Food Makes Babies

Daniel Quinn returned to the theme that “food makes babies” so often in his writings that it would seem he was continually dissatisfied either with the clarity of his case, or with objections people had, or both. I get it. I often return over and over to the same thorny themes, each time thinking I’ll finally nail it. The exercise is as much for improving internal clarity as anything.

Many of the comments following my coverage of Daniel Quinn’s Ishmael focused on the food–baby issue. The more recent post on The Story of B dwells on the topic as well, so I figured it would be worth dedicating a post to the matter, trying to covering all the angles.

The statement that increasing food production leads to increases in population touches a nerve for some people, which is what makes it a valuable topic to explore. For some, the statement seems to be an affront to their notion of control. It implies that humans are “no more than” animals, which takes direct aim at our most prized mythology: human supremacism—relating to Ishmael’s second dirty trick: that we “evolved from the slime”—barely tolerated by modernists, but only in a narrow technical sense.

Now, the objections are not without demonstrable legitimacy. In this post, I will start with the basics, point out key objections, then see what we can make of it.

Food Does Make People

We start with some obvious and incontrovertible facts. As for any animal, humans are built out of the food we eat. It’s where the atoms come from. Humans have many holes (including pores) out of which to lose mass, but only one mouth-hole for adding mass—in the form of tasty assemblies of atoms. Not just any atoms will do, but generally those biologically arranged into sugars, fats, and proteins. No rocks for me, please.

I cant resist pointing out that the simple act of breathing constitutes a significant mass loss mechanism—easily into kilogram-per-day territory. Not only do we take in O₂ and breathe out CO₂ (see what we did there? Kicked out a carbon atom!), but we also lose water molecules in our humid breath—made visible by condensation in cold air.

It is obvious, then, that if human population explodes—as it has—it must necessarily be accompanied by an increase in food supply going into mouths. We’ll get to causal direction later.

Mouse Experiments

Daniel Quinn effectively employed a “thought experiment” using captive mice to demonstrate the point. True: mice are not humans, but I’ll get to objections later. Holding daily food input steady results in a total mouse population fluctuating around a stable value. Providing daily increases of food (so that supply always exceeds demand) results in exponential population increase as long as suitable space avails and the regimen is maintained. Slowly decreasing daily food supply will reduce population, smoothy.

Food is therefore a perfectly effective “knob” for controlling the mouse population: none better. Food availability is a key mechanism in ecology too, for arriving at balanced populations. This knob will work on any animal or plant or fungus or microbe. Humans are among the animals of which we speak (hold the “yes, but…” for a bit longer).

Feedback

The case of increasing daily food sets up positive feedback. More mouths: then more food. More food: then more mouths. The cycle continues as population swells. Exponential, runaway growth is the hallmark of positive feedback. The enormous human population surge could not have happened without a dominant dose of positive feedback. That’s how math do.

Negative feedback can take a number of forms: food depletion, predation, and disease being the most easily identified and common in ecological contexts. A dynamic equilibrium (fluctuation around a longer-term average) involves an approximate balance of positive and negative feedback. Equilibrium doesn’t mean none of them are present: they just act in opposition to each other, leading to partial or total cancellation.

In a “pioneer” stage, positive feedback dominates over the others: not enough population to exhaust food supplies or concentrate disease, for instance. A late-stage population still has the positive feedback mechanism as fully-engaged as ever, but it is offset by the now-more-prevalent negative feedback elements: all together at once—all contributing.

A Human Thought Experiment

To those who resist the notion that increasing food production also means increasing human population, consider this. In 1950, the global population stood at about 2.5 billion people. The Green Revolution was about to explode into global agriculture, substantially increasing crop yields on the back of profligate fossil fuel inputs (for fertilizer, mechanization, energy for irrigation, transport, processing, etc.). Let’s say this tsunami of energy and technology had not arrived on the agricultural scene, and that moreover a global edict (“magically” followed) held annual food production at the 1950 level thereafter.

Would we have 8 billion people today? Impossible. We would still have 2.5 billion, correct? In 1950, the world produced enough food for 2.5 billion people, so that same amount of annual food would sustain 2.5 billion people today…or perhaps 2 billion taller, heavier people; or 3 billion people with more equitable, modest distribution and less waste. But you get the point: hold the food supply steady and you essentially hold the population below some cap. Inarguable. Those additional 5.5 billion people were made possible by a technological wave of food increase.

