I just saw another comment implying that immigration was good because it increased GDP.  Over the years, I've seen many similar comments in the LW / transhumanist / etc bubble claiming that increasing a country's population is good because it increases its GDP.  These are generally used in support of increasing either immigration or population growth.

It doesn't, however, make sense.  People have attached a positive valence to certain words, then moved those words into new contexts.  They did not figure out what they want to optimize and do the math.

I presume they want to optimize wealth or productivity per person.  You wouldn't try to make Finland richer by absorbing China.  Its GDP would go up, but its GDP per person would go way down.

I know why LWers want to say that increasing GDP is good.  Historically, LessWrong is associated with transhumanism and specifically extropianism, and one of the main opponents of transhumanism is the romantic, anti-technology part of the environmentalist movement.  The Extropians were strong supporters of Julian Simon, an economist who argued against environmentalists that population growth would not lead to collapse, but to lower rather than higher prices.

Growth is good.  That was one of our mantras in the 1990s.  And economists measure growth via GDP.  So increasing GDP is good, right?

Wrong.  Increasing GDP is not the same as productivity growth.  Productivity makes sense either per person (when speaking locally) or per planet (or other isolated system, when looking at the big picture or the rate of technological change).  GDP stands for Gross Domestic Product.  It refers to the total yearly market value of the product of those people contained within some rather arbitrarily drawn border on a map, and is used for national budget planning.  Both are surprisingly complex to compute, but in very different ways.

If you have a constant population, and GDP increases, productivity per person has increased.  But if you have a border on a map enclosing some people, and you move it so it encloses more people, productivity hasn't increased.  If instead of moving the border over the people, you move the people over the border, productivity still hasn't increased.  Those people might be more productive on the other side of the border, but I haven't seen people make that argument.  They just say increasing population increases GDP, so it's good.

If your population grows, your GDP will grow.  That's not what Julian Simon was talking about.  Simon argued that as population density rises, wealth per person will grow.  He didn't say that increased GDP is a good thing, but that increased productivity is good.  GDP is only correlated with productivity. (In fact, he argued against using even GDP per person as the sole measurement of growth:  "Every time a human baby is born, the per capita GDP falls."  Increased productivity makes prices fall, which makes wealth go up but GDP go down.)

Remember that the concepts of GDP and productivity are separate.  You might be tempted to call GDP "productivity per country", but it just doesn't mean that when it's used in discussions about changing the population size or density of the country.  It is invalid in that context to infer changes in productivity, and hence positive or negative valence, from changes in GDP.  People would realize that if they remembered why they think increasing GDP is good, but I fear they don't--they just note the positive valence they've attached to the word and assume it's still valid in the current context.

New to LessWrong?

New Comment
24 comments, sorted by Click to highlight new comments since: Today at 3:14 PM

I thought the usual claim was not "immigration increases total GDP" but "immigration increases per-capita GDP". Random example: this paper which, full disclosure, I have not read but only looked at the abstract linked.

I'm not sure any measure of GDP (total or per capita) is a great way of assessing immigration. Consider the following scenarios:

  • A scenario like Phil's: An immigrant moves from a poor country to a rich country. They somehow become less productive when they do this, and earn less than they did before. But their income is positive, so the total GDP of the rich country goes up.
  • A scenario showing the opposite problem: An immigrant moves from a poor country to a rich country. In the rich country they earn less than the average person there, but more than they did before. The immigrant is better off. The other inhabitants of the rich country are (in total) better off. But per capita GDP has gone down.

It seems to me that to avoid Simpson's-Paradox-like confusion what we really want to know is: when some people migrate from country A to country B, what happens to (1) the total "GDP" of just the people who were already in country B and (2) the total "GDP" of just the people who moved from country A? We might also care about (3) the total "GDP" of those who remain in country B. My guess, FWIW, is that #1 and #2 both go up while #3 goes down.

We might also want to compute the sum of the GDP of A and B: does that person moving cause more net productivity growth in B than loss in A?

If you have a constant population, and GDP increases, productivity per person has increased. But if you have a border on a map enclosing some people, and you move it so it encloses more people, productivity hasn't increased.

