Throughout much of the developed world we are seeing a resurgence in populist movements, often with an explicitly nationalist and / or anti-globalist focus. Trump’s candidacy in the US, and the recent vote by the UK to leave the EU are the best-publicised of this sentiment, but it is visible in other countries. Pauline Hanson’s political influence has risen from the dead in Australia, and much of continental Europe is seeing flare-ups of nationalism as well.
It’s also worth noting that while the right-wing nationalists are getting the most attention, there are similar sentiments oc the left as well. Bernie Sanders’s campaign relied on a sense that the economic system was rigged against ordinary Americans, and a large portion of the protests against the Trans-Pacific Partnership have come from the left. While the source of the protests is less about migration on the left and more about trade deals, there is a common sentiment running through both groups – that the people running the countries comprising the developed world are privileging an allegiance to an ideal of cosmopolitan global capitalism over the interests of their own constituents.
It will come to no surprise to most of you that I consider this trend toward nationalist populism deeply disturbing. I strongly believe that having fewer restrictions on the movement of both people and trade across borders is good for the vast majority of people in a society – including most of the people campaigning for change. And that belief comes from over two centuries of economists investigating the topic. If the populist resurgence continues, I am concerned that we could find ourselves in a downward spiral where poverty causes populism causes nationalism causes more poverty. This would lead to unstable or opportunistic demagogues growing in power until they begin to collapse or radically alter governments. This is exactly the sort of discontent that could lead to the rise of fascist governments – after all, it has before. I don’t know if something like that could happen again, but I also don’t want to find out.
Many commenters on the web, including here, have argued that whether or not the rising discontent is justified or focused on the right target, it must be taken seriously for the good of liberal democracy. And after dwelling on this issue for a few weeks, I agree. After all, it’s all too easy for someone like me to outright dismiss a lot of the anti-globalisation sentiment that is floating around right now. An education in economics gives you a very good grounding in the various ways trade between countries benefits people, and how nearly all of the popular arguments against liberalising trade are fallacious. That tends to make you dismissive of people who blame their woes on globalisation. Furthermore I’m a government employee living and working in my nation’s capital. We don’t use the term “beltway insider” in New Zealand, but if we did I would be one.
So let’s begin with the idea that there are a lot of people who feel that they are experiencing economic hardship, and that needs to be taken seriously. On the other hand, that also doesn’t necessarily mean that the people opposing globalism (be that in terms of wanting tighter trade restrictions or tighter immigration restrictions) are right either in terms of their conception of the problem, or their proposed solutions. This is usually the point where the debate collapses into mutual recriminations: we call them ignorant, they call us callous and out of touch; we call them racists and xenophobes, they call us sellouts to an international elite etc. Rather than doing that this time, I’d like to try and get at the fundamentals of the economic conditions that are driving this opposition to globalism.
This post won’t be about offering solutions – the economics of this is complex enough that I can’t do it justice in a blog post – it will take a lot of careful empirical work to tease out what is going on here. But what I can do is try to identify what the key research questions are, so we can least start investigating and arguing about the right things.
Question 1: How bad is it really?
A lot of people feel as though things are worse than they were decades ago – the 1970s are a common period used as a comparison. But subjective experience is a poor guide to reality – human memory is notoriously unreliable, and one’s perception of what the past was really like can be biased by nostalgia – when you’re a child the world can seem freer of problems than it really is. The sources of information we are receiving bias our view of the world as well – crime rates have been falling for 20 years, but fear of crime continues to rise. Also, some changes to your standard of living are less visible than others – your take-home pay is a single, visible number, but a bunch of the goods you consume getting steadily cheaper in inflation-adjusted terms is much less obvious.
While these factors are important, we also must be cautious when pursuing this line of reasoning. It’s really easy to trap yourself in a closed loop where every murmur of discontent is dismissed as the product of biased reasoning. In a previous OT post on Brexit, commenter Chip Daniels analogised this form of thinking to Socialist ideas of False Consciousness, and it’s a warning we should take seriously. Just because the anti-globalists might be biased doesn’t mean they are substantively wrong, nor does it mean our thinking is unbiased either. This is why we need to go back to the data to get a handle on how standards of living are changing over time.
Unfortunately, the available data has its own limitations that need to be worked around:
- GDP is fine for measuring what it was designed to measure (basically a nation’s taxable output), but it provides limited insight into standards of living on its own.
- Income data is better, especially if you have the whole distribution and not just the mean or median. However, to understand what is actually happening to people’s standards of living, you need to analyse household income data instead of personal income. Unless you live in a country where everyone has to register their residence with the government at all times, and have a means of linking that register to your income data, that is a much harder thing to do than you might think.
- And even income data doesn’t capture the whole story. For one thing, you need to account for the change in prices over time. After all, your income is only as good as the goods and services it can buy. This is what the Consumers Price Index is for, but the trouble is that there is a known problem with the CPI that causes it overstate price growth as it doesn’t account for changes in product quality or changing spending habits very well. (This is especially problematic if changes in relative prices are causing spending habits to change.) This makes it very difficult to compare standards of living over long time periods. And if you deal with that problem, you can then start worrying about the fact that people at different points of the income distribution consume different goods and services. The price of tertiary education services matters a lot more to the middle class than it does to the poorest people in a country.
- You also need to account for demographic changes. There are two major sources of demographic bias to standard of living data – age and immigration:
- People have systematically different incomes at different stages of their life. Income rises with age up until a little before retirement age, then falls. If the population’s age profile is changing over time (spoiler warning: it is), then that can affect an individual’s standard of living without the economy doing anything different.
