For the first time in 10,000 years, farming is not the dominating industry

The international reporting on the KILM report mostly rehashed the original Reuters report, and failed to emphasize the most interesting tidbit: that of the three economic activities – farming, industry, and services – farming is no longer the largest global employer.

Living in a post-industrial society – perhaps even post-information society – it’s easy to lose perspective sometimes. Today’s wire from Reuters about the ILO’s Key Indicators of the Labour Market (KILM) was widely reported, including the New York Times coverage where I first read it.

The business and mainstream press focus on our own economic competitiveness vis-a-vis China and Europe. E.g. NYT opens their reporting with “American workers are the world’s most productive, followed by the Irish, though productivity is rising fastest in China and much of the rest of Asia, according to the International Labor Organization.” Wall Street Journal leads with “U.S. workers continue to lead the world in productivity, though many East Asian economies are quickly advancing, according to a new report by the International Labour Organization.

They go on to note various interesting details, for example if you take into account the number of hours worked, the US falls to second place after Norway (Americans are noted for working many hours). More significantly, the productivity head start that the west has over China and other East Asian economies is shrinking; output per worker grew to one-fifth that of an western economy industrialized worker, from one-eighth ten years ago, a dramatic improvement. The economic growth in Asia led to the number of working poor falling by half, though in south-Saharan Africa it is getting worse. World unemployment fell from 6.4% to 6.3%.

By and large the report shows some serious progress in the world economy, coupled with some regional setbacks. But there was a small item about farming that prompted me to dig up the actual KILM report and look through it.

Here’s what the report writes (in Box 4b on page 6 of KILM04):

In recent years agriculture has lost its place as the main sector of employment and has been replaced by the services sector, which in 2006 constituted 42.0 per cent of world employment compared to 36.1 per cent for agriculture. As for the industry sector, it represented 21.9 per cent of total employment, which is almost unchanged from ten years ago. Although textbook theory suggests that economic development entails a structural transformation with a shift away from agriculture to the industry sector, this no longer seems to be reflected in reality. Instead of moving into high-productivity jobs in the industry sector, people are moving directly into the services sector, which consists of both high- and low-productivity jobs.

Therefore, it is unclear if the sectoral shift goes hand in hand with productivity increases and thereby a better utilization of the workforce. Agriculture is still the main sector of employment in the world’s poorest regions. Two-thirds of workers in sub-Saharan Africa and almost half of workers in South Asia and South-East Asia & the Pacific are in agriculture.

So, firstly, modernization of large economies is largely bypassing industrialization and going straight to the service sectors – in the western economies the service sector was about two-thirds of the economy, and has grown further (to 71.2%). But the so many workers are moving straight to service industries that their roles have changed. Worldwide, in 1996 agriculture employed 42%, industry 21%, and services 37%. In 2006, the numbers are 36%, 22%, and 42%. So in the period, the service sector has overtaken farming on a global scale.

To me this stuck out as the news of the day. This is a huge milestone. In the west we’re accustomed to the farming sector being 4-6% or so, but that certainly has not been true in most of the word. You might think the industrial revolution was a long time ago, but the reality is that farming has remained the center of the overall human condition. Until sometime in these past few years, that is.

And thus passes a tremendous milestone in the history of our species. Farming, invented around 8000 BC, quickly dominated human activity and has continued to for some 10,000 years. And we even find that the agriculture->industry->services transition doesn’t hold up globally. The industry segment simply isn’t big enough, so increasingly workers go directly from farming to services.

So despite what they tell you at coffee tomorrow about the adoption rates your marketing people expect for your product, you might want to point out the longest paradigm shift of all time; farming handed over the crown to the service sector only just about now.


  1. bob cobb

    Here’s the important info, from the actual report:

    You’ll note, from this article:

    “Caution should be used, however, where the information refers only to employees or only to urban areas. For some years in certain countries, the sectoral information relates only to urban areas, so that little or no agricultural work is recorded.”

    Also, there is no data culled for the vast majority of African nations, where the sector of choice would be agriculture.

    So, to sum it up – your blog about the rise of services vs. agriculture could only be considered partially correct, at best.

    [PSM] Hm, I thought Africa was accounted for; thanks for posting, I’ll dig back into this later today to check that out.

    [PSM] Well I’ve revisited it, and I don’t understand your point. The sentence you are quoting is specifically about using fine-grained information, which is not what I’m looking at. The sentence immediately prior to the one you quote says “Where information is reported for total employment or civilian employment for the entire country, comparability across countries is reasonable for the employment-by-sector indicator, because of the similarity in coverage.” And in any case I’m not using the data at all, I’m actually quoting from the report per se when it says “In recent years agriculture has lost its place as the main sector of employment and has been replaced by the services sector”. And I don’t know where your “vast majority of African nations” statement comes from. The report lacks several years of detailed data for most sub-Saharan nations, which is by no stretch of the imagination the “vast majority” of the African population; and in any case we’re just looking at endpoints so we don’t need the full time series; and even if we didn’t have those, the numbers in the report indicate a minimal movement in reduction on dependence on farming in sub-Saharan Africa, so it won’t matter to the global averages. Either I’m not understanding your point, or you didn’t read the report, or you need to bone up on your statistics.


