“Manufacturing is Roaring Back Under President Donald J Trump”, the White House claimed in a press release earlier this year. And it’s clearly been a major focus of the Trump administration. One reason for the proliferation of large tariff increases on US imports has been the expectation – or hope – that domestic producers, sheltered from overseas competition, will fire up the steel mills and the car plants again. Other countries have been taking a look at the idea, too.

It’s an interesting policy question. Can you breathe new life into manufacturing? And if you can, should you? 

Spoiler alert: probably no, and probably no.

A trip down memory lane

Let’s start with some long-sweep historical data. Take a look at the chart of employment in US manufacturing, taken from FRED, the very useful – and free – database run by the Fed of St Louis. Employment in manufacturing peaked in mid 1979, 45 years ago. What happened?

US employed in manufacturing

Entering WWII increased manufacturing, jump-started the economy and helped end the Great Depression.

Some of it, for advanced economies like the US or Western Europe, was down to developing economies – particularly in Asia – upping their game. Japanese carmakers, for example, with more reliable, cheaper and fully featured offerings, ran rings round the likes of American and UK models. And some of it was down to productivity gains: as automation took hold, you needed fewer people on an assembly line.

But the biggest answer is a centuries-long structural shift in the pattern of economic growth. As a blog post on the FRED site says, “Historically, economic development has led to a declining share of workers in goods-producing sectors. The first sector to decline is agriculture, whose workers moved to manufacturing and mining during the Industrial Revolution… In the 19th century and beyond, the US economy grew further and progressed to the next phases of development, with mining and manufacturing losing relative importance… Clearly, the US economy is now much less focused on ‘making things’. Rather, the emphasis is now on education, health, leisure, retail, information and finance.”

Just how much less focused the US has become on “making things” was nicely illustrated by Alan Greenspan, formerly chair of the US Fed, when he pointed out that US GDP was five times what it was 50 years earlier, but “the physical weight of our gross domestic product is evidently only modestly higher today than it was 50 or 100 years ago”.

It’s not just a US-centric pattern. In Australia the data shows manufacturing has shrunk from 14.3% of GDP in 1975 (that’s as far back as the ABS website goes) to 5.3% in 2024. In New Zealand, the equivalent data from Stats NZ shows a decline from 24.8% in 1972 to 8.7% in 2023. 

The modern economy still needs steel bars, but it’s now overwhelmingly an economy of stent implants and hip replacements, Taylor Swift concerts and Grand Final tickets, online education courses and social media. Good luck if you think you can resist that incoming tide.

Setting a low bar

It would be unfair to characterise manufacturing as on its last legs: even in our brave new e-world someone’s got to make the routers and mobile phones and solar panels we need, not to mention all the other ‘stuff’ from cars to syringes that have to come off an assembly line somewhere. There’s some real cutting-edge stuff happening within manufacturing: the first US$4 trillion company is one making fancy computer chips (Nvidia). But it’s a real stretch to imagine that manufacturing can magically be restored to where it used to be.

Even if it could, it mightn’t be all you hoped for. There’s a nostalgic view of manufacturing as a solid, well-paying gateway to a decent living for the whole family. But is that right? 

The latest data from the ABS (New Zealand patterns are very similar) shows the average weekly ordinary time earnings in manufacturing were A$1792. That’s not shabby: it’s better than some of the notoriously low-paid industries – ‘Would you like chips with that?’ (hotels and restaurants, A$1430) and ‘That looks lovely on you’ (the retail trade, A$1496). But chances are you’d be a bit better off working on a building site (construction, A$1883) or driving a truck (transportation, A$1906), and you’d be a lot better off heading for the mines (A$3112). 

Weekly earnings in manufacturing are also a bit lower than the A$1976 a week averaged across all sectors, which doesn’t make for a strong case that prioritising manufacturing jobs will make the country as a whole better off. If you were genuinely interested in creating more high-paid jobs, you’d be trying to figure out how we could get more people capable of being employed in IT and communications (A$2507), professional, scientific and industrial services (A$2320), or finance and insurance (A$2290).

A high price to pay

And finally, there are the costs you’ll run up if you take the Trump route of hoping that manufacturing will flourish behind tariff barriers. 

Protectionism is a terrible idea. It protects the privileged few at the expense of the rest of us. It raises prices. It penalises exporters when they need to use imported raw materials. It forfeits the gains from trading with the people who can produce things most efficiently. It creates dozy incumbents with less incentive to innovate. It depresses activity: makes life difficult for Korean microwave manufacturers and they’ll buy less of your wine. It invites a spiral of retaliation, and we saw how that played out in the 1930s’ Depression and how it may be playing out today.

To be fair, I’ve got a lot of sympathy for rust belt communities who have found themselves on the wrong page of the history book. I understand why people want to help them get back on their feet. The right way to do it is to have effective social security programs and to invest in the skills that people need to better make their way in a modern economy. A manufacturing fetish and counterproductive trade wars aren’t going to get them there.