Can you give examples of people confirmed to be actually making the mistake this post discusses? I don't recall seeing any.

The standard economist claim (and the only version I've seen promulgated in LW and EA circles) is that it increases gross world product (total and per capita) because migrants are much more productive when they migrate to developed countries. Here is a set of references and counterarguments.

Separately, some people are keen to increase GDP in particular countries to pay off national fixed costs (like already incurred debts, or military spending).

world total and per capita GDP

Did you mean world to modify GDP? If you did, that's really confusing, because GDP ("domestic") is specifically local. If you concatenate "world GDP" is pretty clear what you mean, but if you separate that like this, it is natural to parse it as "world and national," which is probably not what you mean, since that is pretty much the error Phil is talking about. Your links are careful to always concatenate, though.

I meant GWP without introducing the term. Edited for clarity.

When it comes to funding science higher GDP means more tax revenue and thus more science even when GDP/person stays the same.

The same goes for a new TV serial.

I agree with the substantive point that the changes in living standards we ultimately care about come from productivity growth, not GDP growth as such.

Now for the inevitable disagreements/critiques:

  1. The post title, "Increasing GDP is not growth", isn't actually true as such. Referring to increasing GDP as economic growth isn't a weird LW/transhumanist/etc. affectation; it's totally normal & conventional. If I heard a newsreader talking about economic growth, I'd guess they were most likely talking about (inflation-adjusted) GDP going up.

  2. The people I most associate with this kind of immigration-grows-GDP meme are single-issue-ish advocates of open borders, but AFAIK they refer to immigration increasing global GDP (which is kind of a strange usage of "GDP", since the "D" does after all stand for "domestic", but that's what they say). It's right there in the title bar. This avoids the issue of country-level GDP being dependent on how one draws lines on a map.

  3. It took me a moment to realize the post wasn't using "productivity" in the conventional economist's sense (at least when talking at the national level), namely labour productivity, which is output per worker or output per hour worked. This is distinct from output per person, and made the references to "productivity per person" momentarily confusing (because with the conventional understanding of productivity, writing "productivity per person" is a bit like writing "GDP per capita per person").

  4. PhilGoetz probably knows this already, but I don't know whether everyone reading this does: some advocates of higher immigration explicitly expect immigration to raise productivity, specifically of workers who come from lower-productivity countries to higher ones. The usual term for this is the "place premium". A Google search brings back among other things an illustrative working paper by Michael Clemens, Claudio Montenegro, and Lant Pritchett.

These are all excellent points. The increase in labor productivity accruing to immigrants to e.g. the US is often discussed by economists. I'll grant that it's not often discussed in general media, which is part of PhilGoetz's point, but I'm sure I've seen it there too.

Also, many economists have argued that in certain contexts immigration (even low-skilled) does result in economic gains for the native born. The argument goes that immigrants' negative impact on native born wages is small and that this small change is more than offset by the immigrants' ability to make domestic goods cheaper. People earn less, but things cost less still. In this scenario GDP per person has gone down, but native born purchasing power has increased. And the immigrants are far better off- their labor productivity has increased through the place premium and (related) their wages are almost certainly higher than they would be in their country of origin.

A final point:is that total GDP in some contexts can actually be a good in itself. Having a country with a large GDP is very meaningful politically. The fact the Norway and Qatar have higher GDP per capita than the US is meaningful and worthy of discussion, but the US matters far more to global politics and it's due primarily to one reason- the US is responsible for one quarter of the entire world's nominal GDP. Along some margin, allowing immigration that lowers GDP per capita but raises total GDP can be beneficial to members of a country purely based on international economic strength.

Related to your second paragraph, I was intrigued when I saw economists pointing out that even low-skilled immigrants could raise natives' productivity & income by (1) nudging natives to upskill and move into higher-skill jobs, and (2) lowering childcare & housework costs, making it easier for native women to work paid jobs.

I believe the argument was (or should have been) that immigration increases the GDP of the destination country by more than it decreases the GDP of the origin country, thereby increasing worldwide GDP per capita.

