aggregatejeff - Tumblr blog

aggregatejeff · 3 months ago

Text

How A Deeply Flawed Free Trade Policy Cost the US the Drone Wars, Reduced Our Ability To Innovate, And Threatened Our National Security, And What We Can Do About It

While I was in college, and for many years thereafter, the faith in free trade was near absolute. Those like Ross Perot, who opposed NAFTA or others who raised objections to similar agreements, were ridiculed as simpletons with little understanding of economic theory. Nobel winning economists such as Paul Krugman (who would later at least admit to being wrong) and Milton Friedman, though ideologically quite different, were steadfastly aligned that globalization and free trade was the best possible path forward. Despite my continued education, I remained in the simpleton minority. I understood the arguments proponents made for free trade, and certainly could see how it would result in cheaper goods, but the classic Adam Smith style analysis that a) assumed past data would continue to be predictive, b) was based on an even playing field between parties, and most problematically, c) ignored the negative externalities that resulted from specialization. I was, admittedly, relatively unschooled in higher-level economics, but as it turned out, my lack of knowledge was the thing that made my intuition correct. I wasn’t blinded by adherence to a philosophy, and relied on a filter of common sense vs. applied theory. Which is why I, plus nearly every factory worker in the US, could clearly see what would occur, and your average academic, or even trained policy wonk, could not.

The (Flawed) Implementation of Free Trade

At a purely theoretical level, free trade was, and continues to be beneficial policy. If country A is more efficient at producing wheat, and country B is more efficient at producing flour, then free trade benefits both. Country A produces all the wheat, country B produces all the flour, and everyone gets cheaper goods. Granted, the wheat producers in country B will lose their jobs (what economist and politicians might call “short term disruptions”), but labor is fungible in the medium to long term, and less productive (thus less competitive) employees in both countries would move to other, more productive industries. Apart from the obvious issue that “short term disruptions” meant someone lost their livelihood, and likely couldn’t just shift to a new industry after 20 years in a prior one, for most trade between countries, the theory held true, and strong data existed to back it up. Because of the evidence, academics, and those educated by them, strongly advocated for free trade agreements. But while the high-level theory made sense, details matter, and there were three underlying assumptions that were deeply flawed.

Flaw 1 - past results do not guarantee future outcomes. Free markets are enormously powerful, and generally do guide workers to areas of maximum productivity and return. While the Econ 101 example above does fall apart at some point due to a lack of infinite demand (I.E., no matter how cheaply a country can manufacture flower, at some point people don’t want more of it), so all of Country B’s former wheat growers probably can’t get jobs making flour, it was generally assumed that the invisible hand would guide those workers into new, competitive roles. That sounded nice, but while it might have been true in the 1900s, it seemed an increasingly tenuous assumption due to the requirements of the type of work in which the US could both compete and maintain its standard of living in the face of highly abundant, low-cost unskilled labor. Put another way, while at one point nearly every farmer could just as easily be a factory worker, it wasn’t realistic to assume every factory worker could just as easily become a software engineer. Though no one could publicly admit it, the areas where the US was competitive (high-level knowledge work) could only be performed by a limited percentage of the US populace. So what do we do with the people who aren’t able to be employed in the areas where we now specialize? There were really only two options: a) factory wages in the US fall to the level at which they’re competitive, or b) the non-competitive labor force goes to work in service jobs that can’t be outsourced. Both aren’t positive for a large percentage of the population, which should have made people question the logic of truly free trade, but the bigger problem was…..

