality. As it does not seem feasible or economically viable for the peasant farmers.
    For health and affordability reasons, how practical would it be if CA is implemented on a phase by phase bases and in a partial way? I mean, executing the package without using the “zero tillage” part of it. In this way the small farmers would not have to purchase herbicides.

    • 7 March, 2017 - 9:18 pm | link

      Thanks for your kind comments Howard. You’re right – I do enjoy it.
      You can certainly do the legume part of the strategy without zero till or mulching, but those latter two tend to go together.

  • Dinara
    22 February, 2017 - 12:29 am | link

    Very interesting article.

    my concern is that in most cases climate adaptation options in agriculture are based on zero tillage, crop rotation and mulching, and in addition are focused on small-holders. So there is no win-win really?

  • 27 September, 2017 - 3:21 am | link

    This post is very good and is really a good practice of conserving the agriculture and now-a-days its really a important aspect that needs to be taken in account so that our agriculture world is safe and every farmer also have some rights their voice on any agricultural act taken.
    Thank you for the post and I like it very much.

  • Susan Webster
    9 September, 2018 - 9:40 am | link

    Hi David
    The practice of open access would see grazing livestock adding N to the soil. Would that factor affect the yields? Or would it be considered negligible?

    • 9 September, 2018 - 11:14 pm | link

      Interesting point. In some circumstances I think it would be a partial benefit, although I doubt it’s enough to outweigh the dis-benefits (the loss of residues) in most cases. If it was really beneficial to have cattle defecating in the field , farmers would still be better off if they excluded others’ cattle and ran their own on the crop residues.

  • Ekeh Grace
    20 April, 2020 - 3:48 pm | link

    Thanks for this article, it’s nice. But it can be really difficult to convince poor farmers to adopt this particular technique (CA) or any other technique whose profit isn’t obvious, or perhaps will take a long time to get. And I also think zero grazing should be adopted more in areas that are prone to erosion.

    • 21 April, 2020 - 9:38 am | link

      I completely agree. The practices that get adopted quickly and comprehensively tend to be those that are clearly more profitable than what the farmer was doing before. But it’s still true that non-financial factors influence adoption to some degree in many cases.

  • Leave a Reply

    Your email address will not be published. Required fields are marked *

    Please solve this to show you\'re a real person *