For the last 200 years or so, life has been getting better for the average human in the world.
What about for the 300,000+ years before that?
In order to answer this, one of the hardest things we need to do is get some sense of pre-agriculture quality of life.
- Agriculture is estimated to have started drastically changing lifestyles, and leading to what we tend to think of as "civilization," around 10,000 BCE. Agriculture roughly means living off of domesticated plants and livestock, allowing a large population to live in one area indefinitely rather than needing to move as it runs low on resources.
- So most years of human history were pre-agriculture (and thus pre-"civilization").
- The terms "hunter-gatherer" and "forager" are commonly used to refer to societies that came before (or simply never took up) agriculture.
This appears to be a topic where there is a lot of room for controversy and confusion.
- Many people seem to endorse a "pre-agriculture Eden" hypothesis: that the pre-agriculture world was a sort of paradise, or at least better than life in rich countries today. There are logical reasons that this might be the case; below, I'll lay some of those out and give some quotes from Wikipedia that convey the "pre-agriculture Eden" vibe.
- But there's also a case to be made that the world before agriculture was a world of starvation, disease and violence - that the human story is one of continuous, consistent beneficial progress, and that the pre-agriculture world was the lowest (because the earliest) point on it.
My tentative position is that neither of these is quite right. I think the pre-agriculture world was noticeably worse than today's world (at least in developed countries), but probably some amount better than the world that immediately followed agriculture.
This post will focus on the former: the comparison between the pre-agriculture world and today's world, or whether the "pre-agriculture Eden" hypothesis is right. I'll argue that today's best evidence suggests that today's developed world has significantly better quality of life than the pre-agriculture world. By doing so, I'll also lay the groundwork for a future post about what happened to quality of life in between the pre-agriculture world and today.
- Give more detail on the basic pre-agriculture Eden hypothesis.
- Go through each of the dimensions on which I tried to compare pre-agriculture and current quality of life. These are summarized by the table below, which uses the same structure as for my previous post:
|Pre-agriculture vs. today's developed world
|Mostly assessed via hunger and health - see below.
|Pre-agriculture height looks very low by today's standards, suggesting malnutrition.
|Pre-agriculture infant and child mortality look extremely high by today's standards (20%+ before age 1, 35%+ before age 10 (today's high-income countries are under 1% before age 10). Post-childhood life expectancy also looks a lot worse than today's.
|Pre-agriculture deaths from violence look more common, compared to today's developed world.
|Unknown - while there are some claims floating around about strong mental health for hunter-gatherers/foragers, I haven't seen anything that looks like solid evidence about this, and similar claims re: gender relations don't hold up to scrutiny.
|Substance abuse and addiction
|Presumably not an issue pre-agriculture (though this isn't 100% clear).
|Hard to compare, but pre-agriculture gender relations seem bad.
|Treatment of children
|Education and literacy
|Literacy would be higher in today's world (though it isn't clear whether this matters for quality of life)
|Friendship and community
|Romantic relationship quality
|Meaning and fulfillment
The pre-agriculture Eden hypothesis
- Hunter-gatherers are not "poor," or at least, they "are mostly well-fed, rather than starving," and have more leisure time than most people today.
- Hunter-gatherers "tend to have an egalitarian social ethos," without permanent leaders ("hunter-gatherers do not have permanent leaders; instead, the person taking the initiative at any one time depends on the task being performed").
- (Covered previously) Hunter-gatherers have egalitarian gender relations specifically, "with women roughly as influential and powerful as men."
- I've seen it claimed that "coronary heart disease, obesity, [and a number of other diseases] ... are rare or virtually absent in hunter–gatherers and other non-westernized populations." This is usually given as an argument that hunter-gatherers have excellent diets that we should emulate.
- I've seen more occasional (and as far as I can tell, very thinly cited) claims that "Hunter-gatherers seem to possess exceptional mental health" / "depression is a 'disease of modernity.'"