  2. uchikoma

    indeed this is the news of the day. I wonder if anybody is watching?

    [PSM] I *often* wonder about that when I watch mainstream news coverage. :-}


  3. There’s a simple reason for the direct shift to service – industry can be exported, whereas services cannot. Industry is a net total, and it doesn’t matter whether your steel comes from Pennsylvania or Japan. Service, however, by needs must be local; your waiter can’t actually be on the other side of the city.

    By the way, “if you [liked] this posting.” Not many people are licking your blog.

    [PSM] True, but I think the significance of that is in conjunction with productivity. Industrial productivity quickly becomes so high there simply isn’t enough demand to provide enough employment. And yes, there where many silly spelling errors!


  4. Learn to spell.

    [PSM] I was tired! Since fixed … more or less …


  5. Brady

    Yeah, I licked this article. Come clean my screen.

    [PSM] Hey! I didn’t say you *had* to lick it …


  6. ‘If you licked this posting, then please click here to Digg it. Thx.”

    Sorry, I don’t even lick stamps, anymore.


  7. Danny Greene

    Obviously, you are not a farmer !


  8. It is indeed a milestone that farming no longer is our primary sector. But your conclusion on the industrial sector is incorrect. The shift to services has largely been fueled by the industrial sector and its huge productivity engine. The outsourcing trends in industry have created much of the services sector. The enterprise software industry is a good example. Also consider the transfer of 10s of thousands of employees from GM to EDS in the late 80s. Nothing changed except the relative employment in industry and services.

    The massive growth in the value of output of the US industrial sector, and the dependence of many other service jobs on that sector, belies even the concept of a post industrial economy. The knowledge economy began several hundred years ago and was called the industrial revolution.

    [PSM] John, thanks for your comments, and you make a good point. But it’s to some extent a semantic issue. I wasn’t so much commenting on what was in the end driving the service industries (e.g. paying the bills), but the actual employment per se. Individuals are not transitioning through the industrial sector as much as one might have expected.


  9. Andrew Oliver

    I remarked just yesterday that as a farmer I felt just like a Celtic missionary. I know I am part of a dying breed and too stupid to change religions.

    [PSM] Lol! Andrew, those of us who actually eat as part of our lives (which means most of us), we deeply appreciate you’re efforts!


  10. Excellent article. It really is astonishing. Note: you’ve got a typo right underneath the quote for the report: “ecnomies”.

    [PSM] I had a *lot* of spelling errors; my brain was fried. Thanks for pointing this one out.


  11. stuart

    Do lots of people lick postings?

    [PSM] It’s the latest thing in social networking! Immune system compatibility testing.


  12. The day after you posted this a University of Illinois site, Farmgate, had this to say:

    “The last half of the 20th Century was momentous in terms of US agricultural productivity. Milk production climbed from 5,314 to 18,201 pounds per cow, corn yield went up 39 bushels per acre and the typical farmer was 12 times more productive in 2000 than he was in 1950.”

    I just got back from the Farm Progress Show, one of the top farm shows in the United States. It’d been a few years since I attended, and I was impressed (like I was last time) at the scale of the equipment. I don’t lament it, but just as the world has handed off the agriculture baton to services, the scale of modern agriculture makes farm boys of the last century realize they’ve lost touch. Dropped the baton, to push the metaphor.

    [PSM] It would have been fun to check out that show. The productivity of modern agriculture is indeed amazing.


  13. Ugh, I was tired when I posted this. Just updated it with various spelling, grammar, and clarity fixes.


  14. Thanks Peter

    Nice insight. I’ve been tracing leadership patterns back in time. Seems the old hunter-gatherers were far and away more collaborative within groups despite their need to hunt and kill. Conversely, it was agrarian societies within which innovations such as chariots, beasts of burden and war, and slavery emerged. So are we heading for another cultural turn in which service values help promote a wider sort of social exchange than that of weaponry and status-goods? Don’t think so, somehow.

    Not sure where we might be going … (or, as I get challenged to say: ”who is ‘we’ you go on about”. OK, I’m not sure where I’m going.)

    Best wishes

    [PSM] Thanks for your comments Tudor. Yes indeed, farming is what lead to inventions like large-scale warfare, class society, torture, human sacrifice, slavery, authoritarian politics, toxic waste, deforestation, and all manner of other niceties of civilization! (With apologies to farmers reading this – it’s not your fault!) It’s an interesting but somewhat larger topic whether new technologies for social networking will fundamentally change any of this; the basic problem with the agrarian shift was that people and assets became stationary, and that’s not really looking to change.


