"Smallholder farmers produce 70 per cent of the world's food." What's the source for this number?

“About 70 percent of the MDGs’ target group live in rural areas, particularly in Asia and Africa,

and for most of the rural poor agriculture is a critical component in the successful attainment of

the MDGs. Even though structural transformations are important in the longer term, more

immediate gains in poor households’ welfare can be achieved through agriculture, which can help

the poor overcome some of the critical constraints they now face in meeting their basic needs”.

World Bank, IFPRI. Agriculture and Achieving the Millennium Development Goals, n.d

"Three-quarters of the world’s poor live in the rural areas of developing countries and depend primarily on agriculture for their livelihoods and overall well-being." Rural- A joint donor concept on rural development (Global donar platform for rural development)

Hi Prakash - thanks for engaging with this question. Like you, while pondering the source of this statistic I thought about the fact that around three quarters of the world's poor live in rural areas and depend on agriculture, but that's not the same thing as saying they are responsible for 70 per cent of the world's food production (nor for producing 70 per cent of the world's food, which is almost but not exactly the same thing).

Cheers. Dominic.

A good recent report is by IFAD and UNEP available at http://www.unep.org/pdf/SmallholderReport_WEB.pdf. This highlights some of the stats and issues

Hello Tim and thanks for the link. That UNEP/IFAD document is one of the sources I checked earlier. It's an interesting and useful report in general.

With regard to my specific question, the report includes the 70 per cent figure (on p.11) and attributes it to the IAASTD report, vol V, which I quoted in my original post. However, as I noted above, the IAASTD report itself refers specifically to SSA, and does not provide a satisfactory original source for the 70 per cent figure. So we're still no closer to a source for the claim that small-scale farmers are responsible for 70 per cent of the world's food production.

I'm beginning to wonder whether the 70 per cent figure is simply a (poorly phrased?) inference drawn from the fact that 70 per cent of the population in places like SSA and India rely to a greater or lesser degree on subsistence production, ergo that subsistence cultivation must be responsible for the food security of 70 per cent of the population.

I'm still left wondering whether the 70 per cent figure is a reliable estimate or a factoid.

Regards,

Dominic.

Dominic i also agree with your last opinion that it might have been drawn from the fact of those number of people relying on subsistance production!

this appear in many agricultural production reports and few refer to the sources. However, there are many facts to support the claim that small scale farmers are the main agricultural producers in many nations especially the developing world. Large scale farming is sophisticated and affordable to few of the farming communities,but the statistics are varied. The World Bank and FAO have quite some statistics on this

Hi Glover,

You could try the agricultural census report, prepared in most of the countries in regular time intervals. This would provide some idea on extent of land use by marginal & small farmer categories and their land use in agriculture. With average yield levels in those regions, you could fairly decide their contribution. Yield levels, since, are not reported in censuses, you should depend on the government sources anyway..........

Original by (IAASTD 2009): “In SSA, 70% of agricultural production is subsistence farming”.

To become in (UNEP 2013):”Smallholders […] produce 70 per cent of Africa’s food supply”.

To then become: Andre Leu, IFOAM President, "smallholder farmers produce 70% of the world’s food”.

Thank you Grace, Balaji, Wolfgang and Bill for your recent contributions.  I'm sorry for my slow reaction - this debate had slowed down a bit and I wasn't paying so much attention.

Wolfgang, I think you've encapsulated the problem surrounding this little factoid - a kind of 'Chinese whispers' whereby a claim based very loosely on fact, which sounds plausible, can travel through networks and be transformed gradually in the process.

But of course to say that '70% of agricultural production is subsistence farming' is subtly different from saying that 'smallholders produce 70 per cent of Africa's food supply', and 'smallholder farmers produce 70% of the world's food' is quite a leap from there.

Bill, I have not spent a lot of time on trying to investigate this statistic.  My tentative conclusion from the discussion here is that it is an unreliable factoid, which we should handle with great care.

I think it's true that small-scale farming is the dominant form of production in SSA and other regions of the world, which implies that smallholders are chiefly responsible for feeding very large numbers of people, the majority, particularly in those regions.  But that is not at all the same as saying that smallholders produce 70% of the world's food.

I do think it would be interesting to try to pin down the facts on this but that's not a project I'm planning to take up myself.

Thanks again for all your contributions.

Bill - Agreed.  And small-scale producers in those countries/regions, within reach of cities, are responsible for feeding large urban populations, not just rural backwaters.  So I don't doubt the importance of smallholder production in any way.

Still, I think it's important to be careful about statistical claims.  I also think it would be very interesting to see some robust estimates of what proportion of the world's food is produced in different production systems, and the proportion of the global population that depends on small-scale production.  But, as I say, this is not a project I plan to take up myself!

Thanks again for engaging with my question.

Dominic.

Yes - or to put it another way, they'll fund R&D to fit how they think agriculture ought to be, based on modernist assumptions about the desirable obsolescence of small-scale systems.

I agree with @ Bill Bellotti.

People following this discussion might be interested in a new report published by GRAIN (Genetic Resources Action International), which is entitled Hungry for Land: small farmers feed the world with less than a quarter of all farmland.

The report is really about land grabbing and it analyses international data on access to land, but what's interesting from the point of view of the discussion above is GRAIN's premise, which is as follows:

"It is commonly heard today that small farmers produce most of the world's food. But how many of us realise that they are doing this with less than a quarter of the world's farmland, and that even this meagre share is shrinking fast? If small farmers continue to lose the very basis of their existence, the world will lose its capacity to feed itself."

Again, we see it taken for granted that small farmers produce most of the world's food, and from that starting point it is taken to follow that if small farmers cannot continue to farm, the world will go hungry.

By the way, there is a very brief summary/discussion of the GRAIN report in the latest PLECserv newsletter (July 2014).

I thought readers here might be interested in these new developments.

Dominic.