Why might all of this be? The basic idea would be that:
- For most of human history, humans lived in small, "nomadic" bands (more on this in a future post): constantly moving from one location to another, since any given location had limited food supply. People who did well in this setting reproduced and people who did poorly did not, so we (the descendants of many people who did well) are well adapted to this lifestyle.
- But about 10,000 years ago, much of the world transitioned to agriculture, which meant that instead of moving from place to place, we were able to consistently produce large amounts of food by staying put. This led to an explosion in population, and a division of labor: farmers could produce enough food for everyone, while other people specialized in other things such as religion, politics, and war.
- 10,000 years isn't a ton of time from the standpoint of natural selection, so we're still adapted to the original environment, and we're "out of place" in a more modern lifestyle.
To put some of my cards on the table early, I think this reasoning could be right when it comes to some problems in the modern world, but I don't tend to believe it strongly by default.
- I don't think that "adapting to" an environment should be associated with "thriving" in it - especially not if "thriving" is supposed to include things like egalitarianism. In my view, "adapting to" an environment simply means becoming good at competing with others to reproduce in that environment - you could be fully "adapted" to your environment and still frequently be hungry, diseased, violent, hierarchical, sexist, and many other nasty things that we regularly see from animals in their natural environments.
- Additionally, there are many diverse lifestyles in the modern world. So any problem that seems to exist ~everywhere in modern civilization seems to me like it's most likely (by default) to be "a risk of being human."
That said, I don't think either of these points is absolute. There are some ways in which nearly all modern societies differ from forager/hunter-gatherer societies, and some of these might be causing novel problems that didn't exist in our ancestral environment. So I consider the "pre-agriculture Eden" hypothesis plausible enough to be interesting and important, and I'd like to know whether the facts support it.
Evidence on different dimensions of quality of life
Below, I'll go through the best evidence I've found on the dimensions of quality of life from the table above.
For more complex topics, I mostly rely on previous more detailed posts I've made. Otherwise, I tend to rely by default on The Lifeways of Hunter-Gatherers (which I abbreviate as "Lifeways"), for reasons outlined here.
I discussed pre-agriculture gender relations at some length in a previous post. In brief:
- According to the best/most systematic available evidence (from observing modern non-agricultural societies), pre-agriculture gender relations seem bad. For example, most societies seem to have no possibility for female leaders, and limited or no female voice in intra-band affairs.
- There are a lot of claims to the contrary floating around, but (IMO) without good evidence. For example, the Wikipedia entry for "hunter-gatherer" gives the strong impression that nonagricultural societies have strong gender equality, as does a Google search for "hunter-gatherer gender relations." But the sources cited seem very thin and often only tangentially related to the claims; furthermore, they often seem to acknowledge significant inequality, while seemingly trying to explain it away with strange statements like "women know how to deal with physical aggression, unlike their Western counterparts." (Verbatim quote.)
I think it's somewhat common to find rosy pictures of pre-agriculture society, with thin and even contradictory citations. I think this is worth keeping in mind for the below sections (where I won't go into as much depth as I did for gender relations).
Pre-agriculture violence seems to be a hotly debated topic among anthropologists and archaeologists; the debates can get quite intricate and confusing, and I've spent more time than I hoped to trying to understand both sides and where they disagree.
My take as of now is that overall pre-agriculture violence was likely quite high by the standards of today's developed countries.
This was complex enough that I devoted a separate post entirely to my research and reasoning on this point. Here's the summary on "nomadic forager" societies (which are thought to be our best clue at what life was like in the very distant past) vs. today's world:
|Violent deaths per 100,000 people per year
|Murngin (nomadic foragers)
|Tiwi (nomadic foragers)
|!Kung (aka Ju/'hoansi) (nomadic foragers)
|World (today) - high estimate
|USA (today) - high estimate
|World (today) - low estimate
|Western Europe (today) - high estimate
|USA (today) - low estimate
|Western Europe (today) - low estimate
I'm going to examine both hunger and health, since both seem among the easiest ways to get at the question of whether pre-agriculture society had meaningfully higher "poverty" than today's in some sense.
