China Drops Below Vietnam in US Apparel Sourcing. Near-Shoring Still Can't Fill the Gap

Twelve percent. That's how many US fashion companies still source more than 30% of their apparel from China, down from 22% in 2024 and 19% in 2025. Trade data now puts China behind both Vietnam and Bangladesh in total US apparel import share—a first in decades.

A sourcing director I know at a mid-tier brand in Los Angeles put it bluntly last month: "We spent ten years saying we'd diversify out of China. This year, it actually happened. And nobody's got a clean Plan B."

The 2026 USFIA Fashion Industry Benchmarking Study, conducted by Dr. Sheng Lu and Emilie Delaye at the University of Delaware in collaboration with the United States Fashion Industry Association, surveyed executives from 30 leading US fashion companies between April and June 2026. Roughly 80% were retailers, 65% brands, and about 55% importers or wholesalers. Nearly 80% reported having more than 1,000 employees. These aren't small players guessing at trends—they're the people who write purchase orders worth hundreds of millions of dollars. And what they're saying should worry anyone who assumed the post-China sourcing playbook was settled.

The headline number is stark: China's utilization rate among surveyed companies dropped from 100% in 2025 to 73% in 2026. But here's the part most coverage misses—Vietnam and Bangladesh didn't pick up the slack. Vietnam's utilization rate fell from 100% to roughly 78%. Bangladesh dropped from 88% to around 78%. The orders didn't shift. They scattered.

The Numbers Don't Lie—But They Don't Tell the Whole Story Either

Let's get the data on the table first, because without it we're just trading opinions. The 2026 USFIA study tracks eleven of the most-utilized apparel sourcing bases. Eight are in Asia. That part hasn't changed. What has changed is the intensity of use across the board.

China went from universal utilization—100% of surveyed companies placing orders there—to 73%. That's not a gradual taper. That's a cliff. Vietnam and Bangladesh, the two countries most analysts assumed would absorb China's lost share, both saw utilization declines instead of gains. Vietnam slipped from universal to roughly 78%. Bangladesh went from 88% to roughly 78%. Meanwhile, three non-Asian countries cracked the top ten for the first time: Guatemala, Egypt, and Jordan. All three recorded higher utilization rates than the previous year.

So orders didn't migrate from China to the next cheapest Asian factory. They fragmented across a wider map. That's the real story in this data—and it's more complicated than most sourcing directors are ready to admit.

I've spent enough time in factories across Zhejiang, the Mekong Delta, and Guatemala's highlands to know one thing: when sourcing patterns shift this fast, the supply base can't keep up. Factories don't materialize because a brand's risk committee decided to diversify. They take years to build—if you can find the skilled labor, the reliable logistics, and the fabric supply chain to feed them.

Why Vietnam and Bangladesh Didn't Win

The assumption that "leaving China" equals "going to Vietnam or Bangladesh" was always too neat. Both countries have hard ceilings.

Vietnam's biggest constraint is fabric. The country's garment factories can sew almost anything, but roughly 60-70% of the fabric they use is imported—mostly from China, Taiwan, and Korea. When you shift cut-and-sew operations to Vietnam while leaving the textile upstream in China, you haven't really de-risked. You've just added a cross-border logistics dependency that crumbles the moment there's a port disruption or a new tariff on Chinese inputs. I've seen this play out in real time: a brand moves jacket production to Ho Chi Minh City, only to discover the shell fabric still comes from a mill in Shaoxing. The "Made in Vietnam" label is accurate. The supply chain is still half-Chinese.

Bangladesh has a different problem. Its strength is basic cotton knitwear—T-shirts, polo shirts, basic sweaters. The infrastructure for complex outerwear, tailored garments, or technical sportswear with bonded seams and waterproof membranes is thin. A sourcing director trying to move a ski jacket program from a fully integrated factory in Jiangsu to Bangladesh will spend six months just finding a facility that can handle seam-sealing tape correctly. Factory audits don't catch this. You only find out when the first production run leaks.