That should be the argument.

But the argument is also made that "GDP goes up" in the context of the immigration debate and nationalism. That is the usual context of the question these days.

From a nationalist perspective of the existing citizens of the country, their GDP may go down when immigrants are allowed into the country, even while the GDP of the aggregate including the new immigrants is higher than the original GDP of the existing citizens.

And of course, the influx of immigrants also affects the distribution of GDP within the existing citizens.

Well, GDP, productivity... so 19th century-ish.

How about GIP?

Gross Intelligence Product?

I'd like to take this a step further, even. If you are a utilitarian, for example, what you really care about is happiness, which is only distantly linked to things like GDP and productivity, at least in developed nations. People who have more money spend more money,^[citation_needed] so economic growth is disproportionately focused on better meeting the desires of those with more money. Maybe the economy doubles every couple decades, but that doesn't mean that the poor have twice as much too. I would be interested to know precisely how much more they actually do have, perhaps as measured by decreasing inflation adjusted rent prices, food cost, utility costs, used car prices, bus fare, etc.

Only a tiny fraction of startups are aimed at making rent cheaper or groceries affordable, although such benefits may sometimes trickle down. I'm somewhat hopeful for things like 3D printed luxury homes for the wealthy or Soylent for techies who want to countersignal.

If the rich really were utility monsters capable of enjoying each dollar of luxury items at least as much as the poor would enjoy the dollar, then there wouldn't be a problem according to a pure hedonic utilitarian. (Although, a desire for Fairness is apparently a human universal, according to moral foundations theory, so actual humans would be uncomfortable with this, even if their Liberty value was stronger, and overrode the concerns with fairness.)

It appears that money really can buy happiness for the poor, and that's still somewhat true for the middle class, but once you get to around upper middle class, it becomes extremely difficult to even measure additional gains in happiness. It's not clear to me that there even are gains beyond that point, but feel free to read the article on Happiness Economics yourself, particularly the sections on the effects of income on individuals, and the GDP/GNP section.

In fact, some countries appear to be much happier than you'd predict just based on knowing their GDP, while others are much less happy. This TED talk makes the case for optimizing for Social Progress Index rather than GDP, but I believe the field of Happiness Economics also uses things like Gross National Happiness, Satisfaction with Life Index, and the World Happiness Report, and I don't really know enough to have a strong opinion between them. I just think something in that general direction is a better metric than GDP or productivity.

Only a tiny fraction of startups are aimed at making rent cheaper or groceries affordable

Selling groceries is a low margin business. There isn't much room for a startup to sell radically cheaper groceries.

At the same time companies like Amazon do invest in optimizing grocery sales. Technology such as self-driving trucks have the potential to lower grocery prices. There are also various companies who work on making farming more efficient. Both GMO and putting more robots on farmland has the potential to make food production more efficient.

Agreed. I'd love to see even more of all of these sorts of things, but the low margin nature of the industry makes this somewhat difficult to attack directly, so there isn't anywhere near as much money being invested in that direction as I would like.

I believe NASA has gotten crop yields high enough that a single person can be fed off of only ~25 m^2 of land, (figure may be off, or may be m^3 or something, but that's what I vaguely recall.) but that would have been with fancy hydroponic/aquaponic/aeroponic setups or something, and extremely high crop density. It would be awesome to see fully automated vertical greenhouses pumping out GMO produce for almost 0 cost.

I recently saw someone joke about engineering GMO wheat as an invasive species to out-compete grass. If we wanted to, I suppose we could also replace all the planet's trees with fruit trees, and build ourselves a garden of Eden, with an absurd surplus of food, available for free. That's probably a little extreme, considering that some people are rather attached to nature as it is, but maybe we'll terraform other planets like that?

Just some musings and paradise engineering. It’s interesting to consider various post-scarcity economies where things we work hard for are as common as air.

Fancy hydroponic setups aren't almost 0 cost. Vertical farming is not cost competitive with the farming that we have.

There are companies like http://rowbot.com/ who develop farm robots to more effectively use fertiliser.

