Articles

Balancing the trade deficit is a must for manufacturing. Here's how to do it

By Bart Taylor, GHMA

In the past year, America’s deficit in manufactured goods rose to a record $1.2 trillion. Meaning, we need to import a trillion dollars worth of manufactured products to meet US demand. Every year.

The consequences are piling up.

The products we import aren’t free, meaning that capital is required to fund the acquisition of a trillion dollars worth of goods. “Whenever the United States gets an import from abroad, we're not just getting it for nothing,” says Ian Fletcher, co-author of Industrial Policy for the United States, Winning the Competition for Good Jobs and High Value Industries. “We're either sending back our own exports, like when we get Toyotas from Japan and we export Boeing aircraft. Or we're borrowing money from foreign countries, increasing our foreign debt to pay for all the stuff. Or we're selling off existing assets we own in return for these things that we import. So foreigners get millions of dollars of shares of Microsoft. Or they get partnership interests in a shopping center in Kansas City. You name it,” Fletcher told me in a recent interview. 

“The other thing that happens is that you lose industrial and technological capacity,” he says. “There are a lot of important things that aren't made in the U.S. anymore, which means we don't have the capability to produce them. It's not just that we don’t have the factory, the bricks and mortar. It’s not having the know-how. It's not having the assembled teams of people who understand how to do these difficult and important things. This also has huge implications for national security.”

“The (trade) deficit isn't just toys and shoes,” Fletcher’s co-author Marc Fasteau says. “The last I checked, we have a $297 billion deficit in advanced technology products. That's the stuff where we're supposed to be ahead,” he says. “That part of the deficit is also growing very rapidly. It was $200 billion when we were writing the book. Now it's $297 billion.”

I had called Fletcher and Fasteau to ask how balancing the trade deficit might be a 'true north' for pro-manufacturing “industrial policy.” The singular outcome? Make more and import less. So much has to happen. But it’s achievable. And the payoff for winning is huge – including millions of new US manufacturing jobs.

Not everyone agrees. Experts argue that a couple million new manufacturing jobs aren’t worth the trouble. Also, “industrial policy” is for many, a four-letter word. Free-market capitalism — laissez-faire economics — is mainstream. And policy remedies can be elusive; building consensus on anything, is flat out hard. 

“When we started writing (the book) most Americans barely knew what industrial policy was,” Fasteau told me. “Now the U.S. is pretty much being cornered into having an active industrial policy. It's nice to say, okay, let's just have a laissez-faire solution and let the market sort it out. But when you're an open economy that has tens of trillions of dollars of trade with economies that very much don't do that, doing nothing does not give you a free market, period,” Fasteau says. “It just gives you a market controlled by whatever the other guy's industrial policies are.”

Presidents Biden and Trump have used tariffs as a tool to regain control. I mention that tariffs may be harming the small businesses they were intended to help. “Tariffs have gotten a bad rap for the reasons you suggest,” Fasteau says. “The tariffs have been indiscriminate, even though, in fact, they have not been a major contributor to the kind of inflation that has driven voters nuts — and the real drivers aren't tariffed at all. It's easier to talk about how bad they are and not really look into the actual damage they've done. That said, and if I were running for office, I would say no more across-the-board tariffs. We're not tariffing the stuff that goes into gift shops. We're not tariffing coffee, things we can't make or that are too labor-intensive,” Fasteau added. 

Fletcher adds, “Other industries you want to protect because they're being very aggressively targeted by foreign countries. So if you don't have a tariff, the whole thing is going to get washed away by someone else. So that's what Biden was doing, I think, with the 100% tariff on electric vehicles. And then you’ve got to look at industries that you want to protect because they represent our economic future. It makes a difference whether you protect the production of cornflakes as opposed to microchips, because without microchips, you don't have an economic future — the technological future that we're counting on.” 

The policy tool Fletcher and Fasteau favor most is intervention to lower the value of the dollar. It’s somewhat controversial, in my view, as many of us equate a strong dollar to overall US strength. It’s also wonky, which makes it a hard sell.

More, America’s “finance economy” wants no part of it. 

“What's supposed to happen in free trade and free-floating currencies is when you have a big trade deficit, your currency is supposed to depreciate, which sucks!” Fasteau laughs. “But it makes your exports cheaper and imports more expensive to you. And that's the way trade is supposed to balance. It hasn't happened since 1973 for the United States.”

Fletcher adds, “The first thing you've got to do is move the currency into a trade-balancing range. Because if you have a currency that's overvalued at something between 15 and 20 percent, and if you have an average U.S. tariff which is something between 11 and 14 percent, you don't even have a net tariff at all. And the key thing is, if you bring your currency to a trade-balancing level where exports equal your imports, well, that's the point at which you can start asking the real strategic question: which industries do you want to single out for protection as opposed to others? Now, we know for a start, that some industries have to be protected because they're relevant to national security. If the United States does not produce steel and aluminum, it is not a superpower. End of question,” he says.