Both countries also face their own political risk now. The USTR's May 2026 announcement of a Section 301 investigation into Vietnam's intellectual property practices sent a clear signal: the trade policy focus is expanding beyond China to any country running a large surplus with the US. The "shift to Vietnam" playbook looks less clever when Vietnam itself is in the crosshairs.

The Near-Shoring Paradox: Welcome to Guatemala. Here's Your T-Shirt.

Here's where the USFIA data gets genuinely interesting—and exposes what I'd call the near-shoring ceiling.

The study asked respondents what they plan to source from the Western Hemisphere over the next two to three years. The answers are almost comically narrow. T-shirts from CAFTA-DR countries: 67% of respondents said yes. Activewear and athleisure: 33%. Bottoms: 27%. From USMCA partners, the pattern is similar—47% said T-shirts, 20% said bottoms.

That's it. Three categories. If your product line includes woven blouses with pin-tuck details, tailored wool-blend trousers, or a bonded softshell jacket with laser-cut ventilation, near-shoring isn't on your menu. The factories simply don't exist.

I've walked through apparel factories in Guatemala. They're good at what they do—cut-and-sew knit tops at competitive cycle times, reasonable quality, decent compliance. But ask about garment dyeing capabilities, or if they can handle a polyester-spandex performance knit that requires flatlock stitching and silicone print logos, and the conversation stops. The technical depth isn't there. It's not that the workers can't learn. It's that the ecosystem—the dye houses, the specialty trim suppliers, the bonding and lamination facilities, the testing labs that can run an AATCC 127 hydrostatic pressure test and get you results in 24 hours—simply doesn't replicate what exists within a 100-kilometer radius of Shaoxing, China.

The USFIA study confirms this has been a consistent pattern across previous surveys. Limited product diversification in the Western Hemisphere remains the single biggest constraint on growth. You can write a sourcing policy that says "shift 30% of volume to near-shore suppliers," but if your product mix is 60% woven fashion and 40% performance outerwear, you'll struggle to execute that policy beyond basic replenishment programs.

"Shallow Diversification" Is the Real Risk Nobody's Naming

Let me coin a term here, because the industry needs language for what's happening: shallow diversification.

Shallow diversification is when a company spreads orders across more countries but doesn't build genuine operational depth in any of them. You're in eight countries instead of four, but each relationship is thinner—smaller volumes per factory, less leverage on pricing, fewer dedicated production lines, weaker quality oversight. The sourcing map looks diversified on a PowerPoint slide. On the ground, it's fragile in ways the spreadsheet can't capture.

What deeper diversification actually requires isn't just finding factories in new countries. It's developing the upstream fabric supply, training technical staff who understand your brand's quality standards intimately, visiting production lines often enough to catch problems before shipment, and building enough volume concentration to matter when there's a capacity crunch. None of that happens in a one-year planning cycle.

The 2026 USFIA data shows companies spreading orders across more non-Asian countries. Guatemala, Egypt, and Jordan all gained utilization. But the per-country volumes are small, and the product scope is narrow. This looks like risk management from 30,000 feet. From the factory floor, it looks like a quality control nightmare waiting to happen.

I remember a conversation with a production manager at an outdoor brand about three years ago. They'd just added a factory in Egypt to their sourcing matrix for fleece jackets. The first two seasons went fine. The third season, a fabric shipment from a Turkish mill got delayed at customs for three weeks, and the factory had no alternative local supplier. The entire program shipped six weeks late. The lesson wasn't "don't source from Egypt." It was "if you're going to source from Egypt, you need to understand the fabric supply chain as deeply as you understand the cut-and-sew operation." Most brands don't. They look at labor costs, duty rates, and lead times, and call it a sourcing strategy.

Tariffs Are the Engine—But Not the Whole Machine

The USFIA study lands in a specific policy moment. The Supreme Court struck down reciprocal tariffs, but the Trump administration's commitment to tariffs above MFN rates remains in place. Global tariffs under Section 122 expire on July 24, 2026. The administration plans to use Section 301 and Section 232 to authorize more tariffs going forward. The US-China trade tensions that drove the initial wave of diversification haven't gone away—they've just changed shape.