Dear Dominic,

I meant to respond to this earlier. Your colleague at Sussex, Pedram Rowhani, pointed me to this thread. I have come across a source that a student of mine pointed me to, that may be the original.  See the executive summary of the attached file -- it's not fully clear to me how they arrived at this number, I need to read this more carefully. But from what I have seen so far, it doesn't seem very rigorous. A colleague at UBC does have a paper in review where they have estimated, based on a recent FAO world census of agriculture, that 53% of all agricultural land is managed by "family farms" (and if yields are equal, then by extension, 53% of production).  Note that their definition of family farms is not the same as small farms; it is also much higher than the 25% value in the 'GRAIN' report you pointed to, which is for small farms.

In any case, I have a hard time believing these numbers.  Rice & Wheat are the top food crops (majority of maize being feed), and while I can believe that there are a lot of small rice farmers, with wheat I suspect that a large portion is produced in large farms. In any case, this is an interesting question I would like to revisit.  I have been talking to Pedram and my UBC colleague about this. 

Navin

HI Paul, Navin - welcome and thank you for contributing to the discussion.

Another correspondent drew my attention to the ETC document you linked to.  I agree with your impressions of that document!

I will be very interested to read your UBC colleague's article when it is published.  Could you ask him/her to send me a copy, and also post a link here?

Kind regards,

Dominic.

I'm not sure if anyone is still following this discussion, but just in case, you might be interested in this new article that has just come out in Global Food Security, which considers the importance of family farming and issues related to farm size, worldwide:

van Vliet, J. A., et al. (2015). "De-mystifying family farming: Features, diversity and trends across the globe." Global Food Security 5(0): 11-18. http://dx.doi.org/10.1016/j.gfs.2015.03.001

Abstract:  Family farms are defined by two criteria: the importance of family labour and the transfer of ownership, land tenure or management to the next generation. Most farms across the globe are family farms, and they vary in size from <1 ha to >10,000 ha. Trends in farm size (small farms getting smaller and large farms getting larger) are not directly related to farm ownership and do not necessarily impact global food security. Rather, both the causes and effects of farm size trends depend on the availability of farm resources and off-farm employment opportunities. Similarly, environmental sustainability, though impacted by agriculture, cannot be linked directly to family ownership or farm size. To address issues related to environment, social conditions and food security, focus should not be on the preservation of family farms but on transformations to strive for environmental, social and economic sustainability of farming in all its shapes and forms.

Thank you all for this interesting debate on statistics. I have been working on my bachelor's thesis and found it extremely difficult finding reliable data on smallholder's production capacities. There might be many reasons for that, but it's probably quite close to reality admitting the fact we won't get any stats on that. We can, literally, just guess with certain models.

Speaking of that, I can only imagine how that would be like. I have lived in Malawi and the Philippines. The immense amount of maize being produced by smallholders in Malawi for example, ground in small commercial mills not paying taxes, is just insane. This is happening far from big commercial mills around the country. The infrastructure in so many countries to give reliable stats on production.. I wouldn't trust them based on what I've seen and experienced myself.

Hello H. T. Ngo and Gregor Rahn,

Thanks for contributing to this discussion.  Indeed it seems we are still searching for some reliable statistics and methods for pinning down this question.  And one key problem is the quality of statistics available on smallholder production.  Such statistics are hard to collect and mostly quite unreliable, I believe.

It's also quite correct to point out that it makes a difference which crop we're talking about.  I'm sure that the great, overwhelming majority of the world's wheat and maize are produced by large farms, even though these crops are also important for millions of smallholders.  Rice is another matter I think.  Small-scale production systems are still hugely important in rice.

In a way, this highlights why this question is important for understanding the dynamics and structures of global food security.  In major, globally traded grains and oilseeds, there can't be much doubt that the big producers dominate.  Yet there are many 'minor' crops and locally traded food commodities on which many millions of consumers depend.

I am sure there are substantial rural areas, distant from substantial markets, where people are still fed largely from local or regional production, including many small farms.

But even some big cities are fed to a substantial degree by small and medium producers in surrounding districts who cart fresh produce into urban markets on a daily or weekly basis.

Then there are cooperatives, such as Amul's milk production networks in India, which pool the production of many small producers.

Personally I would find it quite credible that small-scale producers contribute a substantial share of the world's food, but it seems we just don't have dependable statistics to prove it.

Thanks again to everyone for engaging with this question!

Hi Dominic,

I find this discussion highly useful. I was wondering whether you researched this topic further. I can see in the article indicated by Navin, that a reference is made to Hoering's article 2008. Here a Brazilian agro-sociologist is cited indicating that "family farms in Brazil are still producing 70 percent of the food". This statement is very different from the "smallholder farmers producing 70 % of the world's food", statement. I don't argue that it is not correct but that indeed the sources for claims are important. I first heard the claim in one of Patrick Mulvany's podcasts: http://www.theguardian.com/global-development/audio/2012/nov/21/global-development-podcast-farmers-gm-agriculture. He states that 70 percent of the world's people are fed locally and mainly by peasant agriculture. He mentions that the statement is based on research of the International Food Policy Research Institute and several NGOs. 

ETC Group (2009) With Climate Chaos, Who Will Feed Us? The Industrial Food Chain or The Peasant Food Web?

at least explains how they arrived at the 70% figure and gives some sources but probably not definitive

Thanks Anna, John and H. T. Ngo for your contributions to this discussion.

Ngo, I am intrigued that you are digging into this question with considerable persistence and energy and I will watch with interest to see if you come up with a solid statistic in the end!

Hi all, I passed a relevant paper on to Dominic privately earlier (as it was still under review), but it has recently been published (see link below). Please take a look at a paper that finally uses data to estimate how much of the world's farms are in family farms (but pay attention to how they define family farms, and further how they actually derive the data in the supplement). They also estimate how many calories are produced by these farms. In any case, no matter whether you agree with the definitions or quality of data, here is a transparent publication on the topic finally.  