Flaw 2 - everyone wasn’t playing by the same rules. Put simply, it made no sense to have strict environmental and labor standards, yet allow capital and goods to flow unfettered to countries that do not. Holding everything else equal, there simply would be no way for countries with standards to compete. It was going into a fight with one hand (or two) tied behind our back. There’s no way production in a country that has strong environmental regulation would be as inexpensive as production in a country without such laws, because compliance is expensive. It’s no different than saying “Hey, we’re going to put a 30% tax on everything you do here, but if you want to build outside the country, you won’t be taxed.” We were literally incentivizing companies to move production abroad. While politicians could talk about protecting labor rights or cleaning up the environment, what they were actually doing was sending the problems of pollution, child labor, et al. abroad, while fundamentally handicapping the ability of the US to be competitive. This was (and still is) the most glaringly obvious problem with free trade orthodoxy; it requires no math, no assumptions, no predictions, just basic if X > Y logic. It’s also the easiest free trade issue to solve, both because of the simplicity problem, as well as the obviousness of the solution. There are two clear options to address this problem: change the restriction (I.E. remove environmental, overtime, minimum wage, child labor laws, etc.), or place normalization tariffs in place (I.E, if doing business in a country with labor and environmental laws is 30% more expensive, then place a 30% tariff on any goods coming in from countries that lack such laws). The first option wasn’t politically feasible (turns out removing the minimum wage, allowing subsistence living (along with removing housing standards), company-dorms, 80 hour work weeks with no overtime or time off, and all environmental standards is frowned upon by the public), but the second option was far more palatable and should have been implemented. However, option 2 suffered from the achilles heel of corporate distaste. Less expensive production paired with stable pricing (for a variety of reasons, prices are sticky, and companies don’t pass on the lower cost of goods sold in the short to medium term) increases profits, which increases share price. While there might be long-term harm from this time of outsourcing to the company, most executives are compensated on quarterly and yearly time horizons, and rarely are concerned with effects more than a few years out. The corporate community was thus on the whole highly opposed to tariffs, no action was taken to equalize production costs, and offshoring of any labor intensive action that could be was all but assured.

But wait, there’s more. It gets worse. It was widely known that true free trade could be destructive to developing countries. It was so widely known that the ability for developing countries to engage in open trade with partners while protecting and or subsidizing domestic industry is literally written into the WTO rules. Everyone acknowledged that a country without an industry, no matter how industrious, how educated, would have difficulty competing with an established, high-volume producer, because the advantages and economies of scale the high volume producer enjoyed would be impossible to compete with in the near term. And they also acknowledged that private industry is unlikely to devote enough funding to build a competitor in that developing country, because even if successful, they’re emerging into an industry against multiple well-capitalized competitors who have already pushed pricing (an thus profits per unit sold) down, making future profits insufficient to justify the expense. Put another way, scaled manufacturers have already made large investments in equipment, training, logistics, and other areas. Over time, competition against those who have already made the same investments forces prices to get close to the marginal cost of production. Market prices thus become too low for profit per unit to cover the cost of starting from scratch. This is *exactly* the issue I’ll get to later on drones, and also why there are so few new manufacturers in most established industries (cars, appliances, planes, semiconductors, etc.) in any given country (also also why new manufacturers emerge when there’s a step-change in the cost of production).

So the WTO allows developing countries to play by different rules, closing their markets off while enjoying low or no tariff trade with their developed partners. Free trade absolutists, who don’t like complexity with their coffee, think these sorts of one-way trade agreements ultimately harm the developing country, because any type of market intervention leads to a sub-par outcome, but the world’s second largest economy, China, built said second largest economy on top of these one-way agreements, so the idea that they lead to a sub-par outcome is more than a little questionable (Fun fact - as of 2025 China *still* classifies themselves with the WTO as a developing nation). Nor do these barriers have to be limited to tariffs. As an example, for decades any US company wishing to enter the Chinese market was required to a) do a joint venture with a Chinese company and manufacture said goods in China, and b) transfer a significant amount of their IP to that joint venture (and the Chinese partner). This leads to what Noah Smith calls the China cycle, whereby a foreign company enters the Chinese market, funds and educates a domestic Chinese competitor, and eventually gets replaced by that competitor (see more or less every auto, computer, electronics, appliance, or any other complex good manufacturer). And this still goes on today. Put more succinctly, we *knew* free trade could be detrimental without protections, and pushed ahead anyway. I can’t admit to knowing how this would play out in the early 2000s, but by 2007 or 2008, the trend line was obvious, as was the fact that…

Flaw 3 - Low prices could have very high costs. Much of the free trade orthodoxy assumed that lower prices were the ultimate goal, without factoring in the costs to the environment, long-term competitiveness, or national security. We covered the human rights and environmental concerns, and the willful ignorance of simply offshoring one’s problems. Under free trade agreements, the US was able to gain large volumes of smugness, while child labor outsourced to the sweatshops of hard to pronounce cities, and pollution was dumped into foreign rivers. We could scold the world for employing minors or fouling the air, but the total number of children employed and pollution emitted only increased.