Fasteau adds, “The other thing I would say on currency, taking public indifference out of the question, is ‘whose ox would get gored’ if you did this? Basically, the only policy that's serious about bringing the currency down is moderate, variable controls on international flows of capital. Wall Street is going to hate that because their entire business model is taking a piece of money and moving it around the world to where it's going to get what they expect is going to be the highest return,” he says.

“So you're going to hurt the financial interests. You're going to benefit the ‘real’ economy, particularly everybody who exports or competes against imports. So right now, what you're looking at is a political correlation of forces in which the finance side of the economy is still stronger in its influence over the White House and Congress than the real side of the economy. And that is really what I think would have to change,” he explains. 

“Here's another thing that would have to change. The easiest way for President Trump to get investment into our ‘real’ economy seems to be to go overseas, use tariffs, and get other countries to invest. A $500 billion pledge from Japan, Korea, and so on,” Fasteau observes.

“Fine, but why? There's no shortage of capital in the United States. We seem to have two issues. First, we've had a system with cheap imports, subsidized imports, and not just from China, but our friends have industrial policies too. They're not quite as nasty, but they've been extremely effective.

“That has reduced the opportunities for investors and companies to invest more in the ‘real’ economy and make a profit. And making that even worse is the ‘short-termism’ — the over-financialization of our economy — which goes back to probably its most critical element, and that is shareholder primacy, where shareholders do best and ‘forget the other stakeholders’ if the stock goes up,” Fasteau says.

Would two million new manufacturing jobs change America’s employment landscape? Some say no. Would rebuilding America into a creditor nation in trade, change our global standing? If you value manufacturing it would. 

Let’s balance the trade deficit — any way we can.

The Texas Surplus Dwindles

Texas product exports account for roughly 21 percent of the US total — far and away the most of any state and more than the next three (California, New York, and Louisiana), combined. 

As a result the state runs a trade surplus in most years, fueled mainly by Houston’s powerhouse petrochemical industry. More, Texas accounted for 23.4% of all U.S. computer/electronics export value in 2025. It’s a key industry sector in Texas and Houston that’s only getting bigger. 

The story has changed through June 2026. Texas moved from a surplus early in the year to monthly deficits beginning in April, with June's deficit reaching about $1.3 billion. The main driver has been, ironically, the data center and high-tech boom, and the surge in imports of tech-related hardware.

Here’s a trade snapshot:

Texas industry                          2026 trade position                    What is happening

Petroleum & coal products       Large surplus                              Still one of Texas' biggest net-export engines

Chemicals                                 Large surplus                              Strong Gulf Coast production/export base

Machinery                                 Surplus                                        Oilfield, industrial and other machinery exports remain substantial

Aerospace/transportation         Surplus                                        Strong export orientation

Fabricated metals                     Likely surplus                              Significant industrial export base

Food & agricultural products    Surplus                                       Texas remains a major agricultural exporter

Computers & electronics          Deficit / import-intensive             Huge exports, but imports of components/equipment 

Electrical equipment                 Deficit / import-intensive             Significant inbound supply chain

Motor vehicles & parts              Deficit                                         Texas exports vehicles but imports substantial components/vehicles

Pharmaceuticals/medical          Deficit                                          Import-heavy relative to Texas production

Consumer goods                      Deficit                                          Heavy import exposure

Bart Taylor is president of the Greater Houston Manufacturing Association. Reach him at [email protected]. 

AI agents and more: How tech DNA powers Piping Technology & Products

CEO Rakesh Agrawal is equal parts technologist and industrialist — and having a blast leading 700 employees at Houston’s stalwart manufacturer

By Bart Taylor

Rakesh Agrawal leads the company his father founded in the mid-1970’s, “growing up in the business on my dad’s knee,” he told me in a recent conversation. 

Rakesh’s dad, Dr. Durga Agrawal, parlayed an interest in designing piping systems to a side hustle while earning a PhD at the University of Houston, becoming a self-taught expert in pipe stress analysis. “He’d go out to Shell Deer Park and other chemical plants along the ship channel and, well, help them solve piping problems,” Agrawal says. “People started asking for more. ‘You’re recommending we put these pipe supports or expansion joints into the piping system. Would you supply those to us?’” he recounts with a smile. “So we eventually started our own manufacturing.”

Good thing for Houston. Today the company boasts a workforce of over 700 employees, 400-plus in manufacturing, in 500,000 square feet of covered shop space with 75,000 more on the way as part of a major expansion. The company is building on its 40-acre site at the junction of Highways 288 and 610 – notable by a colorful GONZO247 mural that drivers see every day.