Link: http://www.sciencedirect.com/science/article/pii/S0305750X15001217

Navin - great to know that the paper is now online and finally we have some concrete information to think about!  Thanks for posting.

I am most grateful to all the contributors. I am working on my literature review on sustainable farming practices and some of the links/attachments are most useful. Thank you everyone. 

Hi I too am interested in the source and reliability of this statement. I find references to  80% of food produced in some regions, but not a global figure (below). Any further information is very welcome as I hear this quoted at the UNFAO and I can't substantiate it. Many thanks, Emily.

IFAD

There are some 500 million smallholder farms worldwide; more than 2 billion people depend on them for their livelihoods. These small farms produce about 80 per cent of the food consumed in Asia and sub-Saharan Africa.

https://www.ifad.org/documents/10180/ca86ab2d-74f0-42a5-b4b6-5e476d321619 

UNEP

These smallholders manage approximately 500 million small farms and provide over 80 per cent of the food consumed in large parts of the developing world, particularly Southern Asia and sub-Saharan Africa, thus contributing to food security and poverty reduction.

http://www.unep.org/NEWSCENTRE/default.aspx?DocumentId=2718&ArticleId=9524 

Dear all! You are invited to participate in the on-going moderated e-mail conference hosted by the Food and Agriculture Organization of the United Nations (FAO) from 10 to 23 October 2016 entitled “Exploring the contribution of small farms to achieving food security and improved nutrition”. The conference is part of the EU-funded Horizon 2020 research project on “Small Farms, Small Food Businesses and Sustainable Food Security” (SALSA, http://www.salsa.uevora.pt/en/). For particpating in the e-mail conference see: http://www.fao.org/nr/research-extension-systems/res-home/news/detail/en/c/434322/

Hello all - I was a co-lead on the paper Navin linked to.

We spent a lot of time looking over this question and found that, as best we could tell, the 70% figure came from extrapolations of expert "guesses," and a variety of these guesses get repeated and become factoids within the literature. I think the sources you identified, Dominic, pointing to SSA and then extrapolating from there are likely to be the "ultimate" sources (though my co-author Hannah Wittman (https://www.researchgate.net/profile/Hannah_Wittman) and her then-research assistant Anelyse Weller (https://www.researchgate.net/profile/Anelyse_Weiler) spent a good amount of time on a related project and found some specific likely sources for a variety of "guesstimates."

Personally, I consider our 53% estimate to be the best "best guess" and I don't find it likely that a much better estimate can be made at present. Our estimate could be improved by surveying countries' ag censi that we could not read (e.g. Russia's, which may or may not contain pertinent information but that we were not able to translate in order to figure out). And with very detailed statistical analyses and quite a bit of time, it seems likely one could use an information theoretic approach to make data-derived guesses for missing numbers. But I would be surprised if there is significant systematic and reliable data beyond updated censuses, or censuses we missed or could not read, and with a couple of exceptions (like Russia) they would be unlikely to significantly change the estimate.

Personally, I consider 53% to be a conservative estimate since we used fairly restrictive and detailed definitions of small family farms wherever possible, and given the IR (inverse relationship between farm size and productivity), which Barrett et al. 2010 (as we note) admit is definitively the norm, not the exception.*

(*Sidebar: While it is quite possible that data and analysis will eventually be obtained that refutes the common observation that smaller farms have higher per area yields than larger farms, I find it bothersome that the fairly consistent evidence to date is often ignored or discarded because it does not match [economic] theory. Ecological theory does not necessarily have any quarrel with the phenomenon; it seems like we should take a position of greater confidence in the consistent result, contingent on further information, rather than continue to treat it as an anomaly that some day we will explain away.)

Navin's point about wheat and rice is interesting, and offers food for thought (excuse the pun). But given the sheer numerical majority of small-scale family farmers (98% of all farms!) and their occupation of slightly more than half the land, plus the inverse productivity relationship, the idea that they produce a majority of the world's food is quite plausible to me. And of course, since the population suffering from undernutrition is likely equal or less than the population with micronutrient deficiencies, the amount of importance we should accord to the caloric share of food produced is an important question. Arguably, caloric share and nutritional share/diversity should at the least occupy co-equal importance in these kinds of conversations.

Sorry I couldn't be of more help, but hope that is of some help.

Thank you very much M. Jahi Chappell for joining this discussion with some helpful insights into your methodology and analytical reasoning.

Like you, I find it plausible that small-scale growers produce a high share of the total food available on the planet, perhaps the majority.  However it is interesting that we don't have solid numbers to demonstrate whether or not that is the case.

Judging by the steep methodological challenges and data scarcity and quality issues, it seems more likely to be possible to arrive at plausible estimates for individual countries (where reliable data exists and definitions of farm size can be developed contextually) rather than a number for the world food supply as a whole.

I also share your frustration that the inverse farm size : productivity relationship is so well established, yet somehow does not gain any traction, and I agree with your suggestions about why that is so.

Thanks again for pitching into the debate.

I agree. I think the data for Brazil can yield far more information than we explore in our paper, and there are a number of other countries where this is true as well. (And ignoring data quality for the moment, some countries, like Brazil, have both area and sectoral production numbers for different farmers, so the IR can be checked at large scales. I did some VERY initial back-of-the-Excel-spreadsheet calculations on Brazil's data and my initial foray did seem to confirm IR at this nation-wide level...  now that I type that, it seems worth following up on with a paper....)

We (and the FAO) compiled a good amount of data on individual countries in order to derive our final estimates, and most of the individual-level country censuses are publicly available, so (again begging the question of reliabilty) it does seem like plausible estimates for a number of countries can be made. (And indeed, our hope was that our global total was a plausible estimate based on accumulating plausible country-by-country totals!) But we did extrapolate datapoints for some parts of the analysis, so an analysis without that could be done, and of course, an attempt could be made to analyze only those countries where data quality confidence is higher. 

