222 – Technology versus labour in agriculture

A key component of the transition from agriculture-based economies to industrial and service-based economies is the substitution of new technologies for labour in agriculture. This reduces the demand for labour in agriculture, and allows labour to move to more productive jobs in manufacturing and services. Western industrialised countries went through this process in the 19th and 20th centuries, and many developing countries, such as those of south-east Asia, have been following in their footsteps in recent decades.

The process doesn’t all go smoothly, and not everyone who exits from agriculture is better off – just think about conditions in the factories of England during the industrial revolution. But in the long term it’s an essential element of improving the quality of life for the great majority of people.

Sometimes the process involves farmers adopting attractive new labour-saving technologies and shedding labour, which then has to leave agriculture to find work. In other cases, labour is pulled, rather than pushed, out of agriculture, and farmers have to adopt labour-saving technologies because there is a shortage of labour, reflected in rising labour costs.

One of my PhD students, Jessie Beltran, did a study that illustrates the latter process beautifully. In the Philippines, there has been increased demand for labour in non-agricultural sectors of the economy. The resulting increase in labour costs has affected rice farmers, who have traditionally relied on a highly labour-intensive form of agriculture, especially for planting and weeding.

Jessie built an economic model of the planting and weeding decisions of Philippine rice farmers and used it to explore the consequences of rising labour costs. The results were just what you would expect. If labour is cheap, the traditional labour-intensive transplanting approach to seeding is the most profitable option. As labour costs increase, direct seeding becomes more profitable because it requires much less labour.

In addition, if labour is cheap and weed numbers are not excessive, it is best to rely on manual weeding, with people pulling up every weed in the crop by hand. However at higher labour costs, application of chemical herbicides becomes the more profitable option.

Both of these changes are occurring in reality in the Philippines. An inevitable side effect in the medium term will be the evolution of herbicide-resistant weeds. At this stage, farmers there have little or no awareness of the risk of resistance, but the first cases of it have already been discovered. It will be only a matter of time before most farmers who use herbicides have problems with resistance, and have to look again at more expensive, less convenient weed control strategies. But they’ll have to do that in the context of high labour costs, so they are likely to employ different weeding technologies, rather than going back to hand weeding.

Further reading

Beltran, J.C., Pannell, D.J., Doole, G.J. and White, B. (2012). A bioeconomic model for analysis of integrated weed management strategies for annual barnyardgrass (Echinochloa crus-galli complex) in Philippine rice farming systems, Agricultural Systems 112(1), 1–10. Journal web site ♦ IDEAS page for this paper

Beltran, J.C., Pannell, D.J. and Doole, G. (2012). Economic implications of herbicide resistance and high labour costs for management of annual barnyardgrass (Echinochloa crus-galli) in Philippine rice farming systems, Crop Protection 31: 31-39. Journal web siteIDEAS page for this paper


  • Thilak Mallawaarachchi
    27 August, 2012 - 3:35 am | link

    Hi Dave,

    An interesting and topical issue. Withdrawal of labour from agriculture to industry is happening in many countries around the world and its impact on future land use chnage will be an interesting topic for study. Although western economies have benefited from high capital availability allowing mechanisation to help capital substitution for many countries in East and South Asia and Africa in particular high opportunity cots of capital may leave land abaondoned as profitability drops and local costs of production rises above border price.

    In my own experience, in Sri Lanka, direct seeding was always the preferred option and rice planting never took off despite significant push by agronomists. That however led to plant breeding based solutions as high tillering varieties allowed similar gains in productivity. Overall though the use of agrochemicals have increased significantly.

    In our own backyard here in Oz, labour issues are continually coming up as the type of labour that we require now to manage our high-tech agriculture is more multi-tasking in nature than the traditional outdoor type. Lot more issues to study!



  • Rob Bramley
    27 August, 2012 - 7:32 am | link

    I find interesting the notion that technological advance in agriculture “reduces the demand for labour in agriculture, and allows labour to move to more productive jobs in manufacturing and services.”
    Obviously, technological advance reduces labour demand in agriculture, but, industrial revolution aside, does it necessarily allow labour to move to “more productive jobs” – by what measure are these more productive ? One would have thought that producing food was more productive than making toys in a chinese factory for instance.
    Does this suggest that agriculture is missing an opportunity by failing to shift its labour from the farm to the value-added part of the agricultural value chain ? What would be required to encourage the labour shift to be targeted in this way ?