A Little Bit Louder Now…

Around that same period, the Isley Brothers’ song Shout had this fun sequence of “A little bit louder now,” progressively escalating in volume. That’s what actually happened to food supply, year after year. It is little coincidence that the highest-ever rate of human population growth sits right atop the Green Revolution. Food made babies. What else would babies be made of? As long as “decisions” to grow more food each year manifested, the positive-feedback result was essentially guaranteed. It’s certainly how things played out.

But the story did not start in 1960. Oh no. What was that other revolution that initiated the manipulation of plants for a higher level of human food production? Oh, it’ll come to me in a minute.

In the several hundred thousand years prior to agriculture, human population very slowly crept up, at a rate of about 0.0035% per year. Beginning about 10,000 years ago, the rate abruptly jumped an order-of magnitude, reducing doubling-time from the previous 20,000 years to 2,000 years—and monotonically decreasing thereafter as techniques improved (another case in point). By the 1960s, we grew at roughly 2% per year for a doubling-time around 35 years.

The volume knob (for food) was turned up a little bit louder each year, and human population rose along with it.

The Causality Question

Okay, okay. Push-back against the food-makes-babies formulation does not tend to refute the fact that increased food production went hand-in-hand with population increase. After all, babies are made from food. The argument centers on which drove the other. Sure: population increase requires food increase, but the reverse is not strictly true in pure (i.e., decontextualized) logical terms: an increase in food does not require population to increase. Surplus food could theoretically accumulate on the shelf, go to waste, be used for art projects, fuel food fights, etc. It seems perfectly plausible that food increases were always in response to population increase.

This is tangled territory, ill-suited for mental models. First statement: even if the imagined intent was more food in reaction to population increase, it’s quite possible that a misapprehension of the situation hid the causal nature of the positive feedback, so that more food actually preceded demand. In any case, we remained firmly in the exponential loop.

Second statement: let’s say people in the past didn’t get their act together fast enough and let several years or decades slid by before mounting an increase in production. Surely, procrastination is not a new human trait. The only way to keep population increasing (more mouths to feed) in this circumstance is for people to be progressively hungrier, with decreasing energy/stamina. That’s a hard place from which to mount a labor-intensive uptick. This argument—as any other—is not by itself conclusive, but the essential point is that nature offers no financing: to make a baby, food is demanded up front, which biophysically biases the situation for food coming first. Starvation conditions do not tend to produce baby booms that later require food production to increase.

Third statement: uncertainty and prudence tend to result in conservative over-production of food. Planning for surplus yield acts as insurance against all kinds of uncontrollable events. Lots of things can go wrong: low rainfall; a sweltering summer; a too-cloudy growing season; floods; insect waves; rodents finding stores; raids on your food supply by neighboring humans; the list goes on. A safe practice is to produce more than you believe you’ll need, as a matter of policy. In this case, surplus is not in reaction to population growth, but sure as hell enables it. It’s another substantial bias tilted toward a food-first causality.

Fourth statement: imperfect distribution means that even if enough total food is produced annually, some humans still go hungry and starve to death. In the modern era, the “humane” response has always been to seek increased food production—intending to fix the problem once and for all. As long as this is the (collective; automatic) decision, we stay firmly in the grip of the positive feedback loop. Increased food production in a global distribution system means even those who are not starving (but are not necessarily affluent) have greater access to food, which is where population growth tends to be highest. The phenomenon is also much more fine-grained than the whole-country level. A country hosting a starving subset of their population receives food imports and experiences population growth, but not necessarily dominated by the starving segment. The food going to the country stays in the country—some going to starving bellies and others going to well-enough-nourished bellies, some of which are “with child.” Being a global phenomenon, increased food also finds its way into countries that do not have significant starvation, but still express high fertility rates: food increase means more to go around. Many channels operate in parallel, even if our brains tend to tune in to only one at a time.

Fifth statement: aside from plague years, the experimental results are compelling: global population has increased for 10,000 years running—still to this day, for now. Meanwhile, we have never deliberately turned the global food production knob to anything other than “more” each year. If we haven’t, then could we, voluntarily? If we can’t, in practice, can we really claim to be in control—more than just notionally or aspirationally?

Some Data

Before getting into the legitimate confounders to the food-makes-babies recipe, I would like to share a peek I had into the situation, by using data from the Food and Agriculture Organization (FAO) of the United Nations (UN), which provides extensive datasets (here and here) on food production, imports, exports, nourishment levels, and lots more. I also utilize the UN’s demographic data in the form of the WPP (2022). An appendix at the end of this post provides more detail on my nerd methods.