Though ag census data quality also seems to be one of those things that just gets more and more fuzzy as you examine even the best cases more and more closely...

Cheers!

Yes, on the same page about working with data from representative countries around the world, I have tried to search and required global data don't exist. To answer the 'food produced by small farms' question, we need census data with "crop specific" data on farm size distribution.  Many countries provide overall farm size distribution, but not crop specific, which is required to identify what's grown for food versus other things.  We have done this already for Brazil, which has such data, and are doing it for a few other countries.  Let me know if you are interested in collaborating.

And yes, the inverse size-yield relationship seems very well established.  As part of a comprehensive exam, one of my PhD students just did a systematic review of the literature on this, will share with you when we have a good first draft. And as Jahi says, the census data supports this relationship, but I have more confidence in the many more studies that have collected primary data at the farm level.  Agricultural economists have looked at this for over 50 years, taking the inverse relationship as being established, and trying to explain it.  But, on the flip side, this is the only relationship for which I think we have good evidence for an outcome that is related to farm size.

Finally, one pet peeve.  Many studies conflate "small" farms with "family" farms.  Jahi's paper with the 53% estimate is of family farms (at least from my reading of it, although I saw from the supplement that sometimes when censuses don't report ownership, you use size as a proxy, often 10 ha).  The results can be vastly different depending on this definition. Take Brazil for example, where 84% of all farms are family farms, using 24% of total agricultural area.  But, by "size", only 21% of farms are <2 ha (using the universal threshold) using only 0.25% of the agricultural area.  Of course, what is small remains open to question, and clearly small in one part of the world is not the same as small in another part of the world.  And therefore even trying to calculate how much food is grown by smallholders probably needs to be a context dependent calculation (maybe take the lowest quintile of size?  or just based on ownership?).  In the end, I am starting to think that this whole 'farm size' question may be a little bit of a 'black hole', so trying to figure out what the important question really is in this context.

Navin -- I think there are likely trends of disproportionate economic power, differing levels of landscape heterogeneity, and other elements of socio-economic justice tied to farm size. While I feel many of these are likely contingent on historical processes, and are not determinative, that does not change the fact that the relationships that (likely) exist between farm size and socioeconomic equality and rural livelihoods are currently present, even if they can be changed. (Indeed, the evidence for and against the "Goldschmidt Hypothesis" suffers many of the flaws we've been referring to, but so far in my opinion a preponderance of evidence points to smaller average farm size being somewhat correlated with rural community wellbeing. They are likely co-correlates of other social factors, but that again wouldn't mean that there is no "real" relationship between the two.

further, as I'm going to point out in a forthcoming book, land reform is an immensely important issue (and one that will likely become more and more salient in the U.S. context in my opinion). Michael Lipton points out that redistributive land reform has a very strong track record for relieving rural poverty, but that this is perhaps one of the most commonly overlooked insights in the field. (Jun Borras has similar conclusions from the point of view of critical agrarian studies). While again farm size does not DETERMINE land reform dynamics, it also is not at all irrelevant for them.

As for our paper, I think the final draft unfortunately cut some of the details of our analysis. But we were explicitly concerned with "small family farms." Our area based determinations were only when more data wasn't available, but we sought to define them parallel to Berdegue & Fuentealba's multi-faceted and country-specific advisements.

Given the actually existing history and political economy, I'm not sure farm size gets enough attention, though given the choice between EITHER farm size OR land reform, I would certain prefer for more attention to be given to the latter than former.

Oh, and Tim Benton and colleagues did find a strong relationship between farm size and biodiversity in the UK and farm size (though they were careful enough to note that this was true for some taxa and not others). As I implied, I suspect this relationship is not limited to the UK. It is (at the risk of repeating myself) likely contingent in some significant part, but given we have the history we have and not an alternate one, that contingency is only partly relevant.

I found reference to "family farmers produce 80% of the world's food".  http://www.fao.org/3/a-i4036e.pdf

But I don't think family farmers is the same as smallholders. Neither do I know the definition for 'family farmers'. 

Hi All - You might like to read our World Development article which provides definitions for family farm and small farm and the likely contribution of each to food and ag production globally. 

Thanks very much to Sarah Lowder and Emily Lewis-Brown for sharing those useful documents.

The article by Sarah and her co-authors contributes to this debate in an interesting way by distinguishing between family farms and small farms and showing that they are responsible for managing radically different land areas in total.

The article also confirms that while land consolidation continues in richer countries, in poorer countries land continues to fragment into smaller holdings.  While this insight is not about the '70 per cent' claim we've been discussing here, it strikes me as something crucial for policy makers and politicians to understand, especially those responsible for directing aid and development programmes from the 'North' into the 'South'.  The trends in small-scale/family farming are quite different in these two regions (broadly speaking of course).  In some contexts (e.g. in Brazil?) we see the two dynamics operating side-by-side in different regions and production systems.

Reading Sarah Lowder et al's WD paper clarifies the distinction between otherwise interchangeably used terms: Family Farm and Smallholder Farm. It indicates at least for me, Family Farm is much more than Smallholder Farm. The latter may be part of former. It is good to learn from this discussion. Thank you.

A very interesting, even more important and high quality communication. I still have to get and read many of the sources recommended, but it is essnetial to have knowledge about who is producting the food we eat as from this information essential policy implications arise that address food security as well as livelihood security.

Thanks Eberhard for engaging with the discussion. I'm glad you found it useful.

I missed this wonderful debate when it was happening but I read every contribution including the articles. It was very helpful indeed. Thanks very much Dominic Glover for raising this question and all those who contributed to the debate. I benefited from this discussion.

Yes, very useful discussion indeed, even if I missed it when it happened - it's still relevant! Thanks to all.

Very enlightening discussion. I have read every contribution. Thank you Dominic and everyone who contributed.

