Here is a statistic that tells you much of what you need to know about the United States trade debate - only 1 per cent of US companies export.
It is a figure that I uncovered recently in a McKinsey Global Institute (MGI) report on the state of US manufacturing, which confirmed much of what my own reporting has informed me about US trade problems.
They begin at home, not abroad. That is not a line you will hear over the next few days, as negotiators meet in Mexico for the latest round of North American Free Trade Agreement (Nafta) talks.
The Trump administration is likely to point fingers and threaten again to pull out of the whole agreement. But if the White House wants to fix trade, it would do better to look at three crucial mistakes that have been made at home - and then compare them with opposite strategies in China, the world's biggest manufacturing country.
The first mistake was to starve manufacturing of investment. Private sector investment in the sector in the US is at 30-year lows; the ages of the average factory and piece of machinery are 25 and nine years, respectively.
The MGI report estimates that US$115 billion (S$156 billion) of investment would be required annually over the next decade to fix this problem.
But trade and tax policy in the US does almost nothing to provide an incentive for that investment. Companies talk a good game about repatriation, but privately admit that the bulk of any money brought home will go into share buybacks and dividend payments.
Mr Gary Cohn, President Donald Trump's economic adviser, last week appeared surprised (or perhaps feigned surprise) as three hands went up when, at a conference, chief executives were asked how many would invest at home if tax reform were passed.
Why should they, when there are no quid pro quos of repatriation in exchange for equipment investment, training or the funding of an infrastructure bank?
As the past two decades have shown, supply side economics alone does not result in a productive use of capital, particularly when it comes to global multinationals; sometimes, you have to push investment to the areas where it is needed. That is exactly what China is doing with its US$1 trillion infrastructure investment plan.
Second, the US trade debate has focused far too much on the largest domestic companies, which are doing quite well, in terms of share price, profit margins and sales growth. Yet the vast majority of the 250,000 manufacturing groups in the US employ fewer than 100 people.
The current zero-sum game global trade debate hardly applies to them at all. Most do not export, not because they cannot but because access to capital has been tighter for them relative to their rich country peers since the Great Recession (that lack of investment creates a 40 per cent productivity gap between small and large businesses). Reconnecting the dots between the largest exporters and the supply chain could fix that.
Regionalisation is the new globalisation - the largest firms in the US should be given explicit government incentives to work with local suppliers, which would do much to address the President's trade deficit obsession, since 70 to 80 per cent of the value of a finished product is in the supply chain.
Again, this is something that China has prioritised in its 2025 development plan, which states explicitly that the country wants to minimise dependence on foreign markets and technologies.
Finally, the US desperately needs to connect Silicon Valley and the rust belt.
High-tech manufacturing - 3D printing, the Internet of Things, sensors in products that create business opportunities in data analytics and services, rather than just goods - may well be the future. Yet, shockingly, roughly half of US manufacturing companies have no digital strategy whatsoever, according to an MGI survey.
Meanwhile, the technocrats in the Chinese Communist Party are giving speeches about the integration of the physical and digital worlds that would leave many people in US boardrooms scratching their heads.
Of course, it is easier to do economic-dot connecting in an autocracy. No one is arguing for that, but it does speak to the core problem - the Trump administration is trying to push "America First" without any clear and coherent plan of what that means, or how to make an industrial strategy succeed.
"You look at the current trade debate and you just think, 'What a lost opportunity'," says MGI director James Manyika.
The countries that the US left behind when the administration pulled out of the Trans-Pacific Partnership - high-growth emerging market nations with expanding middle classes that are engaged in major infrastructure building works - are exactly those with the largest demand for the US' most valuable exports, such as high-tech products, pharmaceuticals and heavy machinery.
Pulling out of Nafta, particularly without any kind of home-grown industrial policy, would be an equally big mistake. Economic nationalism with no economic plan will not make America great again. In fact, it may be worse than the laissez-faire alternative.