Scaling up

An important theme these days in agricultural development is "scaling up", meaning when an innovative technology or practice rapidly and relatively spontaneously spreads throughout an area, without outside support from donors or government.  The idea is that, if a technology is useful and viable, people will begin to use it of their own accord after seeing it work for their neighbors.  Conversely, if people only use a technology when they are encouraged to use it by outside support or payments, and then they stop using it as soon as that support is withdrawn, then the technology apparently wasn't worthwhile in the first place.

Reality is more complicated than these two extreme examples, of course.  Vaccines, for instance, are a low-cost intervention of proven efficacy, which means that they should have been adopted by the entire world population as soon as Jenner and Salk and company figured them out.  However, individuals can't just go to the store and buy a vaccine.  There needs to be in place a whole physical and social infrastructure of political entities that obtain and administer the vaccines, a cold chain to keep them from losing effectiveness, roads to get vaccines to people, etc.  In the case of vaccines, there has in fact been a gradual, seemingly sustainable scale-up, but lots of support from local and donor governments was necessary.  Likewise, the failure of people to universally adopt condoms as a prevention measure for AIDS, is due not to their high cost or lack of effectiveness, but rather to a whole slew of social and political factors that are bigger than the simple equation of efficacy and price.

So the challenge for development practitioners is to accurately understand which innovations are worth providing some artificial support to until they spread or sustain themselves of their own accord, and which innovations will never be useful or cost effective enough for widespread adoption.

In agriculture, there is much attention given to scaling up of improved seeds and modern inputs like fertilizers and pesticides.  We all know that peasant farmers could grow higher volumes of a given product if they used improved seeds or provided fertilizer to the plants, and the failure of these peasant farmers to do so causes much concern and hand-wringing as we try to figure out what needs to change for these technologies to scale up.  Given my background as a farming systems agronomist, I often see factors that clearly work against widespread adoption of certain technologies by farmers.  Often a given technology just isn't cost effective--the increased brute production of corn or sorghum or whatever isn't worth enough to offset the additional input costs or labor required to achieve those gains.  Sometimes the technology is cost effective, but there are other factors at play.  Maybe there's nowhere to store or sell the additional production, or maybe the new practices exhaust the soil and thus lower long-term yield potential, or maybe growing more of one crop crowds out another component of a polycrop, thus lowering overall unit-factor productivity.  There are lots of things that can go wrong in scaling up modern farming technologies.

I am fascinated by the difficulty of scaling some innovations, compared to the relatively quick spread of others.  I've read accounts indicating that a number of New World crops like maize, cassava, and sweet potato had spread rapidly in Africa and Asia by the 16th century.  There was no intervention from outside "development experts", except of course the probably inadvertent initial introduction of the crops by outsiders in coastal trading ports.  I guess the farmers in those countries knew a good thing when they saw it, and crops like those that yielded vastly more food per unit of land than existing crops would have been an easy sell for them.  This contrasts now with the difficulty we development people often have convincing farmers to even adopt a new variety of maize, just because its color is different than the existing common variety in an area.

Here is a link with an article and video describing one example of more or less spontaneous scaling up of an agricultural technology called natural forest regeneration.  It is a series of practices employed in farmlands of the Sahel, from Chad to Senegal (with its focal point in Niger) that allow the natural savannah tree habitat to regenerate, resulting in more fertile soils that can now support mixed cropping and grazing systems after having been barren before.  The scaling up of these practices happened relatively suddenly, and unperceived until recently by people outside of the region.  Now local governments and donors are trying to figure out a way to support and strengthen the dynamic that made these practices spread in the first place.  It is an excellent example of a natural, person-to-person scaling up of free knowledge.  Who knows if the deliberate efforts to promote these practices will be able to maintain the momentum of their natural spread, or somehow inhibit it?

Another example of free scale-up of very useful knowledge is training on integrated pest management in cotton in West Africa.  By holding a series of farmer field school sessions with farmers over the course of a year, the FAO was able to reduce toxic pesticide use in an entire region by over 90%, with no negative yield impacts.  This is to say that, after learning a slightly different way of managing the cotton they already know how to grow, farmers were able to drastically improve the economic, environmental, and social sustainability of their system.  Outside help was needed initially to carry out the training, but presumably these new practices, which require no special equipment or drastic changes to the farming system, will continue to spread from farmer to farmer basically for free. There are many other examples in which teaching people some simple precepts of integrated pest management can totally revolutionize farming systems.

This contrasts with the bought technologies that are often the focus of organized scaling up efforts.  Most donor money comes from countries and culture with a strong tradition of free markets and private enterprise, and perhaps because of this the most natural impulse is to find innovations that can be commercialized rapidly by private companies.  If there is a strong profit motive, the thinking goes, we can count on entrepreneurs to find an efficient way of rapidly disseminating the technology.  I understand this, and cellphones are a good example of a market-led scaling up (though with little support from donors).  Indeed, many development organizations wish they could find the equivalent of a Steve Jobs to take poor farmers from 0% to 80% adoption of improved crop varieties, just as the iPhone went from not existing to being the most-sold phone in the US in a matter of months or years.  We want rapid market penetration.

I understand this impulse, but the example of cassava in the 16th century, or natural forest regeneration in the Sahel, offer a very different model for how this scaling should or could happen through decentralized, unplanned sharing of free ideas between normal people.  Even for the rapid spread of cellphones in Africa in the past ten years, while there was obviously a market dynamic in the sale of cellphones and the provision of service, there were few proprietary technologies being used--Third World cellphones are often Chinese non-brands that either use no patented technology or just rip it off.