Dear all,

Please find below a recent paper (in Global Food Security Volume 17, June 2018, Pages 64-72) which provides some contradictory insights

Thanks for this stimulating discussion, very useful for teaching: both substantive food security and broader methodological matters :)

I'm amazed that the statement from Graeub et al. 2016 got past World Development reviewers: "We find that family farms constitute ...at least 53% of agricultural land, thus producing at least 53% of the world’s food" (?? - really?). Many thanks to Alcade C. Segnon for the suggested paper.

Not sure why you find it amazing our statement was accepted by reviewers, @Alastair M. Smith. As noted in this discussion and our paper, the inverse productivity relationship between farm size and productivity is an important mediating factor here. Interestingly, Ricciardi et al., using their spatially explicit [EDIT] approach "based on compiling available subnational data on harvested area or production by farm size... for either area or yields, by crop type, for different farm size categories, for 55 different countries" [/EDIT] seem to confirm that. [EDIT] As Navin notes below, their study

"also had to make assumptions about yields (because only a few countries had production data, for the rest we only had area). We assumed uniform yields (conservative assumption that biases against smaller farms) and tested the assumption in the paper. So the real information on 'food' did not come from differences in yields, but based on differences in what types of crops were grown by smaller versus larger farms."

My original response identified their study as using remote-sensing, which was incorrect. I have edited this response to accurately reflect their study. [/EDIT]

In Graeub et al., we assumed a proportion of production equal to land possession since we didn't have direct productivity numbers, and some (internal and external) reviewers argued that perhaps smallholders writ large would produce less because of lack of access to (agroecological or "conventional") inputs. Though we disagreed with this based on the existing inverse productivity data in the literature, we "split the difference" to make the conservative estimate of productivity matching land possession. It would have been questionable, in my mind, to globally extrapolate too much additional productivity based on the previous site-specific inverse productivity research, and it would have been against existing data and theory to extrapolate lower per area productivity. Our approach (to me) then seems both parsimonious and reasonable given the data we had, and seemingly has been confirmed as conservative by Samberg et al. and Ricciardi et al. Given that we didn't have the data they had, I'm pleased that our explicitly rough estimates appear to have been indeed reasonable and in the ballpark of subsequent estimates based on further data.

I just wanted to add some clarifications following Jahi's response above. First, Ricciardi et al. (How much of the world's food do smallholders produce? Global Food Security, vol. 17, 2018) was not a remote sensing based study, it was based on compiling available subnational data on harvested area or production by farm size. We had data on either area or yields, by crop type, for different farm size categories, for 55 different countries. Second, we also had to make assumptions about yields (because only a few countries had production data, for the rest we only had area). We assumed uniform yields (conservative assumption that biases against smaller farms) and tested the assumption in the paper. So the real information on "food" did not come from differences in yields, but based on differences in what types of crops were grown by smaller versus larger farms. In the same paper we show that smaller farms grow more food crops while larger farms grow more crops for feed or processing. Third, Graeub et al. paper was about "family" farms, not "small" farms -- one should not conflate the two. From our paper, "For example, farms in Brazil may be family owned but are large in size (while ~ 85% of farms in Brazil are family owned and cover ~ 25% of agricultural land, only 21% of farms are less than 2 ha in size and cover only 0.25% of the agricultural area)." So hard to directly compare Graueb et al. to our paper. Fourth, the estimate of food produced by "small" farms depends on what threshold you choose for small. We estimate 30-34% (for <2ha threshold for small farms), 44-48% (<5ha threshold) or 62-66% (<50ha). In the end we concluded that smaller farms do produce more food disproportionate to their area covered (e.g., from abstract, "We estimate that farms under 2ha globally produce 28–31% of total crop production and 30–34% of food supply on 24% of gross agricultural area." (note the distinction between production, which is total mass of crops produced, and food supply, which further looks at the use of the produced crop).

Thank you for the corrections, Navin. Though I would say that there are good reasons to associate small farms and family farms, and that conflation is, to a first approximation, a reasonable one, especially if one takes seriously the idea that both "small" and "family" are context-dependent definitions. I personally think grouping farms globally by <2ha or <5ha threshholds obscures as much as it clarifies given social and ecological differences in context.

But yes, one should not use family and small farm interchangeably because they are, in fact, different categories indeed. But I also think that the data implies the categories have substantial overlap for a number of reasons that appear consistent over many, though not all, locations and contexts.

I agree that it is context dependent. But take my Brazil example in above response and let me know if you think one could use the definitions interchangeably there. But yes, the same threshold should not apply for all countries. More broadly, I think we could define "small" in multiple ways, and family owned is one definition people use, while some others use economic measures. Vinny Ricciardi is planning to examine how some of these measures (food produced, crop diversity, etc.) vary by different definitions of 'small' (or other criteria of importance).

It's exciting this thread has been going on for so long now! A very important topic. Navin, thanks for clarifying our paper - food production would definitely be dependent on the crops grown, yields (which vary by farm size), and the actual supply chains of the crop - is it actually food, feed, seed, or other?

In response to the family/size discussion... I always wonder what types of farmers are actually captured by farm size definitions or the family farm definitions - or some combo of farm size and family farm as in the Graeub paper. It would be interesting to dissect which demographics (or other social and environmental characteristics) these two dimensions of smallholders actually capture.

Thanks to Alastair, Jahi, Navin and Vinny for your further contributions to this thread. It is indeed exciting and fascinating that this question has engaged colleagues' interest for several years now, and great that advances are being made in terms of research, data and analysis. Congrats to those who are contributing to this literature on farm size X production/productivity, and thanks on behalf of those of us who appreciate your efforts!