Even more problematically, manufacturing isn’t a process that can be easily turned on and off. It requires physical, monetary, and human capital. Modern manufacturing is also deeply dependent on complex supply chains. Each finished good has hundreds of components, and the speed, integration, and accessibility of those component suppliers’ drastically impact the relative ability of an area to produce competitive goods. Even within countries, regions had distinct manufacturing clusters, both with regard to industry as well as functionality. Over time, these clusters have moved or replicated (computers from Boston -> Silicon Valley, furniture from New England to the Carolinas, entertainment from New York -> Los Angeles), and the cycle has played out enough for it to be known. A major player or two move, and soon there’s not enough business for suppliers, who close or move, and an entire region’s economy disappears. Within a country, labor mobility shifts human capital to the new location, and equipment can easily be transported. Suppliers spring up around the anchor, and the economic engine hums. At a country level however, once a manufacturing capability disappears, any attempt to revive it is far more complex and prolonged. Employees need to be trained from scratch, as there’s no base to draw upon. Entire supply chains need to be created in what are effectively economic vacuums, as there isn’t enough demand for the final manufactured goods to support the supply chain. Any existing physical plants become increasingly outdated. Starting a manufacturing base from scratch takes an extended, coordinated, and concerted effort, usually with government support, hence the WTO rules mentioned above. Given that the US government hadn’t done extended, coordinated, and concerted anything in 60 years, this implied that once we lost manufacturing capacity, we weren’t getting it back, which could have profound implications for both our long-term competitiveness and national security.

The tacit assumption of policy makers from the late 1990s to the mid 2010s was that under a global free trade regime, the US would become the world’s R&D hub. Our strength lay in producing research scientists and software engineers, and it wasn’t an issue if low-skilled production moved off-shore. Hence the rise of “Designed in California, Made in China”, and the push to get everyone to go to college. However, this ignored the tight link between innovation and production for any advanced manufactured good; that how something is made, and the components that go into it, are often as important as any innovation or invention itself. There are number of interwoven threads that tie innovation and production together. They’re worthy of a long post by themselves, but in summary:

1. It’s difficult if not impossible to keep most IP from the people producing it.

2. The more that IP leaks from the manufacturer to the surrounding industry, the faster copy cats emerge, and the shorter the time a company has to recoup R&D investment.

3. Iteration on components and production method is often as, if not more important to effective innovation as any conceptual ideation.

Combined, this means that the ability of a country to be an innovation hub without also controlling the means of production (or at minimum ensuring production is in a place with very strict IP protections) is time limited, and trade policies that compromise this ability compromise the ability of the US to be competitive even in specialized areas.

Similarly, our national security was also at risk from the same policies. A country lacking an ability to produce defensive tools and equipment, does not have a strong defensive ability. Full stop. Do not pass go. There is no security without production. This wasn’t exactly a surprise; maintaining our own food supply was a large part of why we have farming subsidies, and we have domestic defense contractors for a reason. It’s easy to say that people forget the lessons of WWII, where we nationalized so much of our industrial base and that our defense production ability more or less won the Second World War, but I don’t think it’s true that it was forgotten. Rather, it was waived away by a group who felt that large-scale conflicts were “passe” and unlikely to occur given modern tools of war. Further, they felt that with sufficient free trade between countries, the likelihood of that conflict was greatly diminished. I did not agree.

What Happened?

There was a decrease in the cost of goods due to cheaper labor, and an effective collapse of new manufacturing in the US. China, unlike most of the rest of the world, wisely took advantage of the playing field it was given, and due to the allure of its large domestic market, demanded the transfer of a significant chunk of US & European IP in return for access. The core systemic weakness in most of the West - that company leadership is motivated by short-term gains, and politicians are motivated by donors from company leadership, exacerbated this cycle. Because of how production and innovation are linked (particularly when IP leakage isn’t well controlled), the US and Europe lost many of the comparative advantages that accumulated from the knowledge industries it was supposed to be more efficient or better at, leaving it both without a manufacturing capability and with significantly reduced international competitiveness.