There are plenty of devices for growing food at home on kickstarter.

I suppose we could also replace all the planet's trees with fruit trees, and build ourselves a garden of Eden, with an absurd surplus of food, available for free.

Having fruit trees isn't free. Having fruit trees in cities costs the cities enough to prefer other kind of trees.

It's worthy to be suspicious of free. Most of the time there's some work involved. It's worth understanding the hidden costs.

happiness, which is only distantly linked to things like GDP and productivity, at least in developed nations

This implies that no one cares much about periods when the GDP growth turns negative -- and these are called "recessions" and, when the contraction is severe enough, "depressions". Is that so?

economic growth is disproportionately focused on better meeting the desires of those with more money

I don't think that's the case. Take a recent economic growth success story -- China. The major effect was lifting hundreds of millions of people out of poverty.

No one starts startups to make rent cheaper or groceries affordable

I don't think that's the case as well (as long as we are talking about consequences rather than intentions). Uber is cheaper than taxis, AirBnB is cheaper than hotels, etc. And Walmart was a startup before startups were cool :-P

these are called "recessions" and ... "depressions".

Ha, very good point. Our current society is largely built around growth, and when growth stops the negative effects absolutely do trickle down, even to people who don't own stocks. In fact, companies were counting on those increases, and so have major issues when they don't materialize, and need to get rid of workers to cut costs.

I will mention that through most of history and prehistory, the economic growth rate has been much, much, smaller. I haven't read it, so I can't vouch for its quality, but apparently the book The End of Growth: Adapting to Our New Economic Reality suggests that economic growth can't continue indefinitely due to physics limitations, and lays out a framework for transitioning to a post-growth economy. I have no idea how gentle or unpleasant such a transition might actually be. (Also, note that I am hopeful that we can avoid resource limitations by transitioning to a space based economy, and am nowhere near as pessimistic as I think the authors are likely to be.)

China.

China did indeed achieve massive benefits from industrialization. There's a lot of evidence that maximizing economic growth is an excellent way to play catch-up and obtain modern amenities for your population. Perhaps it's even the fastest theoretically possible way, since access to capital is the limiting factor for improved quality of life, and selling cheap stuff gets you lots of capital. I don't think developing countries should try communism or anything like that, unless for some reason they expect it to result in higher economic growth, since the data suggests that free markets are much better for them.

I would, however, suggest that the price of basic amenities appears to me to be a limiting factor in the quality of life of poor people in developed nations, and that increases in national wealth tend not to translate into proportional increases in purchasing power for them, although there is still some gain. (As I said before, I should really look into the details, though.) I see 2 basic classes of solutions:

  1. You can try to funnel more goods and money to them. This might be done through tax structures, aid programs, education, basic income, etc. Either you try and improve their earning potential, or give them things directly, but either way they wind up with more. The end result is that they can purchase more such amenities at the same price, or perhaps a little cheaper due to more economies of scale and more competition for those items.

  2. You can try to funnel more R&D into the sorts of things that the poor want than a free market would otherwise do. Most of the ways of doing this will cut into GDP somewhat, but maybe there are some public good type things that would out-preform the market, but where the benefits are difficult for one company to capture. A dollar spent on specific types of education, for example, may increase GDP by more than 1 dollar. However, since it can be difficult to capture a return on investment,^[1] we have a tragedy of the commons scenario, and government or some powerful entity has to step up and foot the bill for the common good, if we want things like that. (Note that I'm not sure that this is still true on the margin, just that if we cut all funding for education that the GDP would drop by more than the amount saved.)

No one

I was being a bit hyperbolic there, but you'll note that I followed it with 2 examples of startups which might in the future actually be cheaper than what the poor currently use. (3D printing might remove labor costs from construction, and Soylent has aspirations of making food into a utility. I probably should have said so specifically.)

Walmart is a good point. I’m not sure whether the benefits from cheaper goods outweigh the cost to local jobs, but I’m sure we’ve both heard the complaints. That’s getting dangerously close to talking politics, so I’d prefer to avoid getting into details, but I’d be interested if anyone knows of any academic research or cost-benefit analyses.