The most relevant-seeming part of Lifeways is Table 3-5, which gives information on height, weight, and calorie consumption for 8 forager societies. My main observation (and see footnote for some other notes2) is that the height figures are strikingly low: 6 of the 7 listed averages for males are under 5'3", and 6 of the 7 listed averages for females are under 5'0". (Compare to 5'9" for US males and 5'3.5" for US females.)
The height figure seems important because height is often used as an indicator for early-childhood nutritional status,3 and seems to quite reliably increase with wealth at the aggregate societal level (see Our World in Data's page on height). Height seems particularly helpful here because it is relatively easy to measure in a culture-agnostic way and can even be estimated from archaeological remains.
What I've been able to find of other evidence (including archaeological evidence) about height suggests that the pre-agriculture period had average heights a bit taller than the figures above, but still short by modern standards, though this evidence seems quite limited (details in footnote).4
My bottom line is: the evidence suggests that pre-agriculture people had noticeably shorter heights than modern people, which suggests to me that their early-childhood nutrition was worse.
As for Wikipedia's claim that "Contrary to common misconception, hunter-gatherers are mostly well-fed," those who have read my previous piece on Wikipedia and hunter-gatherers might be able to guess what's coming next.
- The citation for that statement appears to be an entire textbook (no page number given), which I found a copy of here (the link unfortunately seems to have broken since then).
- The vast majority of the textbook doesn't seem to be relevant to this topic at all.
- From skimming the table of contents, my best guess at the part being cited is on page 328: "The notion that hunters and gatherers live on the brink of starvation is a popular misconception; numerous studies have shown that hunters and gatherers are generally well nourished." No citations are given.
It seems to me that the best proxy for health, in terms of having very-long-run data, is early-in-life mortality (before age 1, before age 5, before age 15). I've found a number of collections of data on this, and nothing else detailed regarding health for prehistoric or foraging populations (other than one analysis that looks at full life expectancy; I will discuss this later on).
Table 7-7 in Lifeways lists a number of figures for deaths before ages 1 and 15, based on modern foraging societies. Taking a crude average yields 20% mortality before the age of 1, 35% before the age of 15.
Other sources I've consulted (including archaeological sources) give an even grimmer picture, in some cases 50%+ mortality before the age of 15 (details in footnote).5
These are enormous early-in-life mortality rates compared to the modern world, where no country has a before-age-15 mortality rate over 15%, and high-income countries appear to be universally below 1% (source).
What about life expectancy after reaching age 10?
What I've found also suggests that pre-agriculture life expectancy was lower than today's at other ages, too - it isn't just a matter of early-in-life mortality.
Gurven and Kaplan 2007 (the only paper I've found that estimates pre-agriculture life expectancy, as opposed to early-in-life mortality) observes that its modeled life-expectancy-by-age curves are similar for modern foraging societies and 1751-1759 Sweden (Figure 3):
(Note also how much worse the estimate of prehistoric life expectancy looks at every age, although Gurven and Kaplan question this data.6)
As noted at Our World in Data, it appears that life expectancy conditional on surviving to age 10 has improved greatly in Sweden and other countries since ~1800 (before which point it appears to have been pretty flat).
Also see these charts, showing life expectancy at every age improving significantly in England and Wales since ~1800.
See footnote for one more data source with a similar bottom line.7
Bottom line: life expectancy looks to have been a lot worse pre-agriculture than today. I don't think violent deaths account for enough death (see previous section) to play a big role in this; disease and other health factors seem most likely.
What about diseases of affluence?
Diseases of affluence ... is a term sometimes given to selected diseases and other health conditions which are commonly thought to be a result of increasing wealth in a society ... Examples of diseases of affluence include mostly chronic non-communicable diseases (NCDs) and other physical health conditions for which personal lifestyles and societal conditions associated with economic development are believed to be an important risk factor — such as type 2 diabetes, asthma, coronary heart disease, cerebrovascular disease, peripheral vascular disease, obesity, hypertension, cancer, alcoholism, gout, and some types of allergy.