  15. Ooof, new visitor record. Slashdot picked this up yesterday at 1pm and instapundit at 9pm. Welcome new readers! Don’t forget to digg it.


  16. Great post, good info.
    Ignore the typo police, they need to get laid.

    [PSM] Lol! Well I don’t mind them that much. I’m kind of anal about that as well, I’m really bothered when *other* people spell carelessly. But I swear, I was literally falling asleep when I posted this on Monday night. With the wonder of blogs and user feedback, the typos should be fixed now!


  17. Forbes


    John Layden above (3:03) makes an interesting argument regarding the robustness of the data vis-a-vis the transfer of thousands of GM employees to EDS during the ’80s that (apparently) suddenly turned (automotive) industrial employees into (info tech) service employees, and the premise that such industrial outsourcing is responsible for the expansion of the service sector. (I might point out that GM had over a million employees at the time. A reclassification of ~3% of its employees is hardly noticeable on a national scale, much less a distortion of international data.)

    To this, I say, so what. Without doing a disservice to Mr. Layden’s point, it relies, a bit I think, on the argument that the only business model is the fully-integrated vertical business enterprise.

    Just as farm productivity rose (and consequently farm employment declined), first due to mechanization (industrial goods), and then by means of hybridization and other “knowledge goods”, we don’t lament the decline in farm employment as some sort of numbers game due to the outsourcing of the goods and services regularly utilized in modern farming.

    GM outsourced its info tech operations and personnel because they discovered they weren’t just in the automotive manufacturing business–and therefore made a business decision regarding “do it yourself” or buy it from a supplier.

    That there might be some lumpiness in the periodic data collection and classification effort should not detract from observing the trend, as you have pointed out regarding agriculture/industrial/service employment.


    [PSM] Thx for your comment. I agree.


  18. Where did you get it that Norwegians work more hours than Americans? I lived in Norway and that was against all my impression. Office and opening hours in Norway tend to be very short (technical staff in state sector leaving at 3:30 p.m. and shops in the downtown closing at 4 o’clock is not uncommon). I’ve digged in to the report and, indeed, if anything Norway has the smallest percentage of workers with 40+ hours week of the countries listed (US workers were on the long hours side, indeed).

    So where did you get that data on Norway vs. US?

    [PSM] Sorry if I was unclear: Norway has the highest productivity in the world when measured on a per-hour basis. It’s on page 12 in KILM 18: “… Norway showed the highest labour productivity level measured as value added per hour worked (US$37.99), followed by the United States (US$35.63) and France (US$35.08).”


  19. paradoctor

    You speak of a “post-information” society. Now, we all know that a “post-industrial” society is based upon post-industry; that is, offshoring and unemployment. Please tell me what “post-information” is. I suspect that it is another name for fraud and delusion. In that sense, the Iraq war is a post-informational war.

    The trouble with post-industry and post-information is that they tend to lead to post-society.

    [PSM] Well, I meant it a little tongue-in-cheek; after all, the core sentiment of my posting was that the world hasn’t really changed as quickly as we thought. We’ve been busy dreaming up interpretations of how quickly the world is moving, and gee folks, we only just got off the farm. But anyway, I do in fact have a perspective on what constitutes post-information society, and that is the trend for information to be free (open source; wikipedia; blogging) and that it’s networks and interaction (structured data, as it were) that constitute value. But that’s a whole other posting …


  20. John Morales

    Thanks! Good point, good post.

    [PSM] Thx!


  21. P.M.Lawrence

    Unless you’re using the term “farming” very loosely to include pastoral activities and mixed lifestyles, it simply isn’t true that farming has been dominant for 10.000 years or so. It has probably been dominant for around 2.000 years, measured by head of total world population. The thing is, although many societies started practising it from around 5.000 years ago, many of those didn’t do so exclusively until very recently when resource or outside constraints hit. If you read Spenser on Ireland, you will see that only a little over four centuries ago the Irish were still practising this sort of mixed activity lifestyle.

    [PSM] Well, the details around the early rise of agriculture and commensurate early population levels are still quite hotly debated. From the sources I looked at, it seemed pretty conservative to assert that farming (planting crops and domesticating animals) dominated the global labor force by 8K BCE. Yes, there were peripheral geographies where this did not hold true (indeed, there still are), but that doesn’t much impact the global stats. Even in the event that that’s not true, the real significance of the transition is after all that the bulk of human labor is no longer spent on caloric production. As an aside, one might also note that the “modern” notion of a nomadic-pastoral lifestyle presupposes the domestication of the horse, which is a much more recent event than the Neolithic revolution.


  22. Cotton

    We are in a secular bull market in commodities. This is a nice frosting layer on the cake.