Uber may be cheaper than taxis, and AirBnB may be cheaper than hotels, but the poor don't use taxis or hotels. I am hopeful that self-driving cars will make transportation cheap enough that the poor benefit, though.

My point wasn't that the poor aren't any better off decade by decade. That appears to be false. My point is that they aren't 5% better off each year, even though the economic growth rate is about maybe 7.5-ish-percent with maybe 2.5% inflation. So, most (but not all) of that growth is going into sectors which don't benefit the poor much.


[1] Interestingly, this appears to be precisely what Signal Data Science's business model is. They teach you in exchange for a fraction of your future salary. However, perhaps due to irrationality, there doesn't seem to be a wider market for this sort of thing.

through most of history and prehistory, the economic growth rate has been much, much, smaller

True, and I think not many people want a return to those times (some do, though, mostly on environmentalist grounds).

increases in national wealth tend not to translate into proportional increases in purchasing power for [the poor] ... My point is that they aren't 5% better off each year, even though the economic growth rate is about maybe 7.5-ish-percent with maybe 2.5% inflation.

That's the whole growing-inequality debate, a separate highly complicated topic.

outweigh the cost to local jobs

Economically speaking, a job is a cost. If you can produce the same value with fewer jobs, that's a good thing called an increase in productivity.

A job is a cost

Agreed. When I said the "cost to local jobs" I was being informal, but referring to the (supposed) increase in unemployment as Walmart displaces local, less efficient small businesses.

Paying people to do a job which can be eliminated is like paying people to dig holes and fill them back in. I'd rather just give them the money than pay them to do useless work, but I'll take the second option over them being unemployed.

As an interesting side note, I think this might put me on the opposite side of the standard anti-Walmart argument. The meme argues that, Walmart not paying its workers a living wage and making it difficult to unionize forces the government to step in and provide aid, and that this is in effect subsidizing Walmart.

However, because Walmart sells mainly to the poor, I am in favor of subsidizing them in any way that passes through to the poor and doesn’t get skimmed off the top. Maybe that would mean I’d even be against a law forcing them to pay $10/hr or some such, if the benefits to the employees didn’t outweigh the net drawbacks to the customers.

Mainly I just find it depressing that all current political narratives seem to ignore these complexities, and boil down to “Walmart bad” or “markets good” or whatever. Maybe some more intelligent conversations happen behind closed doors, where no one can hear politicians make sane concessions to the other side.

Mainly I just find it depressing that all current political narratives seem to ignore these complexities, and boil down to “Walmart bad” or “markets good” or whatever. Maybe some more intelligent conversations happen behind closed doors, where no one can hear politicians make sane concessions to the other side.

Very few politicians of either side argue that Walmart is bad. At least outside of the local level.

GDP easily becomes a lost purpose, but this argument could have been done more clearly. There are two unrelated things here to talk about:

a) Does GDP truly reflect the generation of value, or is it influenced by technical details in accounting, and how much? For example, if what was previously a gray economy (e.g. friends cooking at home for each other) becomes official economy (e.g. everyone eating in a restaurant), the measured GDP will increase simply because what was previously ignored becomes measured, even without any corresponding increase in productivity. Actually, pushing for artificially increasing GDP by making all gray economy regulated could decrease productivity (as e.g. the owners of the restaurant need to deal with paperwork they didn't have when they cooked for their friends unofficially).

b) Do we care about total productivity of the country, or productivity per capita? Intuitively, people care about productivity per capita, because they care about their personal well-being. However, in certain contexts, such as international negotiation, having a large total productivity might be an advantage.

Speaking of immigration, immigration is not absolutely productive. There are a myriad of factors at play here. If one wishes to inject a population with a group in order to increase the population's overall productivity, they must ensure that the injected group is as-productive or more productive than the original population.

I don't think whether the group's average output increases or decreases is the right metric. What's important is whether the newly enlarged group's output is higher than what the group and its new members' total output would be if the groups weren't merged - do the immigrants become more productive by immigrating, and do they make the native population more or less productive?