I think it's plausible that the pre-agriculture world had less of these "diseases of affluence" than the modern world (especially obesity and conditions connected to obesity, due to the seemingly much greater access to food).
I don't think it's slam-dunk clear for some of these, such as cancer and heart disease. I've dug into primary sources a little bit, and not-too-surprisingly, data quality and rigor seems to often be low. In particular, I quite distrust claims like "Someone spent __ years in ___ society and observed no cases of ____." Modern foraging societies seem to be quite small, and diagnosis could be far from straightforward.
I haven't dug in heavily on this (though I may in the future), because:
- My initial scans have made it look like it would be a lot of work to follow often-circuitous trails of references to often-hard-to-find sources.
- Even if it did turn out that "diseases of affluence" were extremely rare pre-agriculture, this wouldn't tip me into thinking health was better overall, pre-agriculture. When wondering whether undernutrition and "diseases of poverty" are worse than obesity and "diseases of affluence," I think a good default is to prefer the condition with less premature death.
Mental health and wellbeing
I haven't found anything that looks like systematic data on pre-agriculture mental health or subjective wellbeing. There are some suggestive Google results, but as in other cases, these don't seem well-cited. For example, as of this writing, Google's "answer box" reads:
But Thomas 2006 is this not-very-systematic-looking source.
I won't go into this topic more, because having gone through the above topics, I don't find the basic plausibility of "reliable data shows better-than-modern mental health among foraging societies" high enough to be worth a deep dive.
Update: a commenter linked to a study reporting high happiness for one hunter-gatherer society (the Hadza). My thoughts here.
Leisure and equality
I haven't gone into depth on claims that pre-agriculture societies had more leisure, and lower inequality, compared to today's.
Reasons for this:
The claims seem disputed. For example, here are excerpts on both topics from the first chapter of Lifeways:
How much do hunter-gatherers work, and why? Reexaminations of Ju/’hoansi and Australian work effort do not support Sahlins’s claim [of very low work hours] . Kristen Hawkes and James O’Connell (1981) found a major discrepancy between the Paraguayan Ache’s nearly seventy-hour work week and the Ju/’hoansi’s reportedly twelve- to nineteen-hour week. The discrepancy, they discovered, lay in Lee’s definition of work. Lee counted as work only the time spent in the bush searching for and procuring food, not the labor needed to process food resources in camp. Add in the time it takes to manufacture and maintain tools, carry water, care for children, process nuts and game, gather firewood, and clean habitations, and the Ju/’hoansi work well over a forty-hour week (Lee 1984; Isaac 1990; Kaplan 2000). In addition, one of Sahlins’s Australian datasets was generated from a foraging experiment of only a few days’ duration, performed by nine adults with no dependents. There was little incentive for these adults to forage much (and apparently they were none too keen on participating – see Altman [1984, 1987]; Bird-David [1992b]) ...
Others have found that the alleged egalitarian relations of hunter-gatherers are pervaded by inequality, if only between the young and the old and between men and women (Woodburn 1980; Hayden et al. 1986; Leacock 1978; see Chapters 8 and 9). Food is not shared equally, and women may eat less meat than do men (Speth 1990, 2010; Walker and Hewlett 1990). Archaeologists find more and more evidence of nonegalitarian hunter-gatherers in a variety of different environments (Price and Brown 1985b; Arnold 1996a; Ames 2001), most of whom lived under high population densities and stored food on a large scale. Put simply, we cannot equate foraging with egalitarianism.
I'm skeptical that anthropologists can get highly reliable reads on the degree to which foraging societies have high leisure, or low inequality, in a deep sense. I imagine that if an anthropologist (from e.g. another planet) visited modern society, they might conclude that we have high leisure, or low inequality, based on things like:
- Having a tough time disentangling "work" from "leisure." For example, a lot of modern jobs are office jobs, and a lot of on-the-job hours are spent doing what might look like pleasant socializing. (Similarly, foragers "socializing" may be internally conceiving this as necessary work rather than fun - it seems like it could be quite hard to draw this line.)