- Immigration can bias standard of living data in paradoxical ways. Imagine a worker in a poor country who could earn more money if they moved to the US, but less than the average American. If they move to the US they will lower the average income in the US even if they don’t affect the incomes of any native, due to how averages work. If you’re worried about immigration it’s especially important to adjust for this effect.
So, as you can see, there’s a lot to be done here to work out if the perception is based on reality. And that matters because that tells us whether we have an economic problem, a perception problem, or some other kind of problem that’s being expressed as if it were an economic problem. Only if there is an economic problem to solve, should we be trying to develop an economic solution. And to properly target that solution, we need to answer the next question:
Question 2: Who, specifically is losing out?
Hopefully, the research performed will not only help us determine the magnitude of the problem, but also who specifically is losing out. It’s important that this is considered from as many angles as possible. Most political analyses, at least the ones that are discussed publicly, tend to rely on crude classifications of affected groups, such as:
- Workers vs corporate shareholders
- Native-born citizens vs migrants
- Your country vs foreign countries
- The 1% vs everyone else
- The poorly-educated vs the well-educated
But none of these divisions are aligned with the effects of trade or migration very well. The welfare economics of trade is complicated, whether any specific person gains or loses from trade liberalisation depends on a number of variables:
- Do they work in an industry that is vulnerable to foreign competition? If so, how easily could they apply their skills in another industry? People working in import-competing industries with skills that are difficult to transfer to other industries are the most likely to suffer from liberalised trade.
- Do they work in an industry which exports goods? If not, do they have skills that could be applied to an export industry if it expanded? Exports and imports tend to rise and fall together, so more imports from overseas means more opportunities for exporters to expand.
- Are they likely to spend a lot of their incomes on imports, or import-competing goods? A reduction in cost of living is strictly equivalent to an increase in income. The poorest people in a society are often in the best position to benefit from cheap imports.
The story is similar for immigration. What matters is not what broad political demographic someone is part of, but whether the labour of immigrants is more likely to be competitive with yours or complimentary to it, and whether you mostly consume or mostly produce the goods and services that migrant labour tends to produce.
However, on top of all that is another layer of complexity that starts to play in to the political economy of globalisation. Even if you are not directly affected by globalisation much, you may be indirectly affected, and that will depend a lot on where you live. If you live in an area that is dominated by industries that are competing with imports then even if your job is to produce something non-tradeable, you may still suffer because your neighbours have less spending money. This story is less plausible for migration, but if the services consumed by the migrants are sufficiently segregated, you could still see this effect with migration.
Assuming there is a real economic problem here, I think these geographical effects are key to understanding why it is so polarising. If you live in a community that is heavily dependent on a single industry that is dying, then your entire community may be in economic decline. Availability bias would be enough to to make it feel like the entire country was dying – that it needed to be made great again. Conversely, if you live in an area whose main industries have benefited, or not been affected much by globalisation, that same availability bias might make it impossible for you to understand why anyone could think the nation was in decline.
This is just a hypothesis – but i think it merits investigation.
Question 3: Why are the losers losing?
This is the hardest question to answer, but vitally important for developing an intelligent solution. If the analysis I outlined above does identify specific groups whose standards of living have declined, the next step is to figure out why. And by “why” I don’t just mean “the mill closed, so no one can get jobs anymore”. If we want to solve the problem, it’s not enough to identify the proximate cause. We need to figure out what caused the centre of economic activity to leave that community, and why nothing has risen to replace it. Because, make no mistake, the popular narratives do not make sense. Neither trade nor immigration cause persistent unemployment, and the standard version of the “automation is rendering people unemployable” doesn’t really work either.
The thing to understand about unemployment is that there is not a fixed quantity of jobs to be done. This misconception is so common there is a name for it: the Lump of Labour Fallacy. A better way to look at employment is as a cost-benefit calculation. If there is something you can do for someone that is worth more to that someone than you cost to hire, then there is money to be made in hiring you. The market isn’t 100% rational, but someone is going to spot the opportunity sooner or later. So if people are losing their jobs, but no new jobs are being created, we need to figure out where the benefit-cost calculation is breaking down:
- Is this just the tail end of the recession? If so, the best solution may be to wait it out.
- Are there institutional or cultural factors that make an area too risky for new investment? This is an explanation that normally applies to why developing countries have trouble growing, and I have some difficulty believing it would apply in most of the developed world. Still, it is worth pursuing as a hypothesis – is it the case that some communities have structured their regulatory system so completely around a single industry that they are no longer in a position to support new industries springing up?
- Is the problem skill change? Are the people losing their jobs unable to apply their skills to the new ones being created? This could simultaneously result in unemployment and a labour shortage in different parts of the country (which would tie in to a region-based story). It could also result in permanent unemployment of some people, not because jobs have been “shipped overseas”, but because employers may no longer value a person’s skill set enough to justify the cost of hiring that person. This is how technology could cause long-term job loss, by raising the minimum level of skills required to count as having marketable job skills, though I want to emphasise that this is only a hypothetical possibility at this point.
- Relatedly, is it that people are finding new work, but it doesn’t pay as well as the old work?
- Is it a demographic story? Rural areas tend to gain population through births and lose it to migration to urban areas. Birth rates have been declining in most developed countries for decades, so we may now be reaching a point where births are no longer able to replace the people leaving rural communities for urban ones. A community needs a certain population to be viable, so falling rural populations may be damaging the economies of small rural communities through attrition.
Which solutions we might need to consider will depend crucially on which, if any, of the above possibilities is happening (or for that matter, if the truth is something I haven’t even thought of). And some of these are a lot harder to deal with than others. In some cases, the best solution may be palliative rather than a cure. But before we can even think about that we need to put broad assertions and mutually suspicion aside and actually figure out what is really going on here because you can’t make good plans out of bad data.