Navin -- well, I suppose I would say for example that one could use "family farm" and "small farm" interchangeably in Brazil for certain definitions of "small", as you say! ("I think we could define "small" in multiple ways...") That is, while there is a dramatic difference between the 85% of family farms on 25% of agricultural land and 21% of farms being <2ha and occupying only 0.25% of agricultural land, in the context of many parts of Brazil, <2 ha is possibly not a "viable" farm, and certainly should be classed as something like "very small farm" in that country. In other countries, < 2 ha is possibly socioeconomically and environmentally viable, and thus is a small farm of a different kind.

I think Berdegue & Fuentealba's classifications of family farms could similarly apply to small farms (and I would hypothesize, would lead to similar groupings in the end): "the broad term “family farming” can be divided into at least three groups with differing needs: those that are well-endowed and well-integrated into markets (“Group A”); those with significant assets and favorable conditions but lacking critical elements (like sufficient credit or effective collective action) and who may not qualify for social safety nets (“Group B”); and land-poor farmers, who are primarily characterized by family subsistence/non-market activities and who require significant investment in social safety nets (“Group C”)". These groupings lack tidy biophysically-delimited definitions, and are contingent in space and time, but that is perhaps as it should be.

So it seems to me that 85% of farmers being on 25% of the land does not automatically make them "small" farmers, but certainly implies that is a reasonable classification of them given the disparity between their proportions and holdings. As we say, I believe, many of the larger family farmers in our data set would be Group A and B; in Brazil, it seems likely many if not most <2 ha farmers would be Group C. (I would hypothesize < 5 ha farmers would fall mostly into Groups B and C.) Though these are all somewhat open empirical questions! Glad Vinny is working away on beginning to reveal what the data actually can support about some of them.

In the end, I would propose that something like Reddy and colleague's ideas on the context-dependent classification of poverty are necessary for farming and farm typology as well. See http://www.sanjayreddy.com/poverty-global-estimates/ , e.g. or . As in Poverty Beyond Obscurantism, perhaps only measures of small farms "that are consistently grounded in a suitable value framework can provide a sound basis for public discussion and decision-making."

Very interesting discussion. I am researching on the role agrarian community in formal law and decision making using their local knowledge and practices with focus on Kenya and I have come across the 70% of world food contributed by smallholder farmers. I initially took this at face value. However, the exchanges here have been very enlightening. Thanks Glover et al.,

This is a very interesting effort to understand just what is produced by the world's smallest farms...and what contribution they might make to our future food security needs. The topic has my attention because of work we've been pursued to do regarding protecting phytopharmaceutical crops, and wondering how do we get small land holders to participate in the commodity boom windows that the new science applications in crop design will bring to the market? We're trying to look for ways in which to expand those able to benefit from growing for such booms...at whatever level we might find growers.

With so many small land holders locked into agricultural skillsets in the midst of substantial rural poverty, and the very rare likelihood of capacity to replicate China's massive reorganization of so many millions of people off the land and into overnight cities, how do we touch small holders lives in systemic ways to enable them to participate in booms like we're on the front edge of experiencing in hemp/cannabis? The lifting through plants like cannabis/hemp will have applicability in so many environments, MDCs through LDCs, as it's a historical opportunity for people to participate in a 'gold rush' without moving to where the gold might be found. Instead, matching variety to environment of soil and water, the 'gold' can be grown where our lives have left us. We've already reached out to 11 countries to explore how to push the opportunity to small holders...and the discussions are very fascinating. 'Commodity booms' in areas where rigorous regulation of trade practices is lacking, can result in nations being poorer after the boom than before it (e.g., Nigeria and oil). Good regulation can lift significant segments of a population out of poverty. In addressing governments as to how to encompass such a boom as is organizing itself in cannabis/hemp, we keep trying to understand just what is the role of subsistence farms a given national arena. Last thing I want to do is stand before legislatures and regulatory bodies, and not have data rooted some citable reality.

The articles cited above have been very useful when not already a part of the bibliography we've assembled in research on the 'who' of the bottom of the pyramid.

Pakistan FRESH FOOD economy is 70 % RUN BY SMALL FARMERS

Farukh - this is interesting. What is the source for this number? Could you share a reference or link? Thanks a lot!

Agriculture Census http://www.pbs.gov.pk/content/agriculture-census-wing

Great idea Jan! I will be in touch.

Overly overstated apparently. See recent article by Ricciardi et al.

The FAO document states: "As providers of nearly 70 percent of the world’s food supply, smallholders and family farmers (....)". Therefore, FAO is not considering only smallholders.

Additionally: Is it reasonable to consider that smallholders are those with less than 2 ha?

Our research among small-scale tree-crop farmers in Ghana (cocoa, oil palm) and South Africa (macadamia nuts, avocado) shows that average land sizes are more than 2 ha (but generally under 10 ha). Reason to unpack what smallholders are!

I agree, Grace Mwaura. The received wisdom is that small farmers are ageing and dying, and that most young people want to leave agricultural occupations and rural livelihoods, migrate to cities, and seek other kinds of work (perceived to be more glamorous, modern, higher status, etc.).

Of course, many young people still have little choice about what they do for a living - some of whom end up working in farming-related occupations against their will.

Equally, some might be forced to migrate because it's too difficult to make a decent living from small-scale agriculture.

On the other hand, we also see some evidence here and there (from global North and South) that some young people, including college graduates, see opportunities in agricultural occupations, to build the kind of life they want for themselves and their families. This includes some of the work you have published.

Maybe this is a topic for a new thread. Shall we consider starting a new discussion? We could begin by gathering examples of published research into the topic of young people and small-scale agriculture-based livelihoods.

Dominic Glover If you create a new topic, could you notify me? I study family farming in Brazil and I have interest in this topic.

Dear and distinguished Dominic Glover and Lilian De Pellegrini Elias

My next research project is smallholders and family farming agricultural productivity, innovative agricultural water management solutions and efficiency: With Special Reference to SSA. Hence, I would be very appreciative if you could kindly notify me when you have (create) new topics.