The evidence for the latter claim is undeniable and staggering. Innovation in physical goods has drastically dropped in the US over the last twenty years, and has grown dramatically in China. Of the new large-scale manufactured products categories that emerged since 2000, China now dominates or is on the way to dominating every single one. Household robotics? 3D printers? Electric (and non-electric) vehicles? Even something with substantial software lock-in such as smartphones are not immune. All of these innovations started in the US, and now the US is at best a competitive player, and in more cases, a laggard. Environmental objections were permitted to strangle any domestic production of critical minerals and materials, while indirectly supporting the building of that capacity in our strategic adversaries. Our defensive posture was drastically weakened over the last 30 years, and we’re seeing the fruits of those decisions (or lack thereof). We no longer have the ability to produce multiple rare earth elements in anything near the capacity we need for chips, electric systems, or most any technology, and are completely subject to China’s whims on materials. China now has a near monopoly on many critical areas, and behaves like it. Any attempt to start a new company in any of those critical areas outside of China is crushed, just as a regular monopoly does, by a temporary lowering of prices to make any competition economically unfeasible. As a result, our ability to produce the equipment needed for modern warfare, such as planes, drones, unmanned naval vehicles, and even basic items like artillery shells, can’t even keep up with peacetime needs, let alone any future conflict. In the past, we used the Defense Production Act to rapidly increase our ability to manufacture defensive and offensive equipment, but the DPA relies on their being a manufacturing base to reuse and repurpose. We’ve lost much of this, a reality we saw staring us in the face with the simple task of producing masks during COVID, and still, nothing was done.

Drones as a Canary

Which brings me to drones, as perhaps the best and most painful example in the group. Not only did we lose the industry, but it’s abundantly clear that it’s the future of warfare and that the US’s lack of domestic capability is more than just a black eye; it’s a life-or-death strategic risk.

Drones, at least as we know them today, largely started in the US, but today nearly 100% of the market (by both unit sales and dollar value) is manufactured in China by Chinese companies. Not only were the core technologies for drones pioneered in the US (covered well by the MIT Technology Review), but drones themselves emerged out of a primarily American hobbyist community who started using mobile-phone-based IMUs and communication sets to power quadcopter. U.S. companies like Dragonfly kicked off the industry with drone kits. US-based 3D Robotics formed in 2009 with substantial venture backing, focused on building consumer drones, and French company Parrot launched the first commercial “ready-to-fly” drone in 2010. DJI meanwhile started in 2006 as a company selling RC helicopter components, and was still doing so in 2009 . Sometime after that (public accounts, at least in English, are very sparse), DJI shifted to manufacturing drones, and in 2013 launched the Phantom 1, a fully-integrated consumer drone. 3D Robotics continued to launch products, with 2015 featuring a face-off between 3D Robotics flagship“Solo” drone vs DJI’s Phantom 3. Unlike 3DR, DJI was fully vertically integrated and able to capitalize on the Shenzhen-based supply chain, with easy, inexpensive access to nearly every drone component. 3D Robotics had to either design and source their parts from China (as little to no domestic supply chain existed), or manufacture those parts themselves. The former would result in near-immediate IP leakage (I.E., any innovation 3D Robotics created would quickly walk out the back of the factory and spawn copycats), or manufacture their own parts in the US or Mexico at much higher costs. The result was a disaster for 3D Robotics, who couldn’t compete on price. “While a Solo with a gimbal and GoPro camera cost more than $1,700, the vertically integrated Chinese company, which controlled its own factories and already sold its comparable Phantom 3 Professional package for $1,300, aggressively slashed prices. By 2016, that Phantom with a gimbal and camera cost $1,000.”