- Being confused by social norms encouraging people to downplay real inequalities. For example, I've seen a fair number of references to the fact that people in foraging societies will sometimes mock a successful hunter to "cut them down to size" and enforce equality. But if this were strong evidence of low inequality, I think we'd have similar evidence from modern society from things like the Law of Jante; "humblebragging"; the fact that many powerful, wealthy people in modern times tend to dress simply and signal "authenticity"; etc.
Even if I were convinced that pre-agriculture societies had large amounts of leisure and low amounts of inequality, this wouldn't move me much toward believing they were an "Eden," given above observations about violence, hunger and health. It would be one thing if foragers were healthy, well-fed, well-resourced, and lived in conditions of high leisure and low inequality. But high leisure and low inequality seem much less appealing in the context of what looks to me best described as "poverty" with respect to health and nutrition.
Having vetted other "Eden"-like claims about pre-agriculture societies, I've developed a prior that these claims are likely to be both time-consuming to investigate and greatly exaggerated. See previous sections.
With all of that said, as I'll discuss in future posts, I do think there are signs that at least some foraging societies were noticeably more egalitarian than the societies that came after them - just not more so than today's developed world.
Next in series: Did life get better during the pre-industrial era? (Ehhhh)
Use "Feedback" if you have comments/suggestions you want me to see, or if you're up for giving some quick feedback about this post (which I greatly appreciate!) Use "Forum" if you want to discuss this post publicly on the Effective Altruism Forum.
"Contrary to common misconception, hunter-gatherers are mostly well-fed, rather than starving ...
"Hunter-gatherers tend to have an egalitarian social ethos, although settled hunter-gatherers (for example, those inhabiting the Northwest Coast of North America) are an exception to this rule. Nearly all African hunter-gatherers are egalitarian, with women roughly as influential and powerful as men. For example, the San people or 'Bushmen' of southern Africa have social customs that strongly discourage hoarding and displays of authority, and encourage economic equality via sharing of food and material goods ...
"Anthropologists maintain that hunter-gatherers do not have permanent leaders; instead, the person taking the initiative at any one time depends on the task being performed. In addition to social and economic equality in hunter-gatherer societies, there is often, though not always, sexual parity as well ...
"At the 1966 'Man the Hunter' conference, anthropologists Richard Borshay Lee and Irven DeVore suggested that egalitarianism was one of several central characteristics of nomadic hunting and gathering societies because mobility requires minimization of material possessions throughout a population. Therefore, no surplus of resources can be accumulated by any single member ...
"At the same conference, Marshall Sahlins presented a paper entitled, 'Notes on the Original Affluent Society' ... According to Sahlins, ethnographic data indicated that hunter-gatherers worked far fewer hours and enjoyed more leisure than typical members of industrial society, and they still ate well. Their 'affluence' came from the idea that they were satisfied with very little in the material sense. Later, in 1996, Ross Sackett performed two distinct meta-analyses to empirically test Sahlin's view. The first of these studies looked at 102 time-allocation studies, and the second one analyzed 207 energy-expenditure studies. Sackett found that adults in foraging and horticultural societies work, on average, about 6.5 hours a day, whereas people in agricultural and industrial societies work on average 8.8 hours a day. ↩
I've spot-checked the primary sources a bit: I looked up the first three rows and the last row, and found the methodology sections. I confirmed that these are all adults. I didn't check the others. I put the table in Google Sheets and added some of my own derived figures here. In addition to the observations about height, I note:
- I don't think we can make much of raw calorie consumption estimates, since more active lifestyles could require more calories.
- The BMI figures would qualify as "underweight" for Ju/'hoansi and Anbarra females; others seem to be generally on the low side of the normal range. (These are averages, and could be consistent with having a significant percentage of individuals outside the normal range.