Meanwhile, the US-China tariff ceiling agreement reached at the May 2026 presidential summit created a new dynamic: future US tariffs on Chinese goods cannot exceed the levels established during the Kuala Lumpur economic and trade consultations. Each side committed to tariff reductions on $30 billion worth of goods. For apparel specifically, this means some breathing room on landed costs for China-sourced goods, at least in the near term.

But here's what I'd bet on: near-term tariff relief won't reverse the structural shift underway. Once a sourcing director has spent two years qualifying new factories, running fit samples through new production lines, and building relationships with new supplier leadership, they don't throw all that work away because a tariff dropped three percentage points. The diversification train has left the station. The question isn't whether US brands will continue reducing China exposure—the survey numbers answer that definitively. The question is whether the alternatives can support the full complexity of their product assortments.

What the Survey Doesn't Ask—And Why It Matters

One limitation worth flagging: the USFIA study surveyed 30 companies. That's a small sample by conventional research standards. But these aren't random respondents. They're executives from the largest US fashion brands, retailers, and importers—companies that collectively account for a massive share of US apparel import volume. When nearly 80% of them have more than 1,000 employees and they're telling you they've cut China orders to single-digit percentages, that's directionally significant even if the exact percentages would shift with a larger sample.

The bigger gap in the data isn't sample size. It's that the survey captures stated intentions, not observed behavior. Sourcing directors fill out surveys based on their current plans. Actual purchase orders get placed based on factory performance, fabric availability, shipping costs, and last-minute quality issues that no strategic plan predicted. The gap between what people say in April and what they ship in October is real.

I've also noticed something the survey doesn't directly measure: the quality of the supplier relationships that remain in China. When a brand cuts its China sourcing from 40% of total volume to 8%, it's not just reducing order quantity. It's fundamentally changing the factory relationship. The highest-volume programs move out first—basic replenishment styles where unit cost is the primary driver. What stays in China tends to be the complex stuff: styles with multiple fabrications, intricate construction, short lead times, small minimums. The factory that was your strategic partner for 200,000 units a year becomes your specialty house for 30,000 units of the hard-to-make styles. That's not the same relationship, and it doesn't command the same pricing or priority when capacity tightens.

This bifurcation has consequences. The Chinese factories that retain business are the ones with deep technical capability—the mills that can run a 20-denier nylon with a PU microporous membrane and hit 10,000mm hydrostatic pressure after 20 washes. Those factories become more valuable to their remaining brand clients, not less. The factories losing business are the commodity knitters and basic woven houses that can be replicated elsewhere. China's textile industry isn't shrinking uniformly. It's polarizing.

AI Enters the Sourcing Conversation

A detail in the 2026 USFIA study that's easy to miss: Figure 2-24 shows AI increasingly integrated into sourcing and business operations among surveyed companies. The study doesn't break this down in granular detail, but it tracks with what I'm hearing from sourcing teams on the ground.

AI in apparel sourcing right now clusters around a few specific applications. Supplier risk scoring—pulling trade data, compliance records, shipping performance, and financial indicators into a predictive model for factory reliability. Material cost forecasting—modeling cotton and polyester price movements to time purchase commitments. Production tracking—using computer vision on factory floors to monitor work-in-progress and flag bottlenecks before they delay shipments.

None of this replaces the sourcing director's judgment. What it does is compress the timeline for evaluating new suppliers in unfamiliar countries. When you're diversifying into Guatemala and Egypt for the first time, you don't have five years of personal relationships to fall back on. You need data to fill the gap. AI that's trained on trade-level shipping data, third-party audit histories, and financial filings can surface risks that a factory visit might not reveal—like a supplier whose on-time delivery rate drops 30% during Ramadan because their workforce composition shifts, or a mill whose energy costs just spiked because of a local fuel subsidy change.

The companies that navigate this fragmented sourcing landscape best won't be the ones with the cheapest unit costs. They'll be the ones with the best information about supplier reliability in countries where they have limited operational history.

Three Things Sourcing Directors Should Do Right Now

If I were still sitting on the brand side, here's what I'd prioritize based on this data.

First, audit your actual fabric origin, not just your garment origin. The country of origin on a commercial invoice tells you where the final assembly happened. It doesn't tell you where the yarn was spun, the fabric was knit, the dyeing was done, or the membrane was laminated. If you've shifted garment production to Vietnam but your fabric still comes from China, you've reduced your tariff exposure but not your supply chain concentration risk. Map the full chain back to the greige fabric level. You might be surprised how much "Made in Vietnam" volume still traces back to Zhejiang and Jiangsu.

Second, resist the temptation to over-diversify. The USFIA data shows companies spreading orders across more countries. That looks responsible from a risk committee perspective. But eight sourcing countries with thin relationships is more operationally fragile than four countries with deep ones. Pick your diversification targets carefully. Build genuine technical capability in those locations—train the factory's quality team on your standards, not just your own QC inspectors. Accept that you'll pay more per unit during the ramp-up phase. If you're not willing to invest in building depth, you're not really diversifying—you're just adding logistics complexity.

Third, be realistic about near-shoring's product ceiling. The USFIA data confirms that Western Hemisphere sourcing is concentrated in knit tops, activewear, and basic bottoms. If your assortment includes complex outerwear, tailored clothing, or fashion-forward woven styles, near-shoring isn't going to move the needle for those categories—at least not in the next three to five years. Plan accordingly. Don't write a sourcing strategy that assumes near-shore capabilities that don't exist yet.

What China Still Has That Nobody Else Does

Some context the survey data doesn't explicitly capture, but that anyone who's spent time in textile supply chains knows: China's advantage was never just about price. It was about completeness.

The distance from Keqiao fabric market in Shaoxing to the weaving hub in Shengze, Jiangsu is roughly 100 kilometers. Within that radius, you can source greige fabric, get it dyed, have it finished with a DWR treatment or antimicrobial coating, run it through a testing lab for ISO 811 hydrostatic pressure and JIS L 1099 B1 moisture vapor transmission rate, and ship it to a garment factory. Every dye class, every finishing chemical, every specialty trim supplier exists in that ecosystem. If a fabric development requires a 20-denier nylon with a specific PU microporous membrane that hits 15,000 g/m²/24h moisture vapor transmission, you can find three mills within a two-hour drive that can prototype it in two weeks.

Guatemala, Egypt, Jordan—these countries have capable garment factories. They don't have that upstream ecosystem. The fabric for a bonded softshell jacket produced in Guatemala is likely still coming from Asia. The specialty reflective trim for a workwear program in Egypt probably ships from a supplier in Dongguan. The production is near-shored. The supply chain isn't.

This is the structural reality that no trade agreement and no tariff policy can change quickly. Building a textile ecosystem takes decades. China built theirs over 30 years with massive state investment in industrial parks, wastewater treatment infrastructure, and logistics networks. Expecting Guatemala or Egypt to replicate that in five years misunderstands how textile supply chains actually develop.

The USFIA data reflects this indirectly: the same survey that shows declining utilization of Asian suppliers also shows Western Hemisphere sourcing stuck in a narrow product range. That's not a coincidence. It's the same supply-side constraint expressing itself across two different metrics.

Five Years from Now

If the current trajectory holds—and I see no reason it won't—we're heading toward a sourcing landscape where China handles the technically complex, small-minimum, high-mix production that can't easily be replicated elsewhere, while basic volume programs scatter across a wider map of countries that can handle commodity knit and woven categories. Vietnam and Bangladesh will remain major players but at lower concentration levels than the industry assumed five years ago. Guatemala, Egypt, and Jordan will grow their share, but almost entirely within T-shirts, basic activewear, and simple bottoms.

The companies that navigate this well will be the ones that stop treating sourcing diversification as a spreadsheet exercise and start treating it as an industrial development problem. You can't diversify into a country that doesn't have the upstream supply chain to support your product mix. You either invest in building that supply chain—training suppliers, connecting them with fabric mills, funding testing lab capacity—or you accept the product ceiling and plan your assortment accordingly.

Either approach is defensible. What's not defensible is pretending the ceiling isn't there.