But it’s also the talent mix at PT&P that’s noteworthy – shaped by its tech DNA. “Everything is engineered to order, built to order,” Agrawal says. “We do what could be called ‘mass customization’, where we make a quantity of product, but all different based on the customer’s application. So half of the work we do is engineering, and the other half we do is manufacturing,” he explains. “We’re a mix of knowledge workers – programmers, engineers, buyers and more – and craftspeople – welders and technicians that operate various CNC machines or forklifts.” All arrayed to design and manufacture products to solve piping system challenges like thermal expansion and contraction, vibration control, and energy containment, in a mass customization model. 

The markets PT&P serves are self-evident – but changing. “I think there was a time when oil and gas was 70 or 80% of our business, but we’ve always been diversified. It’s how we survived the crash in oil and gas in the ‘80s while many did not,” he says. “In the last few years, we’ve expanded to selling to mechanical contractors as we do a great deal more work in data centers, pharmaceuticals, and automotive. Data centers, in particular, are a big driver in our business right now.” 

What’s also noteworthy is how the company is navigating a technology journey that for many, can be a challenge. PT&P is a manufacturer — but also a tech lab, including the use of AI agents.

Agrawal’s background is a catalyst. "I studied mechanical engineering and computer science, and got two degrees at Rice University. Mechanical engineering is where I grew up, right?, the son of an engineer. When we had a flat tire, it was like, ‘oh, we're gonna do a soap and bubble test on the tube’,” he laughs. So I grew up around the practice of mechanical engineering, and computers were my hobby, and I was very lucky to have parents who indulged my curiosity in computers. There was always a new computer at the house,” Agrawal says.

“And even at the business, we were always employing technology. I had just finished college, and my first job was building the company's software systems. We're in a business where traditional ERP doesn't really suit us very well because of the mass customization thing I mentioned. ERP likes it when you're making the same thing over and over again – for example, if you're making a million AirPods and they're all only slightly different from one another,” he says. “But in our business you lose your mind trying to use a traditional ERP. So I hired a few friends from Rice, and we built a custom ERP from scratch – and it became the technology core of the company.”

Agrawal actually left the company for a time to scratch his technology itch. “I started a software company, SnapStream, and, as an investor, I built a portfolio of software. SnapStream enables media companies like New York Times, the Daily Show and Colbert, before they went off the air, to source all their short video clips,” he says. 

But I had called Agrawal to chat about an AI case study I’d read — at Tasklet that featured PT&P — where his team is using AI agents to enhance productivity. “I started a group within the company two years ago to attempt to drive broader usage of AI at the company. We paid for people’s ChatGPT accounts, and there were good ideas and people were inspired by one another. But it kind of ‘hit a wall’ in terms of broader usage,” he says. 

“Then I was introduced to this AI agent technology, Tasklet, through a venture fund I’m an investor in – and they had automated their entire backoffice using it,” Agrawal says. 

Tasklet says its “AI command center…takes ownership of work: it sends emails, updates spreadsheets, files tickets, builds apps, and runs recurring processes across your tools. Describe the outcome in plain English, and Tasklet figures out the rest.” Its use cases span the operational spectrum of any business, it seems, including shop-floor employees at PT&P.

For his part, Agrawal uses a personal "Chief of Staff" agent connected to 22 services — Gmail, Google Analytics, Google Chat, calendars, and others. 

It’s a leap of faith I haven’t yet taken. Agrawal can relate, to a degree. “People have old information,” he says. “They read a headline where OpenClaw came out earlier this year and it drained someone’s bank account. The fact is that these agents have guardrails on them. Sure, they can make mistakes — but generally the mistakes are pretty bounded.”

His employees provide much-needed context. “One of the people on my team — shoutout to Fiona — calls these AI agents ‘genius math babies’,” he laughs. “There's this weird dynamic where AI is really good at just, like, ridiculously hard things, but then really dumb when it comes to really easy things. Genius math babies. We need to learn how to use them to get work done accurately, to be effective but not drain our bank accounts!” 

For Agrawal, his people also provide the ‘true north’ for his company’s tech roadmap. “I kind of think about AI agents as a new productivity primitive. It's like a spreadsheet or a word processor, and I want us to hurry up and learn how to use word processors and spreadsheets so we can all be more productive.”  

Agrawal also acknowledges that the technology he’s harnessing may profoundly impact his workforce – every workforce. I get around to asking what impact this technology will have on jobs. “The job’s question is a hot topic when you talk to people in the AI world. Nobel Prize-winning laureates are apart on the issue,” he says. 

“My personal view is that, yes, at some point it's going to be pretty bad and weird, at least before it gets better. And I see it as my job as the CEO of this company to prepare our workforce and our company for whatever’s next, that next level where new jobs will be created — and get in front of it.”

Read PT&P’s Tasklet use-case study here. We’ll also circle back with Rakesh Agrawal and Houston’s powerhouse manufacturer in later editions. There’s a lot more.

Bart Taylor is president of the Greater Houston Manufacturers Association. Reach him at [email protected].