This is possibly one of the most useful threads I've ever read on this platform. Many thanks to all. Dominic Glover - I would also love to be notified of any sub-threads that are started related to this one.

Just a note as to the continued relevance of Dominic Glover 's question: the figure (now over 80% of food) lives on in draft background documents for the UN's Decade of Family Farming, but the supporting reference is FAO 2014.

Hi everyone, I have been tracking the 70% statistic, as Dominic initially asked then as Jan suggested to track the narrative. I was thinking of writing a blog post on it. Is anyone interested in helping catalogue where the 70% statistic (or similar figures) have been used in the media, grey literature, or academic literature? So far, I've found about 20 different sources. If you want to contribute to tracking this number with me, please reach out or simply start adding your findings to this Google Spread sheet (link below). It would be neat to track this "zombie statistic" and people on this thread seem to be very keen on finding where its been mentioned. https://docs.google.com/spreadsheets/d/1PuJa7oJXTSGh5Xikdwd_Y3SNZT21Bahn6CMvJwkRQIc/edit?usp=sharing

And I should mention, any blog post/writing about this would want to contextualize each source and not seek out to shame the authors. I see this as a way to highlight how little has been quantified about smallholders at the global level - a key action item needed.

Hi everybody. I am following the discussion some time ago. My PhD thesis (that I am writing at this moment) is about smallholders aspirations in Mexico. I am working with small cattle ranchers, and I analyzed their aspirations according to their narratives. Technically maybe they are "familias campesinas", they have more than 2 ha, but least than 20. However, after the reform to agrarian law in 1992, anybody could buy more land, and today there are some families with more land surface. Their aspirations change by elders and young adults groups (I don't have information about younger people because of my field work design). I am very interested in the new forum that Dominic Glover propose. I can see it is strongly interlinked with the future of smallholder (and "familias campesinas") livelihoods around the world.

Hi All! Since 2012, my area of research and practice has been in engaging young people in agriculture in the Philippines. My team and I have extensively written about this and they are all in my page. I agree this requires a lengthy discussion. At some point, in one of our papers, we asked-- is it really a case that young people do not want to farm or they just to do farming differently?

Alejandra Tauro and all - Despite the 1992 (counter) reforms, a very large proportion of Mexico's agriculture land remains under multi-layered tenured regimes of collective ejidos/communidades, coupled with de facto family owned farmlands. For a comparative analysis relevant to your PhD, see, https://www.amazon.com/Democracy-Woods-Environmental-Conservation-Comparative-dp-0190935502/dp/0190935502/

The trap of policy numbers: The policy space is so complex that a number dangled will attract commentators. Before long, the number becomes the defining force of the policy space. So, we must laud the few people who dare question the dominant numbers. However, the policy space is opening fast to the public, and particularly in the lower income countries. Yet, the sanitisers are too few. The paper below exposes another misused number but in infrastructure.

I found this information about Colombia: 83% of the food is produced by the peasants in Colombia. https://www.minagricultura.gov.co/noticias/Paginas/El-83-de-los-alimentos-que-consumen-los-colombianos-son-producidos-por-nuestros-campesinos.aspx

I have just stumbled across this thread and I have a question about type of foods produced. Specific to meat production/consumption, I am under the impression the bulk of meat consumed globally emanates from industrial (large-scale) meat production. However, I am wondering if anyone knows of a citation with data that provides evidence for this claim? Or counters the claim?

Herrero et al 2017 found that farms < 20 ha produced ~50% of the world's livestock and farms < 2 ha produced ~25%. See their Fig 1 in https://www.thelancet.com/journals/lanplh/article/PIIS2542-5196(17)30007-4/fulltext#seccestitle130

I think Herrero is the best number for this to date, but keep in mind that they used a few global overlays (field size map, national farm size data, and crop/livestock production datasets) to come up with the number - so they did not have direct farmer reports of which size farms produce which commodities or quantities. In the SI to our article, we wrote a bit more detail on why this distinction matters:

Vincent Ricciardi Thank you! Very helpful!

I've always found the studies put out by ETC Group incredibly well-sourced. Their report "Who Will Feed Us?" updated in 2017 at least gives the best stats they could find. http://www.etcgroup.org/sites/www.etcgroup.org/files/files/etc-whowillfeedus-english-webshare.pdf.

I always wondered about this statement. It is conjecture. Not based on facts!

In India, where smallholders contribute 70 percent of the country’s agricultural production (Nwanze and Fan,2016). Globally, there are about 525 million farms out of which small­holdings of less than two hectares (ha) constitute 85%. Out of this 87% of smallholding farms are located in Asia, followed by Africa (8%). The rest 5% smallholdings are located in Europe and America. In Asia, China stands first in con­centration of smallholdings followed by India, Indonesia, Bangladesh and Vietnam. The number of small farms in China in 1997 was 189.4 million accounting for 47% of smallholders in the world (Chand et al.,2011).

Nwanze, K F., and Shenggen, F. (2016) CLIMATE CHANGE AND AGRICULTURE; Strengthening the Role of Smallholders chapter 2; In 2016 Global Food Policy Report. Washington, D.C.: International Food Policy Research Institute (IFPRI). http://dx.doi.org/10.2499/9780896295827

Ramesh Chand, P A Lakshmi Prasanna, Aruna Singh (2011)Farm Size and Productivity: Understanding the Strengths of Smallholders and Improving Their Livelihoods.EPW. xlvi (26 & 27);5-11

Hi there, certainly a very intriguing question.

During my literature review, I´ve come across these three sources, yet they don´t clearly state the global percentage of family farms, they do show its importance when providing food.

- Herrero et al. Farming and the geography of nutrient production for human use: a transdisciplinary analysis

- Lowder, S.K., Skoet, J. and Singh, S. 2014. What do we really know about the number and distribution of farms and family farms worldwide? Background paper for The State of Food and Agriculture 2014. ESA Working Paper No. 14-02. Rome, FAO.