Another advantage of scaling up of nonproprietary technologies is that there is not as much risk of coercion or abuse.  When a seed company wants to push farmers to replace their freely-available seeds (that are constantly, dynamically being developed by local people and adapted to local conditions) with a proprietary seed that farmers must buy year after year, people understandably become suspicious.  Because seeds are alive, if you give up your traditional landraces or even open-source improved varieties for a few planting seasons, those varieties can disappear, and you will from then on be dependent on the privately-owned varieties in the market.  This isn't always a question of coercion--farmers in the US largely gave up their corn landraces in the mid-20th century because they made a relatively free choice to use the privately-sold seed.  It's still a tragedy to use valuable landrace diversity, but at least there's no abuse or tricks involved.

But let's get deeper into the question of seeds and scaling up.  The spread of transgenic corn and soybeans (and cotton, to a lesser extent) has been one of the most amazing, striking examples of rapid scale-up of a proprietary technology.  In this case, having useful genetic traits controlled by just one or two huge companies allowed the companies to scale up use of seeds with these traits, first in the US, and then in Argentina and Brazil--though the expansion in South America was often via "pirated" seed that was shared between growers with no royalties paid to the private companies that held theoretical rights over those traits.  This has been touted by the companies pushing it as the fastest scale-up of any technology in modern agricultural history.

Unfortunately for us who work in agricultural development, it is not easy to replicate this kind of rapid scale-up for other productivity-enhancing technologies.  This is so much the case that an author on the Grist site recently argued that transgenic technology is thus far a two-trick pony, because 99% of the world's transgenic crops possess either Bt insect resistance, glyphosate herbicide resistance, or both.  There are no other major transgenic technologies that have proven worthwhile on a global scale, thus far.  The author perhaps unfairly discounts the promise of genetic engineering of crops based on the narrowness of its achievements thus far.

There is though one corn seed innovation that some people are lauding as the next important, revolutionary breakthrough, at least for peasant farmers in Africa.  Farmers in many parts of the continent suffer from Striga or witchweed, a parasitic plant whose dormant seeds "wake up" when they sense corn or sorghum germinating nearby, then send out eel-like roots that suck onto and insert themselves into the growing corn roots, thus sapping all the corn's energy.  By the time you see the witchweed above ground, it's too late to do anything; it's already latched onto the corn, and you can't pull or spray the weed without killing the corn.  Even if you do kill it one year, there is a store of millions of seeds in the soil that will keep coming back. 

Enter BASF, a chemical company that had a competitor product to Monsanto's Roundup herbicide.  This product, Clearfield, operated on the same logic as Roundup:  it used a broad-spectrum herbicide called imazapyr that kills any plant it touches.  Normally it would kill corn, too, but the geneticists at BASF induced a mutation in corn that made it resistant to imazapyr.  They were able to say that it was not transgenic, because the "new" gene conferring imazapyr resistance was not inserted into corn from another species, but rather arose when they exposed corn seed to highly toxic mutagenic chemicals and high doses of ionizing radiation (a pretty scary process that is just as morally objectionable as genetic engineering, if you believe in the sanctity of ecological integrity, and certainly more dangerous).  Clearfield never captured a big share of the US market, I think because the herbicide imazapyr is more expensive and/or less effective than Roundup. 

A novel use of imazapyr-resistant corn was that you could coat the seeds with very low doses of the herbicide, and the corn plant would take up the herbicide in its system as it germinated, without being hurt by it.  This is potentially useful for African farmers struggling with Striga, because when the parasitic witchweed seeds germinate and send their tentacles into the corn roots, the parasite gets killed off by the herbicide.  It's a pretty ingenious innovation called StrigAway, and lots of development agencies are now trying to push the Strigaway system (imazapyr-resistant corn seeds coated in imazapyr herbicide) as the next big scale-up, to varying success. 

I don't know if Strigaway will take off.  If it does, I would be somewhat pleased because it really is a well-thought technology that can potentially address a huge problem using very low doses of herbicide.  Some people (fairly or not) worry that Strigaway may be part of a larger BASF strategy to gain worldwide market share for its Clearfield products, thus throwing into the mix concerns about that element of coercion that I'd referred to.  Of course, the real long-term solution to fix Striga would be to have longer, diversified rotations incorporating things like peanuts, cassava, beans, cotton, sesame, soy, okra, melons, and any other number of typical African crops into your mix.  These are all non-hosts that slowly wear down the germinating witchweed seeds looking for fresh blood.  Strigaway corn could be a part of such rotations, but shouldn't be the only or the dominant crop. 

So on the one hand I laud the innovation of the inventors of the Strigaway team, but I also recognize that it's in some ways an overly complicated and ultimately unimaginative alternative when compared with the common-sense good husbandry of maintaining diversified, healthy cropping systems.  Beyond this though, my earlier thoughts about proprietary technologies apply here.  Why did/would an herbicide-based Striga control strategy have to limit itself to the proprietary imazapyr herbicide?  There are plenty of other broadleaf herbicides like atrazine and 2,4-D that would potentially control witchweed without hurting the corn.  I've only found one study on the use of seed treatment with these other chemicals to control Striga.  It didn't give promising results, but surely if there were a number of great minds working on this question, they might find an answer, just as the profit-driven BASF researchers did.

Of course a strategy based on these freely-available technologies, much like the FAO's farmer field schools or the Sahel's forest regeneration management, would not bring great profits to any one company, since no one company has rights to those herbicides.  But what if the public sector had put just a little bit of money into research on such possibilities, in order to come up with a totally free-source innovation that could do what Strigaway is doing?  For me, such open-source, non-propietary innovation is the best chance we have of scaling up groundbreaking technologies.  This model of innovation won't make any one company a lot of money, and it won't even be possible for development agencies to track with indicators, and certainly not to take credit for.  But scaling up through organic innovation and sharing processes would create a huge impact in terms of incrementally improving quality of life for lots of people.