    • 27 August, 2012 - 8:12 am | link

      The measure of productivity is the amount of wealth that can be generated in each of the jobs, which reflects what people in the community (or in export markets) are prepared to pay for the things that are produced, and the technologies used to produce them. Producing food is not more productive than making toys, at least at the margin, if the amount of wealth that a worker can generate is greater in toys than in agriculture. This is not a problem. It’s actually a very positive feature of our economic system – resources (including labour) move to production of those goods and services that have the highest value added. It’s a big part of the explanation for why capitalist economies have been so successful. If food production becomes more valuable for whatever reason, resources move back in that direction.

      It doesn’t follow that agriculture is missing an opportunity in the agricultural value chain. If there was an opportunity there, labour would move there. It wouldn’t necessarily be the same labour as had left the farm sector.

  • Bolirvia
    27 August, 2012 - 7:49 am | link

    It’s still happening in Australia today and will continue. The high cost of labour is a continuing complaint from sections of the horticulture and dairy industries. The other evidence is the continuing debate about the “loss of young people from agriculture”.

    • 27 August, 2012 - 8:00 am | link

      Yes, that’s absolutely right. One of the big challenges facing agriculture at the moment, especially in Western Australia, is competition for labour with the mining sector. Farmers generally can’t compete with the sorts of salaries being offered by mining companies.

  • Peter Ampt
    27 August, 2012 - 9:48 am | link

    The apparent ‘inevitability’ of this shift of labour from rural to urban troubles me. Economists often argue that it is an essential part of ‘development’ and wealth generation that improves quality of life. I find this a symptom of the type of thinking that loses sight of broader social, cultural and ecological issues. Unless I am missing something, It appears linear and fatalistic.

    In China they are re-thinking some of the policies that encouraged rural urban migration because of the other problems it causes. Africa is ‘poised’ to potentially become the global food bowl due to it’s ‘underdevelopment ‘ of agriculture. If development follows the labour replaced by technology path it will dispossess rural smallholders, cause greater problems in urban areas that are already often dysfunctional, when improving the productivity and profitability of small holders and villagers may be far better suited to the benefit of most.

    • 27 August, 2012 - 10:54 am | link

      The ‘inevitability’ I mention in the piece is about the onset of herbicide resistance, not rural-urban shifts. Nevertheless, I would say that, if economic development does proceed in the industrial and service sectors of an economy, a sizable shift from rural to urban is inevitable. It largely occurs because people leave rural areas voluntarily and go hunting for a better life. In any such major shift, there can be significant social problems, as I acknowledged in the article. Individuals who move, or their children, can potentially be really badly affected – living in terrible conditions on terrible and uncertain incomes. If the receiving areas are dysfunctional, of course that makes things worse. It’s intriguing that people often choose to put up with those conditions rather than move back to rural areas.

      Is this linear? What does that mean? I presume it means there is cause and effect operating. Yes, there certainly is; people respond to differences in wages.

      Is it fatalistic? Fatalism: “a doctrine that events are fixed in advance so that human beings are powerless to change them”. Of course it is possible for individuals to choose not to make the move. So in that sense it’s not fatalistic. But at the aggregate level, I think it is part and parcel of the development process. If a country experiences economic development (as we would want every poor country to do), a significant level of rural-urban migration will surely follow. I don’t think that’s fatalism, just realism. The migration might be moderated by government policies to some degree, but trying to stop such a strong force would be difficult. And they won’t want to moderate it too much because the labour is needed in the industrial and service sectors. (If it wasn’t, then wage rates wouldn’t have increased.)

      If it’s possible to improve the productivity and profitability of small holders and villagers, that would be a great thing, and would moderate (but not eliminate) out-migration. But realistically it’s not a substitute for broad-based economic development.

  • Peter Ampt
    27 August, 2012 - 9:53 am | link

    I forgot to there are also Australian farmers and graziers who, for a whole range of reasons that might appear economically irrational, have stepped off the linear technology treadmill and are developing strategies that are replacing purchased inputs (like fuel, fertilizer and pesticides which will inevitably continue to rise in price) with one form of labour (moving stock around and pasture cropping) while attempting to regenerate natural systems.

    • 27 August, 2012 - 10:59 am | link

      Economists would never describe those decisions as economically irrational. They are decisions by individuals to pursue particular personal objectives, which we are all free to do. But the dominant trend in agriculture in the developed world has been (and still is) for labour to leave, to be replaced by a variety of labour-saving technologies.

  • Michael Harris
    4 September, 2012 - 11:29 am | link

    Interesting post, Dave, and an interesting (albeit somewhat confused/ing) follow-up discussion. In my new role I am thinking a lot about “adaptation”, what it means analytically, and what the policy implications might be. (First commenter Thilak is doing much the same!)