There’s some question of how DJI was able to price compete so aggressively. Undoubtedly, it was significantly advantaged by a robust domestic supply chain, low-cost labor, and the previously discussed free trade policies that actively disadvantaged US manufacturers. This alone would likely have been enough for long-term success, but DJI’s, and China’s, position in drone manufacturing was solidly cemented through China's aggressive industrial policies, specifically under its "Made in China 2025" initiative, with combined both explicit and implicit state subsidies totaling hundreds of billions of dollars to drive both manufacturing scale and scale and overcapacity. The US Intelligence community called out DJI’s price cuts as dumping (the deliberate use of lower-than-cost pricing to drive competition out of business) without explicitly stating how DJI was able to drive prices so low. Regardless of the cause, the net result was that by 2017, there were effectively no US-based competitors to the Chinese drone industry, except for a few small players who focused on defense-related contracts. Benefitting from Wright's Law (technological costs decrease as production scales up), Chinese manufacturers were “encouraged” to produce millions of drone components, moving their production rapidly down the cost curve, and making it near impossible for any non-Chinese manufacturer to compete.

The result is a market distorted to the point that the US can’t compete without government intervention. The overwhelming market presence of subsidized Chinese drones means that most private sector investors simply won’t invest in a US-based drone company, as there’s no viable path to establishing a competitive product with a sufficient profit margin to recoup their investment. Put another way, even if someone were to risk the billions of dollars it would take to build an American drone company capable of outputting products at the scale of DJI, the pricing needed to be competitive would be so low as to make profits minimal or negative, and thus no sane investor would put their capital at risk. Similarly, many in the drone industry fail to understand the core issues, and say something like “the problem is that US companies aren’t innovative”. They fall victim to the genius inventor fallacy, and don’t think through the production required to capitalize on that innovation. Without a domestic supply chain, that innovative product gets manufactured in China, where the IP gets copied, giving a Chinese player like DJI the opportunity to reverse engineer it, and bring a similar product to market at a lower cost. The innovative company thus only has a few years at most to recoup any investment, which is generally not enough to justify the investment in the innovation in the first place. The core problem is not the lack of innovation, but the lack of incentive to innovate. Even where domestic drone companies have survived through defense and security-related contracts, they’re subject to the whims of the Chinese government, who can and did restrict the battery supply of Skydio, one of the remaining US-based drone manufacturers, to limit their ability to compete.

The same dynamic has played out in every emerging manufactured product, from drones to batteries to smartphones to chips and now to electric vehicles. For all of these, the lack of scaled US-based manufacturing reverberates throughout the economy. For every missing end-use product, 100 suppliers also fail to scale, and thus no effective competition can emerge in the US or anywhere outside of China. Only in a few select industries, such as software or pharmaceuticals, where manufacturing has little to do with the value chain, has the US remained dominant.

So what do we do?

Which leads us to the core question - how to we fix this mess? The obvious answer is that we need to resume domestic manufacturing, both for the sake of our long-term economic future and our national security. The how, like the initial problem, doesn’t have a single solution. Are Trump’s tariffs part of the answer? Likely. They would have been much more effective when the US was a larger share of the world’s market, but strategically applied the domestic US market is still large enough to spur investment. The biggest problem with tariffs is that in most cases, they’re far too small. A 100% tariff on a good produced at scale will still leave that good dramatically less expensive than one produced in small runs. That’s the power of Wright’s law. A product that might cost $15,000 to produce at a prototype level would likely cost a few hundred dollars at a scale of millions. It might be enough to convince investors that it makes sense to commit sufficient dollars to bring the small run US manufacturer to scale, and thus spur viable domestic competition. Other answers? Stronger enforcement of IP protections. Much faster anti-dumping rulings. Outright product / category bans. Industrial automation to address labor costs and availability gaps. A renewed focus on Federal R&D funding to deliver more upstream innovation. Direct incentivization of production (this is the only response if we want to keep costs the same as they are today) ala the Chips Act. But what’s really necessary is a longevity of policy driven from a common awareness of the problem. All of the above solutions are largely useless if they only last until the next administration. We need long-term policy to convince the private sector that their investments in US manufacturing won’t get wiped out in four years when someone decides lower prices are the way to a voter’s heart.

Fortunately, there seems to have been an awakening in Washington, at least among a select few that our prior system wasn’t, and continues not to be, working in our favor. Trumps election is in part a rejection of that system. Yes, stocks and corporate profits surged over the period, but at the cost of our collective future. Do we continue to mine our past for the last vestiges of profit while hollowing out the shell of our economy, or do we try to move the ship back to the right tack, and take a long-term, multi-pronged approach?

I hope, and will work on ensuring, we do the latter.

0 notes