- Note that calculating a BMI from average height and average weight is not the same as looking at average BMI. But I played with some numbers here and it seems unlikely to be a big difference. I'd guess my BMI calculations would lead to slight overstatement of average BMI (and a slightly larger overstatement if well-fed people were both taller and higher-BMI).
Our World In Data cites a figure for the Mesolithic era (shortly before the dawn of agriculture) of 1.68m, or about 5'6". This figure comes from A Farewell to Alms, which in turn cites a study I was unable to find anywhere. It isn't explicitly stated which sexes the height figure refers to, but this chart implies to me that Our World in Data (at least) is interpreting it as a quite low height by modern standards.
Searching for recent papers on height estimates from archaeological remains, I found a 2019 paper claiming that "The earliest anatomically modern humans in Europe, present by 42-45,000 BP (5, 6), were relatively tall (mean adult male height in the Early Upper Paleolithic was ∼174 cm [about 5'8.5"]). Mean male stature then declined from the Paleolithic to the Mesolithic (∼164 cm [about 5'4.5"]) before increasing to ∼167 cm [about 5'6"] by the Bronze Age (4, 7). Subsequent changes, including the 20th century secular trend increased height to ∼170-180 cm [about 5'7" to 5'11"] (1, 4)." Its main two sources are this paper, noting a relative scarcity of data and trying to fill in gaps using mathematical analysis, and this book that looks interesting but costs $121. I haven't found many other recent analyses of this topic (and nothing contradicting these claims in any case).
Table 3.9 of A Farewell to Alms also collects height data on modern foraging societies, and the median figure in the table is about 1.65m, or about 5'5". ↩
Table 1 in Volk and Atkinson 2012 (the source used by Our World in Data) reports a mean of 26.8% mortality before the age of 1 ("infant mortality rate") and 48.8% mortality before the age of 15 ("child mortality rate"), again based on modern foraging societies.
The only sources I've been able to find pulling together estimates from archaeological data are:
- Trinkaus 1995, on Neanderthals. In Table 4 (based on a small number of sources), it gives an average of 40.5% mortality before age 1, 13.2% mortality between ages 1-5, 6.6% mortality between ages 5-10, and 7.9% mortality between ages 10-20 (which would cumulatively imply 68.2% mortality before the age of 15).
- Trinkaus 2011, which does not give mortality estimates but argues for a conclusion of "low life expectancy and demographic instability across these Late Pleistocene human groups ... [the data] provide no support for a life history advantage among early modern humans."
"Estimated mortality rates then increase dramatically for prehistoric populations, so that by age 45 they are over seven times greater than those for traditional foragers, even worse than the ratio of captive chimpanzees to foragers. Because these prehistoric populations cannot be very different genetically from the populations surveyed here, there must be systematic biases in the samples and/or in the estimation procedures at older ages where presumably endogenous senescence should dominate as primary cause of death. While excessive warfare could explain the shape of one or more of these typical prehistoric forager mortality profiles, it is improbable that these profiles represent the long-term prehistoric forager mortality profile. Such rapid mortality increase late in life would have severe consequences for our human life history evolution, particularly for senescence in humans." ↩
Trinkaus 1995 has a more detailed breakdown of mortality rates by age range for both a few modern forager societies (Table 3) and for archaeological remains (Table 4):
Here "Neonate" means "before 1 year," "Child" Is age 1-5, "Juvenile" is age 5-10, "Adolescent" is age 10-20, "Young adult" Is age 20-40, and "Old adult" is age 40+. These numbers seem extremely high for all ages under 40 by today's standards; for a comparison point see this detailed life expectancy table for the US, which implies (based on calculations I'm not showing here but that are straightforward to do) "Juvenile" mortality of 0.06%, "Adolescent" mortality of 0.6%, and "Young adult" mortality of 3.7%. (All figures are for males; female figures would be lower still.) ↩