    Some of the confusion in the discussion relates to scale issues, and some relates to normative judgements, and both of these confusions seem to be interacting with each other.

    I agree, as would most economists, that at the macro scale, the substitution of labour for technology could be described as “inevitable” — the process being virtually inexorable. As you note, this is a combination of “jumping” (e.g. moving to the city, or working for the mines) and “being pushed” (labour being crowded out by new innovations). Poor King Canute is misremembered for attempting to hold back the tide, whereas in fact he was demonstrating the limits of his powers. Like Canute, our powers to stem the tide of labour out of farming at the macro scale are very limited indeed.

    At the “micro scale” we can raise serious questions about the costs and impacts of particular rates of change (resource reallocation, in effect), and whether there are some serious concerns to be addressed (that can effectively be addressed, which is another issue). Since at least the 1970s work of Harris (no relation) & Todaro, we’ve had some formal understanding of the costs and consequences of people leaving farming for the city without jobs to go to — they buy the equivalent of a lottery ticket (the chance of a job) and the end result is a large pool of urban unemployed.

    In China, the magnitude of the shifts involved are pretty close to unprecedented as far as I can tell, with consequent impacts on both urban congestion and urban sprawl. I’m not surprised they’re re-thinking that. But that’s not to say they will, or could, reverse or even halt the process at the macro scale.

    Again, “improving the productivity and profitability of small holders and villagers” is a (relatively) micro-scale intervention, and one that should be on the policy agenda anyway, irrespective of rates of rural-urban migration. It’s not an alternative to migration, and it wouldn’t be sufficient to halt it.

    That’s some commentary on scale-confusion. There’s also confusion based on implicit normative judgements, in areas where economists like to be explicit.

    One thing that frustrates some agricultural scientists (and others) is that agricultural economists don’t “privilege” agriculture in the way many others do. Agriculture is a sector that produces food. If we get better at doing so, and can consequently release resources to other sectors to produce other things, great! If we find ourselves unable to produce as much food, but we can source it from elsewhere, that’s OK too.

    Of course, we can complicate that simple story with externalities, with resource (mis)management, with risks, with lifestyle preferences, and so on, and end up with a set of interventions designed to deal with those problems. But none of that requires us to privilege agriculture and assert that the sector requires special treatment.

    The notion that agriculture is “missing an opportunity” for keeping labour somewhere in the agriculture value chain, or similar concerns about resources moving from agriculture (as other resources move in, of course), tend to be products of unspoken “oughts” and “shoulds”, and associated opposites, the oughtn’ts and shouldn’ts. Agriculture “shouldn’t” be losing workers. We “ought” to try and keep people on the land and out of cities, and we “ought” to try to make low-tech farming more profitable.

    I would find these arguments and debates easier to navigate, and to contribute to, if people were better attuned and habituated to clearly articulating their oughts and their shoulds, and the basis on which they arrived at them.

    Dave, I think you’ve already handled “linearity” and “fatalism” so I’ll leave those alone!

    • 4 September, 2012 - 11:55 am | link

      Hi Michael. Thanks for that clear thinking and, for me at least, clear communication. I agree with all of that. Your comments should be very helpful to some of the earlier commenters, I hope.

  • Prem Bhandari
    20 May, 2015 - 3:38 am | link

    This is an interesting issue, Dave. I have been working on this area of scholarship. If you are interested, please see the article –

    Rural Agricultural Change and Fertility Transition in Nepal*

    Using longitudinal panel data from the Western Chitwan Valley of Nepal, this study investigates the impact of the use of modern farm technologies on fertility transition—specifically, the number of births in a farm household. Previous explanations for the slow pace of fertility transition in rural agricultural settings often argued that the demand for farm labor is the primary driver of high fertility. If this argument holds true, the use of modern farm technologies that are designed to carry out labor-intensive farm activities ought to substitute for farm labor and discourage births in farm families. However, little empirical evidence is available on the potential influence of the use of modern farm technologies on the fertility transition. To fill this gap, the panel data examined in this study provides an unusual opportunity to test this long standing, but unexplored, argument. The results demonstrate that the use of modern farm technologies, particularly the use of a tractor and other modern farm implements, reduce subsequent births in farm households. This offers important insight for understanding the fertility transition in Nepal, a setting that is experiencing high population growth and rapidly changing farming practices.

  • Leave a Reply to David Pannell Cancel reply

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

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