- Sarah K. Lowder, Jakob Skoet, Terri Raney. The Number, Size, and Distribution of Farms, Smallholder Farms, and Family Farms Worldwide, World Development, Volume 87, 2016, Pages 16-29.

Hope it helps.

Best regards

I have also been questioning where this figure comes from also.

Without any direct source to confirm otherwise it seems the data may have been extrapolated from regional analyses and/or confused with population data rather than actual yield.

I found this paper, which seems to try to address the issue where they note that the 70% figure has little empirical basis and set out an analysis.

Small scale farms producing 30% of food on 24% of land claimed here. While this is not as alluring as other claims, I think it still makes a strong case for the effectiveness of small scale production.

I also found examples where data on 'family farms' is presented, but this seems less meaningful 'family farm' needn't equal small scale.

thank you vey much. i'm wondering if the 70% figure was meant to say 70% of people are fed from smallholder farms. do you think? thank you

There are more than 570 million small farms in the world (Lowder, et al., 2016) and about 500 million smallholder farmers depend on agricultural businesses in developing countries (World Bank, 2017; Smith et al.,2020). Worldwide, 84% are farms are smaller than 2 ha (approximately 475 million farms), but only operate 12% of all farmland. Most of the farms are found in upper- and lower-middle income countries with China and India making up roughly 60% of the farms in the world.(Fanzo,2018). The most commonly used thresholds for designating small farms are one and two hectares (Lowder, et al., 2014; 2016), while >98% of the farms are characterized as family farms and they operate at least 58% of the agricultural land. It is estimated that family farms in sub-Saharan Africa and Asia that meet 36–114% of the domestic caloric requirements Graeub et al.,2016). According to FAO (2014) classification, smallholders can be small-scale farmers, pastoralists, forest keepers, fishers who manage areas varying from <1 to 10 ha. Of those farms that are considered small-scale, approximately ~3.4 billion people currently live and work on those farms across the developing world as per United Nations report ESA/P/WP/241(2014). Nearly 85% of the farmland in sub-Saharan Africa and Asia is managed by smallholders working on up to 10 ha and these smallholders provide 80% of the food supply in Asian and sub- Saharan Africa. Whereas in Australia, Latin America and North America, food coming from rural places are from medium to large holdings (Herrero,et al.,2017). The smallholders constitute the backbone of global agriculture and therefore extremely important for ensuring food and nutritional security of ever increasing population which is likely to be 9.5-10 billion by 2050. Nevertheless, this is the most vulnerable group in terms of adaptation and mitigations to the increased frequencies and intensities of the adverse affects due to climate change and therefore jeopardizing the agricultural sustainability as a whole. In this regard, Platteau et al (2017) reported that less than 30% of smallholders take out any form of insurance(Smith et al.,2020) as an adaptation measure. Roughly 2.5 billion of rural population derive their livelihoods from agriculture. Paradoxically, most of people suffering from hunger and malnutrition are in rural areas and only 20% are in city slums. According to FAO, 50% of them are small peasants, 20% are landless, 10% are nomadic herdsmen or small fishermen, and 20% live in city slums(Kanianska,2016).

A UN report has been cited by Robert Evans (2011), as “ for most crops of the optimal farm is small in scale and that is at this level that most gain in terms of both sustainable productivity increases and rural poverty reduction can be achieved” stressing towards a major shift in small scale farming if endemic food crisis are to be overcome and production boosted to support the global population. In India, National sample Survey data has been cited in favor of small farms having higher productivity (Chand et al., 2011). It is the fact that small holdings in Indian agriculture still exhibit a higher productivity than large holdings (Kadapatti and Bagalkoti, 2014). The UN Special Rapporteur on the Right to Food, Olivier De Schutter emphasized that small-scale ecological farming is very productive and promising even better with the fact that small and diversified farms in comparison to larger ones exhibited greater productivity per unit area, and named this phenomenon as the ’paradox of the scale’ or the ‘inverse farm size-productivity relationship’ (De Schutter, 2011). Moreover, the diminishing traditional agro-biodiversity along with its associated indigenous and local knowledge which in turn driven by agricultural intensification. A process involving high input in terms of assured irrigation, fertilizers, quality seed of only a few genetically uniform high yielding varieties (Mono-culturing). Agro-biodiversity loss has increased at an alarming rate which may jeopardizing food security and biodiversity loss result reduces productivity, while economically, biodiversity loss increases yield variability with possibly decreased average yield (Lanz et al., 2018). Moreover, Herrero et al., (2017) demonstrated that globally, diversity of agricultural and nutrient production diminishes as farm size increases and smaller landscapes with more agro-biodiversity provide more nutrients into the global food supply, particularly micronutrients (53 to 81% of micronutrients in the food supply).

Please view this source to have some of your questions answered (Practical Action, 2011. Hunger, food and agriculture: responding to the ongoing challenges)

With best regards

I am familiar with the UNEP Green Economy report of 2011.

The report describes research claiming the following (see attached screen capture, the full report is to heavy to upload here but it is easy to find).

Swaminathan, M.S. 2009. Need for an Ever-green revolution. Keynote speech at the Second World Seed Conference. Responding to the challenges of a changing world: the role of new plant varieties and high quality seed in agriculture, 8-10 September 2009, FAO, Rome.

FAO(2011) STRENGTHENING SEED SYSTEMS: GAP ANALYSIS OF THE SEED SECTOR. Item 3.3 of the Provisional Agenda COMMISSION ON GENETIC RESOURCES FOR FOOD AND AGRICULTURE Thirteenth Regular Session

Rome, 18 – 22 July 2011. CGRFA-13/11/Inf.13 May 2011

I have just found this source:

seems plausible.

Thank you Johannes Balas for this interesting information. The main issue is how we define small farmers, family farmers, smallholder farmers...