  23. The core observation on farming is important at a global level for what it means to standards of living. The US economy passed this point (50% of population in farming) in about 1910.

    My original point was that we are not by-passing the industrial economy phase and going directly to a services economy. We are redefining the boundary between industrial companies and their supporting services organizations. The huge increase in productivity in agriculture over the past two centuries resulted in increased output and, since the peak in 1910, decreased employment. But many farming jobs are now in manufacturing at Deere. Exactly the same thing has happened in the industrial economy. The value of industrial output continues to grow even as direct employment declines from its peak in 1979. The employment picture is clouded by the fact that some jobs were eliminated by improved efficiency, and some were transfered to services companies. I’ve seen no analysis capable of isolating the two. The key is that these service jobs are still dependent on the industrial economy they support. In any service industry you need to ask the question “…in service to whom?” The expansion of the knowledge economy began and is still largely fueled by the needs of the industrial companies. This industrial investment has driven a global explosion of industrial productivity unlike anything we’ve ever seen.


  24. cesium

    1) I think you need to be careful about whether or not people are moving directly from agriculture to services, or whether they are moving from agriculture into manufacturing and thence to services. Certainly the data provided by the KILM paragraphs you quote do not shed any light on this, and I would want to see the data before I believed KILM’s conclusion. The world certainly has an appearence of Chinese workers moving from the farm into manufacturing as US and European workers move from manufacturing into services.

    2) A stronger distinction needs to be made between the knowledge economy portion of services and the labor economy portion of services. Certainly the labor portion is local — your waiter does need to be at your table. But the knowledge portion is more global than manufacturing. Contributors to Wikipedia may be anywhere in the world. Updates to the Linux OS can come from anywhere.

    3) Your article is not a good example of how slowly things are changing. We were on the farm for 10,000 years. Nearly everyone in the world farmed or herded or hunted and gathered. 100 years ago, nearly everyone in the world was still on the farm. Half of the United States had gotten off the farm, but that was small relative to the world as a whole. All of a sudden in the past 100 years, half the population has moved off the farm. That’s a horrifically fast rate of change.

    I think if the services numbers were split into labor and knowledge segments, further insights into the rate of change and the potential for future rapid change would become more evident.


  25. Not really 10,000 years. For quite some time after the appearance of agriculture in Mesopotamia most people around the world remained still hunters and gatherers 🙂 For example in Eastern US (a spot were agriculture appeared independently, rather than being imported from other places) plants were not domesticated until 2500BC.

    [PSM] My understanding is that what you’re saying is not true, because of population densities. Agriculture supports such a dramatically higher density – and the order of 100 times more – that the agriculture-based societies rapidly outnumbered the others. See for example page 259 in “The evolution of human society”.



  1. buzz
  2. For the first time in 10,000 years, farming is not the dominating industry « Terryorisms
  3. Amused Cynicism
  4. International Relations » The relegation of Farming
  5. The Conjecturer » News Brief, Some Loud Thunder Edition
  6. A Second Hand Conjecture » News Brief, Some Loud Thunder Edition
  7. We live in a truly unique time « Lunchtime!
  8. Below The Beltway » Blog Archive » A 10,000 Year Transformation
  9. Framtidstanken - Accelererande förändring i en global ekonomi » Blog Archive » Jordbrukssamhällets tid officiellt förbi efter 10000 år
  10. » Blog Archive » The world gets a new job
  11. NuVatsia » Go figure
  12. Farming no longer #1 « Godpigeon’s Weblog
  13. Today’s Links « Wintry Smile
  14. For the first time in 10,000 years, farming is not the dominating industry « a shade left of happy
  15. Pickerhead :: Pickings from the Webvine ::september 6, 2007
  16. Uncle Bear » Blog Archive » Random News Table
  17. Yankee Wombat | An American in Oz
  18. Smith and Engels » Blog Archive » Capitalism Oversees Monumental Transition
  19. Trumpai: pagaliau nebe artojai : než
  21. Pebre » Blog Archive » Por primera vez en 10000 años, la agricultura no es el sector económico dominante en el planeta
  22. Ashutosh Nilkanth’s Blog » For the first time in 10,000 years …
  23. links for 2007-09-09 « Romulo Lopez Cordero
  24. 2007 September 14 — EXCELER8ion | People ARE The Social Media
  25. Progress Daily » Blog Archive » Service-ization
  26. El mundo urbano « Marketwatcher’s Weblog
  27. purple motes » COB-16: bureaucracy is more important than farming
  28. Yankee Wombat | An American in Oz
  29. FamilyGreenberg.Com - Farmers - A Dying Breed. Finally.
  30. Fabrikerna är rivna och arbetarna dansar på dess gravar « Skiften
  31. On Agriculture | My Blog

Leave a Reply to tamgo Cancel reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: