La Paradoja
del Ensamble
组装悖论
Mexico exports $593 billion per year, ranking 12th globally. Its industrial complexity ranks 26th in the world — more sophisticated than Brazil, South Korea, or Israel in product space. Almost none of it is Mexican. The automotive sector is 100% foreign-owned OEMs. Electronics assemblers are Foxconn and Samsung. Medical devices are Medtronic and Abbott. Mexico provides the territory, the labor, the geographic position. The value-added flows to shareholders in Detroit, Tokyo, Seoul, and Shenzhen. The Assembly Paradox: world-class exports, thin domestic capture. 墨西哥每年出口5930亿美元，全球排名第12。其工业复杂度在世界排名第26——在产品空间上比巴西、韩国或以色列更复杂。但几乎没有一样东西是墨西哥的。汽车行业100%为外资OEM厂商。电子组装商是富士康和三星。医疗器械是美敦力和雅培。墨西哥提供土地、劳动力和地理位置。增值流向底特律、东京、首尔和深圳的股东。这就是组装悖论：世界级的出口，稀薄的国内价值捕获。
Not a polemic. A position: Mexico does not own its export economy. It rents its territory to it. 这不是一篇论战文章，而是一个立场：墨西哥不拥有其出口经济。它只是向其出租领土。
$178B in vehicle and parts exports. Eight foreign OEM assembly complexes. 3.26 million vehicles produced per year. Zero Mexican-owned brands. Compare: South Korea produced the Hyundai Pony in 1975 — just fifteen years after assembly lines opened. Mexico has had foreign-owned assembly since 1965. Sixty years. Zero brands.1780亿美元的汽车及零部件出口。八家外资OEM总装综合体。每年生产326万辆汽车。零个墨西哥所有品牌。对比：韩国于1975年生产了现代马驹——仅在装配线开放十五年后。墨西哥自1965年起即有外资组装。六十年。零个品牌。
Mexico retains ~30 cents of every dollar of gross exports. South Korea retains 72%. The United States retains 85%. Brazil, with its heavily protected and often inefficient domestic industry, retains 75%. The remaining 70% — profits, IP royalties, imported components — flows to headquarters in Detroit, Tokyo, Seoul, Guadalajara California, and Shenzhen. This is not a trade statistic. It is a description of who owns Mexico's economic output.墨西哥从每一美元总出口中留存约30美分。韩国留存72%。美国留存85%。拥有高度保护且往往低效国内产业的巴西留存75%。其余70%——利润、知识产权特许权使用费、进口中间品——流向底特律、东京、首尔、加利福尼亚瓜达拉哈拉和深圳的总部。这不是一个贸易统计数字。它是对谁拥有墨西哥经济产出的描述。
Herfindahl index of destination concentration: 0.64, extreme. OECD "high concentration" threshold is 0.25. Mexico's entire manufacturing export base — sixty years of accumulated foreign investment, 48 assembly plants, 900,000 direct automotive jobs — can be disrupted by a single political decision in Washington. This risk stopped being theoretical in February 2025, when a 25% tariff threat briefly eliminated the financial rationale for Mexico's auto sector.目的地集中度赫芬达尔指数：0.64，极度集中。OECD"高集中度"阈值为0.25。墨西哥的整个制造业出口基础——六十年积累的外国投资、48家总装工厂、90万汽车行业直接工作岗位——可能被华盛顿的单一政治决策所颠覆。这一风险在2025年2月不再是理论性的，当时25%关税威胁短暂地消除了墨西哥汽车行业的财务理由。
Mexico exports $593 billion per year. Not a single major export sector is owned by a Mexican company. The automotive sector is 100% foreign OEMs. Electronics assemblers are Foxconn and Samsung. Medical devices are Medtronic and Abbott. The aerospace firms are Bombardier, GE, and Safran. This is not a policy failure — it is the designed outcome of sixty years of the maquiladora development model, which was explicitly structured to attract foreign capital by offering cheap labour and geographic proximity to the United States. That model succeeded completely on its own terms. The problem is that those terms assigned Mexico a permanent role as lessor of land and labor, not as owner of production. When Korea ranked ECI #26 in the 1990s, the complexity was inside Korean-owned conglomerates. Mexico's ECI #26 exists inside foreign firms' processes. Move the firms, and the complexity migrates with them. 墨西哥每年出口5930亿美元。没有一家主要出口行业的企业是墨西哥本国所有。汽车行业100%是外资OEM厂商。电子组装商是富士康和三星。医疗器械是美敦力和雅培。航空航天公司是庞巴迪、GE和赛峰。这不是政策失败——而是六十年马基拉多拉发展模式的设计结果，该模式被明确构建为通过提供廉价劳动力和与美国的地理邻近性来吸引外国资本。该模式在其自身条件下完全成功。问题在于，这些条件将墨西哥永久性地定位为土地和劳动力的出租方，而非生产的所有者。当韩国在1990年代经济复杂度指数排名第26时，复杂性存在于韩国所有的财团内部。墨西哥的第26名存在于外国企业的流程之中。迁移这些企业，复杂性也随之迁移。
$593B across six sectors. Ownership: overwhelmingly foreign. 六大行业5930亿美元。所有权：压倒性地属于外资。
Sources: INEGI SNIEG, Banxico, ITC Trade Map, OEC (Observatory of Economic Complexity), AMIA 2023. Electronics figure encompasses HS84 (machinery/computers) and HS85 (electrical equipment) — the two largest HS2 chapters in Mexico's export basket.来源：INEGI SNIEG、墨西哥央行、ITC贸易地图、OEC（经济复杂度观察站）、AMIA 2023。电子数字包含HS84（机械/计算机）和HS85（电气设备）——墨西哥出口篮子中最大的两个HS2章节。
$178B in vehicle and parts exports. Eight OEMs. Zero Mexican-owned brands. 汽车及零部件出口1780亿美元。八家OEM厂商。没有墨西哥品牌。
Mexico is the world's seventh-largest vehicle manufacturer and the largest supplier of automobiles to the United States, holding 38.5% of the US auto import market. Forty-eight automotive assembly and engine plants operate in twelve states. The sector employs approximately 900,000 workers directly (and 2.5M in the supply chain). Every vehicle rolling off those assembly lines belongs to a foreign brand: General Motors, Nissan, Stellantis (Fiat/Chrysler/Peugeot), Ford, Volkswagen, KIA, Mazda, BMW, Honda, Toyota. The sector is Mexico's single largest export and has been since 2010. None of the firms that built it are Mexican companies. 墨西哥是世界第七大汽车制造国，是美国最大的汽车供应国，占美国汽车进口市场的38.5%。十二个州运营着48家汽车总装和发动机工厂。该行业直接雇用约90万工人（供应链另有250万）。从这些装配线上驶出的每辆汽车都属于外国品牌：通用汽车、日产、斯特兰蒂斯（菲亚特/克莱斯勒/标致）、福特、大众、起亚、马自达、宝马、本田、丰田。自2010年以来，该行业一直是墨西哥最大的单一出口类别。没有一家构建这个行业的公司是墨西哥企业。
| Manufacturer制造商 | Origin母国 | Units (000)产量（千辆） |
|---|---|---|
| General Motors | USA | 722 |
| Nissan | Japan日本 | 616 |
| Stellantis | NL/FR | 468 |
| Ford | USA | 365 |
| Volkswagen | DE | 349 |
| KIA | South Korea韩国 | 256 |
| Mazda | Japan日本 | 203 |
| BMW | DE | 117 |
| Mexican brand墨西哥品牌 | — | 0 |
Source: AMIA (Mexican Automotive Industry Association) 2023 annual data. Total industry production ~3.26M units.来源：AMIA（墨西哥汽车工业协会）2023年年度数据。行业总产量约326万辆。
What Mexico captures: wages and taxes墨西哥捕获的：工资和税收
The domestic value captured in Mexican automotive exports is primarily payroll, local services, and some domestic parts content. Assembly plant wages average $5–9/hour (vs $20–30 in the US for equivalent work), representing Mexico's fundamental competitive advantage. Taxes on automotive production are moderate — the sector benefits from IMMEX and other investment-facilitation tax schemes. Corporate profits and IP royalties flow entirely out of Mexico. 墨西哥汽车出口中捕获的国内价值主要是工资、本地服务和部分国产零部件。装配厂工资平均5-9美元/小时（相比之下，美国同等工作为20-30美元），代表了墨西哥的根本竞争优势。汽车生产的税收适中——该行业受益于IMMEX和其他投资便利化税收计划。企业利润和知识产权特许权使用费完全流出墨西哥。
USMCA pressures and Plan MéxicoUSMCA压力与墨西哥计划
USMCA (2020) requires 75% regional content for duty-free access (up from NAFTA's 62.5%), 70% of steel/aluminium from North America, and 40–45% of auto content from "high-wage" workers ($16/hour minimum). The high-wage rule creates pressure for Mexican wages to rise, while simultaneously providing OEMs incentive to source more North American content — which can mean Canadian and US parts, not necessarily Mexican. Mexico's auto sector is navigating a narrowing tariff channel. USMCA（2020年）要求免关税准入需含75%区域含量（较NAFTA的62.5%提高），70%的钢铝来自北美，40-45%的汽车内容来自"高工资"工人（最低16美元/小时）。高工资规则为墨西哥工资上涨创造了压力，同时激励OEM采购更多北美内容——这可能意味着加拿大和美国零部件，不一定是墨西哥的。墨西哥汽车行业正在应对一个收窄的关税通道。
Trump tariff threat: structural risk特朗普关税威胁：结构性风险
The Trump administration's 25% tariff threats on Mexican automotive exports (Feb–April 2025) created acute uncertainty. Mexico's auto sector sends ~$120B to the US annually. A sustained 25% tariff would price Mexican vehicles above comparable domestic US production and create manufacturing incentive to onshore. Sheinbaum negotiated temporary exemptions under the USMCA review process, but the structural risk — a single political decision in Washington can unwind 30 years of Mexican automotive investment — has been made explicit. 特朗普政府对墨西哥汽车出口的25%关税威胁（2025年2-4月）造成了急剧不确定性。墨西哥汽车行业每年向美国出口约1200亿美元。持续的25%关税将使墨西哥汽车价格高于可比美国国内生产，并产生回流美国的制造业激励。谢因鲍姆在USMCA审查程序下谈判获得了临时豁免，但结构性风险已被明确揭示——华盛顿的一个政治决定可以瓦解30年的墨西哥汽车投资。
~$179B in electronics. Foxconn, Samsung, LG, Flex. Jalisco is not Silicon Valley. 约1790亿美元的电子产品。富士康、三星、LG、Flex。哈利斯科不是硅谷。
Mexico's electronics exports at ~$179B encompass computers, televisions, set-top boxes, circuit boards, and an enormous volume of intermediate electronic components. The sector is geographically concentrated in the North and in Jalisco. Guadalajara's "Silicon Valley" reputation is well-earned in one sense — major global electronics assembly firms operate large facilities there. In another sense, it is a misnomer: Guadalajara is not producing chips, designing processors, or filing hardware patents. It is assembling, testing, and packaging products designed elsewhere. The firms: Foxconn (Taiwan/Apple supply chain), Samsung (South Korea), LG (South Korea), Flex (Singapore), Jabil (USA), Sanmina (USA). Mexican electronics IP: negligible. 墨西哥约1790亿美元的电子产品出口涵盖计算机、电视、机顶盒、电路板，以及大量中间电子元件。该行业在地理上集中在北部和哈利斯科。瓜达拉哈拉的"硅谷"声誉在某种意义上是名副其实的——主要全球电子组装公司在那里设有大型设施。但从另一个意义上说，这是一个误称：瓜达拉哈拉不是在生产芯片、设计处理器或申请硬件专利。它是在组装、测试和包装在别处设计的产品。相关公司：富士康（台湾/苹果供应链）、三星（韩国）、LG（韩国）、Flex（新加坡）、Jabil（美国）、Sanmina（美国）。墨西哥电子知识产权：可忽略不计。
Medical devices: Mexico as USA's #1 supplier医疗器械：墨西哥是美国的第一大供应国
Mexico exported $19.3B in medical devices in 2023, making it the United States' largest single-country medical device supplier — surpassing Germany, Japan, and China. The devices include surgical instruments, catheters, stents, orthopaedic implants, diagnostic equipment, and hospital supplies. The manufacturers: Medtronic (USA), Abbott (USA), Stryker (USA), Boston Scientific (USA), B. Braun (Germany), Cardinal Health (USA). Mexico's contribution: skilled assembly-line labour, compliance logistics, and Baja California's clustering advantages. Baja California accounts for >50% of Mexico's medical device exports. R&D, IP, and corporate profit: entirely abroad. 墨西哥2023年医疗器械出口193亿美元，成为美国最大的单一国家医疗器械供应商——超过德国、日本和中国。器械包括外科器械、导管、支架、骨科植入物、诊断设备和医院用品。制造商：美敦力（美国）、雅培（美国）、史赛克（美国）、波士顿科学（美国）、B. Braun（德国）、主要健康（美国）。墨西哥的贡献：技术熟练的装配线劳工、合规物流和下加利福尼亚州的集群优势。下加利福尼亚州占墨西哥医疗器械出口的50%以上。研发、知识产权和企业利润：完全在境外。
Aerospace: $10.7B and growing航空航天：107亿美元且持续增长
Mexico's aerospace sector has grown 15-fold since 2000, concentrated in Querétaro, Chihuahua, and Baja California. Key firms: Bombardier (Canada), Honeywell (USA), GE Aerospace (USA), Airbus (EU), Safran (France). Mexico produces fuselage structures, landing gear, avionics, and engine components. Mexico holds ~96 aerospace manufacturing certifications (AS9100, NADCAP). This is genuine technical complexity — but all of it sub-contracted to foreign OEM programmes. Mexico has not produced an indigenously-designed aircraft or aircraft engine. 自2000年以来，墨西哥的航空航天行业增长了15倍，集中在克雷塔罗、奇瓦瓦和下加利福尼亚。主要公司：庞巴迪（加拿大）、霍尼韦尔（美国）、GE航空（美国）、空客（欧盟）、赛峰（法国）。墨西哥生产机身结构、起落架、航电设备和发动机部件。墨西哥持有约96项航空航天制造认证（AS9100、NADCAP）。这是真正的技术复杂性——但所有这些都是转包给外国OEM项目的。墨西哥没有生产过本土设计的飞机或飞机发动机。
$52B in food exports. Avocados and tequila are Mexican. Beer is 98% foreign. Berries are complicated. 520亿美元的食品出口。牛油果和龙舌兰酒属于墨西哥。啤酒98%是外资。浆果比较复杂。
Agriculture is the one major export sector where Mexican-owned entities hold significant positions — but the picture is more complex than headline numbers suggest. The sector has four visible sub-segments with different ownership structures: Beer ($6.2B, 98% foreign — AB InBev/Modelo and Heineken/FEMSA dominate); Tequila and mezcal ($4.4B, majority Mexican-owned distilleries, but Diageo/Becle/ABInBev own the top brands); Fresh produce — avocados/tomatoes/berries ($12B combined, Mexican growers but US packing/distribution chains); Grains, sugar, livestock ($25B combined, complex mix of ejido, private, and foreign agribusiness ownership). 农业是墨西哥自有实体持有重要地位的一个主要出口行业——但实际情况比标题数字揭示的更为复杂。该行业有四个所有权结构不同的明显子领域：啤酒（62亿美元，98%外资——AB InBev/模型龙舌兰酒和喜力/FEMSA占主导）；龙舌兰酒和梅斯卡尔（44亿美元，多数为墨西哥所有的酿酒厂，但帝亚吉欧/贝克利/AB InBev拥有顶级品牌）；鲜产品——牛油果/番茄/浆果（合计120亿美元，墨西哥种植者但美国包装/分销链）；谷物、糖、畜牧（合计250亿美元，ejido、私人和外国农业企业所有权的复杂混合）。
| Product产品 | $B | Ownership所有权 |
|---|---|---|
| Beer啤酒 | 6.2 | 98% foreign (AB InBev / Heineken)98%外资（AB InBev/喜力） |
| Tequila / spirits龙舌兰酒/烈酒 | 4.4 | Mixed — brands foreign-acquired混合——品牌被外资收购 |
| Berries (strawberries, blueberries)浆果（草莓、蓝莓） | 3.7 | MX growers, US distributors墨西哥种植者，美国经销商 |
| Avocados牛油果 | 3.0 | MX private (Michoacán families, cartel-adjacent)墨西哥私人（米却肯州家族，毗邻卡特尔） |
| Tomatoes番茄 | 3.0 | MX growers (Sinaloa, Sonora)墨西哥种植者（锡那罗亚、索诺拉） |
Beer classification note: Grupo Modelo (owned by AB InBev since 2013) and Cuauhtémoc Moctezuma (Heineken since 2010) produce all major Mexican beer labels including Corona, Modelo Especial, Carta Blanca, and Indio. The "Mexican beer" consumed globally is made in Mexico but profits are Belgian and Dutch.啤酒分类注意：Grupo Modelo（自2013年起由AB InBev拥有）和Cuauhtémoc Moctezuma（自2010年起由喜力拥有）生产所有主要墨西哥啤酒品牌，包括科罗纳、模型、卡尔塔布兰卡和印第奥。全球消费的"墨西哥啤酒"在墨西哥生产，但利润属于比利时和荷兰。
Mexico was OPEC's equivalent of a mid-size exporter. By 2015 it became a net petroleum importer. 墨西哥曾是相当于OPEC中型出口国的存在。到2015年它成为石油净进口国。
| Year年份 | Crude exports $B原油出口（亿美元） | Petroleum imports $B石油进口（亿美元） | Net position净头寸 |
|---|---|---|---|
| 2004 | 28.6 | 10.2 | +$18.4B |
| 2008 | 49.3 | 22.0 | +$27.3B |
| 2012 | 52.7 | 30.4 | +$22.3B |
| 2015 | 18.5 | 21.1 | −$2.6B |
| 2019 | 22.4 | 30.2 | −$7.8B |
| 2023 | 22.0 | 36.5 | −$14.5B |
Sources: Banxico, PEMEX Statistical Yearbook, UN COMTRADE. Mexico exports crude (mostly to Asia and US Gulf Coast) and imports refined products (gasoline, diesel) due to insufficient domestic refining capacity.来源：墨西哥央行、PEMEX统计年鉴、联合国商品贸易数据库。墨西哥出口原油（主要流向亚洲和美国墨西哥湾沿岸），进口精炼产品（汽油、柴油），因为国内炼油能力不足。
The refining paradox: Mexico exports crude oil and imports gasoline and diesel. Domestic refining capacity is approximately 830,000 bpd across six state-owned refineries (plus the new Dos Bocas/Olmeca). Actual utilisation in 2024: approximately 40–45%. This means Mexico imports ~$15B/year in petroleum products while sitting on oil revenues of ~$22B — a net petroleum trade deficit that erodes the fiscal contribution of oil. 炼油悖论：墨西哥出口原油并进口汽油和柴油。国内炼油能力约为83万桶/日，分布在六个国有炼油厂（加上新的多斯玻卡斯/奥尔梅卡）。2024年实际利用率：约40–45%。这意味着墨西哥每年进口约150亿美元的石油产品，而石油收入仅约220亿美元——净石油贸易逆差侵蚀了石油的财政贡献。
The refining paradox is a classic agency problem compounded across decades. The government needed PEMEX's cash flow; PEMEX could not retain earnings to invest in refinery upgrades; deteriorating capacity required product imports; import bills eroded the fiscal surplus from crude exports. The net result: a $22B oil revenue stream partially consumed by a $15B petroleum product import bill — not because Mexico lacks oil, but because the state extracted so much from the producing entity that the entity can no longer productively transform its own resource. 炼油悖论是一个被数十年层层叠加的经典代理问题：政府需要PEMEX的现金流；PEMEX无法留存利润投资炼厂升级；炼油能力恶化迫使进口成品油；进口账单侵蚀了原油出口产生的财政盈余。最终结果：220亿美元的石油收入被150亿美元的成品油进口账单部分吞噬——不是因为墨西哥没有石油，而是因为国家从这个生产主体抽取过多，使其已无法有效地转化自己的资源。
$593B exports. ~$178B domestic value added. The rest leaves Mexico. 5930亿美元的出口。约1780亿美元的国内增加值。其余部分离开墨西哥。
The OECD/WTO Trade in Value Added (TiVA) framework estimates Mexico's domestic value added in gross exports at approximately 30–35%. This means for every dollar of goods manufactured and shipped from Mexico, roughly 65–70 cents of value — in the form of intermediate inputs, corporate profits, IP royalties, and headquarter services — was created and captured outside Mexico. The corresponding figure for the United States is ~85%. For South Korea: ~72%. For Taiwan: ~62%. OECD/WTO增加值贸易（TiVA）框架估计墨西哥在总出口中的国内增加值约为30-35%。这意味着每一美元在墨西哥制造和出口的商品中，约65-70美分的价值——以中间投入、企业利润、知识产权特许权使用费和总部服务的形式——是在墨西哥境外创造和捕获的。美国的对应数据约为85%。韩国：约72%。台湾：约62%。
Mexico's low domestic value capture is not a failure of Mexican workers but a structural feature of its development model. The maquiladora model — attractive for foreign firms precisely because it maximises their value retention — was never designed to develop Mexican productive capacity. It provides employment, wages, and a limited tax base. It does not generate domestic technology, supplier networks, brand ownership, or R&D infrastructure. A country can be among the world's top exporters and simultaneously be economically thin at home. Mexico is the proof case. 墨西哥较低的国内价值捕获不是墨西哥工人的失败，而是其发展模式的结构性特征。马基拉多拉模式——之所以对外国公司有吸引力正是因为它最大化了它们的价值保留——从未被设计用来发展墨西哥的生产能力。它提供就业、工资和有限的税基。它不会产生国内技术、供应商网络、品牌所有权或研发基础设施。一个国家可以跻身全球顶级出口国之列，同时在国内经济上保持单薄。墨西哥就是这一证明。
| Country国家 | DOM value %国内增值% | Model模式 |
|---|---|---|
| USA | ~85% | Integrated industrial base一体化工业基础 |
| South Korea韩国 | ~72% | National champions + assembly国家冠军企业+组装 |
| Brazil | ~75% | Protected domestic market受保护的国内市场 |
| Taiwan | ~62% | Own-IP semiconductor hub自有IP半导体中心 |
| China | ~70% | Domestic supply chain build-out国内供应链建设 |
| Vietnam | ~48%* | Textiles + Samsung electronics assembly纺织品+三星电子组装 |
| Mexico | ~30–35% | Foreign-owned assembly model外资所有组装模式 |
Source: OECD-WTO TiVA Database 2023, World Bank WDI. Mexico's low figure reflects the depth of its foreign-assembly structure — not product composition (which is sophisticated).来源：OECD-WTO增加值贸易数据库2023，世界银行WDI。墨西哥的低数字反映了其外资组装结构的深度——而非产品构成（产品构成很复杂）。
* Vietnam's structural split: the ~48% overall average conceals an extreme sectoral divide. Textiles and garments retain ~60–70% DVA — Vietnam has a genuine domestic fabric and thread industry. Electronics (dominated by Samsung, which alone accounts for ~20–25% of Vietnam's total exports) retains only ~15–25% DVA: chips, screens, and key components are shipped from Korea and China; Vietnam provides the factory floor and assembly labour. This makes Vietnam's export profile even more concentrated than Mexico's — a single foreign firm's production decisions can swing Vietnam's national export figures by several percentage points. Samsung's ongoing capacity shifts to India are the clearest proof of that structural fragility. * 越南的结构性分裂：约48%的总体平均数掩盖了极端的行业分化。纺织品和服装保留约60–70%的国内增加值——越南拥有真实的本土面料和纱线产业。电子产品（由三星主导，三星一家约占越南全部出口的20–25%）的国内增加值仅约15–25%：芯片、屏幕和关键元器件从韩国和中国运入，越南提供工厂场地和组装劳动力。这使越南的出口结构比墨西哥更为集中——单一外资企业的产能决策就能使越南的国家出口数字大幅波动。三星持续向印度转移产能，是这种结构性脆弱性最清晰的证明。
The ECI paradox explained: Mexico's Economic Complexity Index rank of #26 reflects the diversity and sophistication of what is made in Mexico — the product space occupied. It does not measure who owns the knowledge generating those products. Korea at a similar ECI rank in the early 1990s had Samsung, Hyundai, LG, and POSCO — Korean-owned firms investing in Korean R&D. Mexico at ECI #26 today has Foxconn, GM, Nissan, and Medtronic — foreign-owned firms that can and occasionally do relocate assembly to lower-cost jurisdictions. The complexity exists in the global companies operating in Mexico. Mexico hosts it; it does not own it. 经济复杂度指数悖论解释：墨西哥经济复杂度指数第26名反映了在墨西哥生产的产品的多样性和复杂性——所占据的产品空间。它不衡量谁拥有产生这些产品的知识。1990年代初期经济复杂度指数排名相近的韩国拥有三星、现代、LG和浦项制铁——投资于韩国研发的韩国所有企业。今天排名第26的墨西哥拥有富士康、通用汽车、日产和美敦力——能够并偶尔确实将组装迁移至低成本司法管辖区的外资所有企业。复杂性存在于在墨西哥运营的全球公司中。墨西哥是主办者；它不是拥有者。
Six border states generate 52% of all exports. The interior of Mexico is largely absent from export production. 六个边境州产生全部出口的52%。墨西哥内陆地区在很大程度上缺席出口生产。
| State州 | Exports $B出口（亿美元） | % national占全国% | Key sectors主要行业 |
|---|---|---|---|
| Chihuahua | 82–113 | 13–19% | Auto parts, electronics (Foxconn, BMW)汽车零部件、电子（富士康、宝马） |
| Coahuila | 65–72 | 11–12% | Automotive (GM, Chrysler, KIA)汽车（通用、克莱斯勒、起亚） |
| Nuevo León | 54–61 | 9–10% | Steel, glass, chemicals, auto钢铁、玻璃、化工、汽车 |
| Baja California | 52–61 | 9–10% | Medical devices (#1), electronics医疗器械（第一）、电子 |
| Jalisco | 42–50 | 7–8% | Electronics (Guadalajara hub)电子（瓜达拉哈拉中心） |
| Tamaulipas | 33–42 | 5–7% | Auto, energy, maquiladoras汽车、能源、马基拉多拉 |
| Interior + South内陆+南部 | 200–240 | ~38% | Agriculture, petrochemicals, tourism农业、石化、旅游 |
The geographic distribution of Mexico's exports is a near-perfect map of its integration with the United States. The six states that share the 3,180-kilometer US border — Chihuahua, Coahuila, Nuevo León, Tamaulipas, Sonora, Baja California — account for the vast majority of manufacturing exports. The three southern states (Guerrero, Oaxaca, Chiapas) that contain 30% of Mexico's poorest population contribute less than 2% of exports between them. 墨西哥出口的地理分布几乎完美地映射了其与美国的整合程度。共享3180公里美墨边境的六个州——奇瓦瓦、科阿韦拉、新莱昂、塔毛利帕斯、索诺拉、下加利福尼亚——占制造业出口的绝大多数。包含墨西哥30%最贫困人口的三个南部州（格雷罗、瓦哈卡、恰帕斯）合计贡献不到2%的出口。
This geographic duality — an industrialised north integrated into the US economy and an agricultural/informal south largely disconnected from export chains — is one of Mexico's deepest structural fault lines. Nearshoring investment since 2020 has been almost entirely captured by northern states and Bajío (Querétaro, Guanajuato). AMLO's attempt to redirect investment southward through the Tren Maya and Corredor Interoceánico has not materially changed the export geography. The south exports informality; the north exports to the United States. 这种地理二元性——整合进美国经济的工业化北方与很大程度上脱离出口链的农业/非正规南方——是墨西哥最深层的结构性断层之一。2020年以来的近岸外包投资几乎完全被北部各州和巴希奥地区（克雷塔罗、瓜纳华托）捕获。AMLO通过玛雅火车和跨洋走廊将投资引导至南方的尝试没有实质性改变出口地理格局。南方出口非正规性；北方出口给美国。
Sheinbaum's structural response to the assembly paradox: make the foreign firms buy Mexican. 谢因鲍姆对组装悖论的结构性应对：让外资企业购买墨西哥商品。
Announced in early 2025, Plan México is the Sheinbaum administration's industrial policy framework designed to increase domestic value capture from the manufacturing export base. It operates through a tiered national content incentive: firms achieving "Nivel Básico" (30% Mexican-national content) receive preferential access to public procurement, expedited permits, and reduced administrative burden. Firms achieving "Nivel Avanzado" (50% national content) receive additional fiscal incentives. The plan is particularly targeted at the automotive, electronics, aerospace, and medical device sectors. 2025年初宣布，墨西哥计划是谢因鲍姆政府的工业政策框架，旨在提高制造业出口基础中的国内价值捕获。它通过分级国内含量激励机制运作：达到"基础级别"（30%墨西哥国内含量）的公司可优先获得政府采购、快速审批和减少行政负担。达到"高级级别"（50%国内含量）的公司获得额外财税激励。该计划特别针对汽车、电子、航空航天和医疗器械行业。
What Plan México targets计划目标
The plan explicitly targets increasing backward linkages — getting foreign OEMs to source more inputs from Mexican-owned or Mexican-incorporated suppliers. The government has identified 12 "strategic supply chains" (automotive electronics, medical devices, semiconductors, aeronautics, agro-industry, green energy) where domestic supplier development is specifically incentivised. NAFINSA (national development bank) has committed $8B in credit for domestic supplier capacity building. 该计划明确针对增强反向关联——让外资OEM从墨西哥所有或墨西哥注册的供应商处采购更多投入品。政府已确定12条"战略供应链"（汽车电子、医疗器械、半导体、航空工业、农业工业、绿色能源），其中国内供应商发展受到专项激励。NAFINSA（国家开发银行）承诺提供80亿美元信贷用于国内供应商能力建设。
The structural obstacles结构性障碍
Industrial policy to deepen supply chains requires: scale financing (Mexico's SME lending market is thin and expensive), technical capability development (skills take a generation to build), certification infrastructure (AS9100, ISO 13485, cybersecurity compliance), and OEM willingness to switch suppliers when incumbent global chains are more reliable. Mexico's Rule of Law Index (0.44) also deters complex long-term B2B contracts. Plan México faces the same obstacles as every previous Mexican localisation initiative since the 1970s. 深化供应链的工业政策需要：规模融资（墨西哥的中小企业贷款市场薄弱且成本高昂）、技术能力开发（技能需要一代人才能建立）、认证基础设施（AS9100、ISO 13485、网络安全合规）以及当现有全球链更可靠时OEM愿意更换供应商。墨西哥的法治指数（0.44）也妨碍了复杂的长期B2B合同。墨西哥计划面临着与自1970年代以来每一个此前墨西哥本地化计划相同的障碍。
Nearshoring: opportunity and capture近岸外包：机遇与捕获
The US-China supply chain decoupling has created a nearshoring wave: FDI into Mexico reached $36B in 2023 and $39B in 2024, the highest in Mexican history. Monterrey, Saltillo, and Querétaro are constructing industrial parks at record rates. But the investment composition: ~90% is foreign-owned manufacturing for US export. The nearshoring boom is, so far, a larger version of the same assembly model Mexico has operated since 1965. Plan México aims to change who benefits from the boom; whether it can do so at scale against global supply chain inertia is the central question. 美中供应链脱钩创造了近岸外包浪潮：流入墨西哥的FDI在2023年达到360亿美元，2024年达390亿美元，是墨西哥历史上最高的。蒙特雷、萨尔蒂略和克雷塔罗正以创纪录的速度建设工业园区。但投资构成：约90%是面向美国出口的外资所有制造业。到目前为止，近岸外包繁荣是墨西哥自1965年以来运营的同一组装模式的更大版本。墨西哥计划旨在改变谁受益于这一繁荣；面对全球供应链惯性，它能否大规模实现这一目标是核心问题。
Mexico has the export profile of a developed industrial economy and the domestic ownership structure of a prize assembly zone. The assembly paradox is not a failure of policy implementation — it is a structural outcome of the maquiladora development model, which was explicitly designed to attract foreign capital by offering cheap labour and geographic proximity to the United States. That model succeeded on its own terms. Mexico is the world's 12th-largest exporter. It has not developed the ownership, R&D, brand equity, or supply chain depth that would make those exports drivers of Mexican domestic wealth accumulation. 墨西哥拥有发达工业经济体的出口特征，却拥有优质装配区的国内所有权结构。组装悖论不是政策执行的失败——它是马基拉多拉发展模式的结构性结果，该模式被明确设计为通过提供廉价劳动力和与美国的地理邻近性来吸引外国资本。该模式在其自身条件下取得了成功。墨西哥是世界第12大出口国。它没有发展出能使这些出口成为墨西哥国内财富积累驱动力的所有权、研发、品牌资产或供应链深度。
The two gravitational forces at work: First, US dependency concentrates 80% of exports in one market, creating catastrophic fragility to US tariff policy. A single Trump executive order can eliminate the financial rationale for dozens of assembly plants. Second, foreign ownership concentrates value capture outside Mexico, so GDP growth from exports does not translate proportionately into Mexican wages, tax revenue, or investment. Mexico's export machine is a gravitational force in its own right — but it pulls value northward and outward, not inward. Plan México is an attempt to redirect that gravity. The structural inertia of four decades of embedded foreign supply chains is the force it must overcome. 两种引力在发挥作用：首先，美国依赖将80%的出口集中在一个市场，造成对美国关税政策的灾难性脆弱性。特朗普的一项行政命令就能消除数十家装配厂的财务理由。其次，外资所有权将价值捕获集中在墨西哥境外，因此出口带来的GDP增长不能按比例转化为墨西哥的工资、税收或投资。墨西哥的出口机器本身就是一种引力——但它将价值向北方和向外拉，而非向内拉。墨西哥计划试图重新引导这种引力。四十年根深蒂固的外国供应链的结构性惯性是它必须克服的力量。