Below are two plots, each averaging the five-year period from 2009–2013 inclusive (latest period in the dataset I used). The first looks at births minus deaths in each country, divided by population and expressed as percent. By counting births and deaths, migration in or out is neutralized. The second has Total Fertility Rate (average number of live child-births per woman for the period in question, where 2.1 is the nominal replacement rate for modern societies). Both are plotted against caloric food import fraction, where each dot represents a country—its area sized by population, and colored according to fraction under-nourished (blue is well-nourished and red is under-nourished).

See this list for three-letter (ISO-alpha3) codes. The South American countries far to the left are tremendous exporters of cereals, but also of oil crops, and animal products to a lesser extent.

See caption for previous plot. Both tolerate zooming in.

The different appearance of the plots is mainly a reflection of demographic inertia. TFR is a more reliable “now” measure, while higher net birth rates carry an echo of TFR a few decades ago when the current child-bearing population came into being. In other words, demographic bulges created in the past have worked their way to child-bearing years to keep the absolute number of births high. Both are valid captures in their own ways.

For each, I outline a box in the upper right in which net food importers are making babies. What this says is that population growth is largely in countries living beyond their local production means, requiring imports to fuel their growth. Overproduction in more affluent countries therefore supports population increase in countries of lower domestic food capacity. Below replacement-level TFR, import fraction seems basically random. Above replacement, the situation is heavily skewed toward food importers. One might say that this is the domain in which negative feedback has not yet overpowered the main positive feedback effect of food making babies (which has been the dominant story of the ages).

In Necker Cube fashion, the same plots support the opposite position: food excess (exporters) have low population growth (food makes not babies). The situation will never be settled if focusing on only one or the other aspect, as both are simultaneously true in a heterogeneous context—also complicated by a balance that shifts over time. Exceptions abound, but the net effect has been clear: the population growth manifesting today is largely supported by food importation. I’m not attempting to make a judgment here as to that practice so much as pointing out the biophysical underpinnings—as supported by data.

Real Grounds for Objection

In aggregate, then, I think we all agree that no food increases would mean a cessation of population growth. The hangup is that continued food increase need not translate to population growth, in theory. Except it always has, globally. But it doesn’t have to, and doesn’t do so in all locations—just in global aggregate throughout history.

Two related phenomena provide legitimate and powerful reasons to doubt the formula’s universality. First, affluent countries have the greatest access to food, yet show the lowest population growth rates (top plot)—lately even negative for some few. That’s real, and counter to the formula.

Secondly, total fertility rate (lower plot) is now at sub-replacement levels for countries hosting over two-thirds of global population, and these countries tend to be OECD (affluent). That’s also real. The sense this carries is that “civilized” countries have “arrived,” are in control, have defeated the “animal” feedback, and serve as a template to which others ought aspire.

Any theory needs to then account for the historical connection as well as recent exceptions, which is essentially impossible in a single-formula model in the context of a world teetering on a massive modernity-busting transition. Lots of past truths are about to be up-ended.

Nothing is Forever

Positive feedback never carries on forever, or the universe would break. Some influence (or many simultaneous influences) act to counter runaway growth at some stage. Having done so, the positive feedback contribution is not null-and-void. It doesn’t vanish from the equation: it’s just counterbalanced or overpowered by negative feedback. It’s more fair to say: food makes babies—historically—but only to a point. Food surplus made more babies for 10,000 years (and continues today), but not universally, and not forever.

In the current context of late-stage modernity, children become a costly affair in affluent societies. No number of all-you-can-eat buffets will offset the crippling cost of raising and educating children in a competitive market (obesity is also not the biggest turn-on). Adding to this is a sense among the younger-than-me generations (where babies originate) that prosperity peaked with the Boomers, that existential threats loom on all sides, and that the dream of the 1950s rings hollow today. Fertility rate is plummeting globally as the system that has been thumping along for 10,000 years is cracking up. The positive feedback is finally becoming overwhelmed, and is very unlikely to be exactly offset by the pile of negative feedback influences—more likely trampled by them. It appears we’re heading for a major population haircut, even if by “benign” demographic adjustment. But again, just because the historically dominant positive feedback mechanism is becoming overwhelmed is not the same as being invalid, or plain wrong. The context is changing, and I would say not by the exercise of control.

Speaking of context, the emerging fertility decline was not on the radar when Daniel Quinn kept hammering the point that continuing an annual increase in food production was a form of insanity: doing the same thing for 10,000 years in a row and expecting a different result. I have to say that I can’t see how he was incorrect. Had we stopped annual food increases at any point along the way, it is very clear that population could not continue to soar in some sort of biophysical detachment. That was his whole drive for returning to the subject: stop the insanity of food increases and the insanity of population increases will necessarily stop—because food, indeed, makes babies. Not wrong, but a complete political dead-end.

Reactions to the Knob

And to that point… Since Taker culture is obsessed with control of the planet, wouldn’t the ultimate control fantasy be to dial annual food production in such a way as to control human population? We know it to be effective. We also know that food production has only ever increased (minor, uncontrolled hiccups aside), while global population has only ever increased (again with the rare hiccup).

Let me be clear that I’m not proposing food curtailment as the right solution: attempts at (illusory) control are always going to go sideways, have more unintended consequences than intended consequences, and invariably hurt those who are not jerking at the levers of (partial) control. I just find it odd that those who cite our superior control as a basis for rejecting the animal-adjacent food-makes-babies formulation do not tend to advocate actual control.

This, I believe is the crux of why I am drawn to this “fight.” What can we learn about fundamental drivers motivating push-back against the biophysical formula? So, I ask you to consult your own feelings on the matter, in a thought experiment, to get at the core nerve the subject touches (if it even does, for you).

• What objections arise to the proposal of capping global annual food production?
• Does it seem cruel?
• Does it seem wrong to control people?
• Does it seem wrong to impose limits on humans?
• Should humans get whatever they want?
• Are limits evil, when applied to humans?
• Is the only way to be “fully human” to live outside of biophysical/ecological restrictions?

I’ll just say that some of these attitudes have disastrous consequences (witness modernity). Now, consider what these objections might translate to.

• Does this mean, in effect, that we will never voluntarily reduce or deprive humans of food?
• Are we therefore effectively powerless to the food-makes-babies phenomenon?
• Do we really have control, or is it more imagined/theoretical/aspirational?
• Is it more the case that other unbidden factors assert/manifest before the cycle is broken?
• Should it be: food will make babies until other factors beyond our control intervene?

To what extent, then, is our sense of control largely illusory? Who is it that planned the meta-crisis; 8 billion people; the future population plunge; initiation of a sixth mass extinction? It seems to me we got swept up in the currents, now imperiling the world. Along the way, a lot of food made a lot of babies—packing the stadium for the great spectacle of collapse under the weight of the assembled crowd.

Appendix: Procedure

For the nerd-curious: I used the FAO Food Balance data (available here) as well as the UN WPP data from 2022, and the FAO food security data here.

FAO changed their format after 2013 for Food Balance, and because I wanted the ability to explore a longer time interval, I went with the 1961–2013 dataset. This dataset is a spreadsheet 238,419 lines long. For each of the countries (and pooled regions), a variety of fields are provided to track food at different levels of aggregation (total food; animal vs. vegetal; about 18 sub-groupings; and then roughly 50 fine-scale designations). For instance, wheat and associated products are at the fine scale, belonging to the cereals group, which itself is in vegetals, and contributes to the total. Import/export data only becomes available at the group (e.g., cereal) level, and that’s where I worked.

In each category, total domestic supply is computed as domestic production plus imports minus exports plus anything that came from stock (which is negative if being cached). Of this supply, some is used for animal feed, some may be used for seed, some is lost, some is used in processing, and there’s even an “other” category (decorative macaroni art?). What’s left is called “food.” This quantity is also provided in translated form as kilograms per capita, and then kilocalories (kcal) per capita per day (and further broken into protein and fat content).

So, for all the categories, I compare the total net imports (negative if exports exceed imports) to total domestic supply (in mass terms) to compute the aggregate fraction of imported (exported) food energy for that group. To compare apples to beefsteak, I express each group’s imports in terms of kilocalories (multiplying by import fraction for that group to track imported food energy) in order to be able to add up across all groups.

A careful eye might notice that the plots showing fractional food import look like they are weighted (considering population, also) toward imports. This can’t be right! All imports come from exports somewhere in the world, so the population-weighted balance should be neutral. The reason it’s not is subtle. Because I am using each country’s fractional import/export based on caloric content, but not all countries operate on the same caloric basis (variation in kcal per capita), and exporters tend to enjoy higher kcal per capita, the result is skewed. In effect, the importers utilize the exports more frugally, so that a larger import fraction among light-eaters is offset by a correspondingly smaller export fractions in high-calorie countries. Anyway, I checked that absolute imports balance exports to my satisfaction.

Kilocalories per day per capita vs. import fraction. See caption for first plot in post. Net exporters (left) are biased toward higher caloric supply, unsurprisingly.

The bias in question can be seen in the plot above: exporters (left of zero) have higher-than-average caloric supply, and importers trend lower. This skew offsets the imbalance in the other plots to make it all come out right. Note that the color scheme for nutrition levels track per-capita caloric supply relatively well (blue on top; red on bottom). Note also that not all supply is eaten, given (sometimes substantial) food waste.