資深台海觀察家Chris Horton先生，於7月28日在《日經亞洲》評論版發表〈Taiwan's pivot away from China offers a lesson for democracies〉一文，主張台灣過去十年的貿易轉向是一個良好示範，證明了民主國家可以在降低對中共經濟曝險的同時保持繁榮。我認同貿易與國家安全早已無法切割，世界雖要降低中共風險的文章出發點，台灣過去十年的轉變也確實值得世界關注。 但文章接下來的具體論斷我難以認同。Horton把這個轉變幾乎完全歸功於蔡英文政府的政策能動性——太陽花運動之後的社會共識、蔡英文以「降低對陸依賴」為政見當選、八年任內主動把貿易重心移向美國，並直言「我們現在活在蔡英文打造的世界裡」。這個敘事把功勞幾乎全部算在單一政府頭上，卻迴避了同一時期規模遠大於任何一個政府所能主導的結構性變化。 這個結構變化源於美國和中國大陸的關稅戰，雙邊貿易砍掉近三成，中國大陸逼於內需疲弱、出口導向模式並未因此收斂，反而以更高的補貼強度衝擊全球市場，逼迫其他經濟體加速尋找替代供應鏈；與此同時，AI資本支出進入超級週期，輝達等美國企業的訂單需求暴增。中華民國剛好站在這兩股外部力量的交會點上。這種全球範圍的出口結構異動，與其說是台北的戰略選擇，不如說是台灣被這波全球重組浪潮推著往前走。如果把外生的景氣循環和地緣政治衝擊，包裝成單一政府的治理成果，這個因果順序值得商榷。 更值得商榷的，是Horton所依賴的官方數據本身。台灣對中國大陸投資佔比降至0.9%，經常被引用為「幾乎完全脫鉤」的證明。但拆開投審司原始報表會發現，2026年上半年台灣對外投資總額中，高達85%是台積電透過境外子公司投資銀行定存與美元債券操作，是一種用來管理匯率避險成本的資金停泊，跟產業佈局、供應鏈重組毫無關係。把這筆財務操作從分母剔除，中國大陸在台商「實際生產性投資」裡的佔比，會從0.9%躍升到7%至8%之間，落差近八倍。同一份報表裡，金融保險業、專業科學技術服務業對陸投資年增率分別高達479%與224%，雖然爆發式的百分比成長源於低基期因素，方向卻與官方宣示的「降低曝險」背道而馳。這種只引用籠統的整體數字、迴避數字背後結構的做法，恰好與台灣執政黨近年慣用的敘事手法如出一轍：兩者都傾向把混雜的現實，簡化政策成功曲線，Horton在這一點上，幾乎是原封不動地接收了這套敘事。這種選擇性引用是台海觀察的一個大盲點。 這條籠統曲線底下，還有一個更少人談的落差，就是經濟紅利分配並不均。關稅戰重創台灣機械、金屬這類傳統製造業。出口數字的亮眼成長幾乎全部集中在半導體與資通訊供應鏈。總體數字上揚的同時，台灣經濟內部正在被切成兩個世界——一個吃到AI紅利，一個承受關稅代價。這個落差，不會出現在「對美出口首度超陸」這種整體統計裡。 最後，也是最根本的問題：即使台灣真的降低了對中國大陸的曝險，換來的也不是風險分散，而是風險轉移。彭博社的數據顯示，台灣超過53%的出口集中在電機、電子與電腦相關產品；輝達六成三的生產成本流向台灣供應商；台積電來自美國市場的營收佔比，十年間從64%升至74%。這不是把雞蛋放進更多籃子，是把雞蛋從一個敵視台灣的政權手中，整批搬進另一個超級經濟體的籃子裡。而這個超級經濟體，眼下正親手重新定義國際貿易規則——川普政府一邊要求台灣擴大在美投資，一邊公開稱台灣是「談判籌碼」；台灣經濟研究院商業發展研究中心主任孫明德已公開示警，持續擴大的對美順差，本身就可能成為觸發下一輪關稅施壓的引信。台灣換來的不是確定性，是把一種曝險換成另一種曝險，而後者的規則，還在被單方面改寫中。 Horton的問題不在於他對台灣的關注，也不在於他所主張的大方向，而在於他跟他所批評、卻又不自覺套用的那套政治敘事一樣，太急於把一個仍在進行、仍然混亂、仍然充滿結構性風險的過程，講成一個已經完成、可以打包輸出的成功案例。台灣的故事還沒寫完，這是我認為眼下最誠實的說法。 Taiwan’s trade dependence runs deep The global chip boom and weak Chinese import have driven Taiwan’s trade realignment Over the past decade, Taiwan has reduced its share of exports and outbound investment directed at mainland China. The shift is often presented as a success of Tsai Ing-wen’s eight years in government, at a time when democracies are seeking ways to bolster their economic security. This narrative fails to grasp the structural forces unfolded over the same period. One such example is a recent opinion piece penned by Chris Horton, the veteran Taiwan-based journalist, which argued that Taiwan's trade realignment over the past decade offers democracies a model: proof that a country can reduce its economic exposure to Beijing while continuing to prosper. Taiwan’s trade realignment has taken place against two broader developments. Squeezed by weak domestic consumption, mainland China has doubled down on its export-led model, flooding global markets with subsidized goods. At the same time, AI capital expenditure has entered a supercycle. Orders from Nvidia and other American firms are surging exponentially. Taiwan sits at the intersection of these two external forces. This global reshuffling of export patterns looks less like a strategic choice made in Taipei. Taiwan is being carried along by a wave of realignment it did not set in motion. Presenting mainland China’s weak import and an exogenous business cycle and a geopolitical shock as the governance achievement of a single administration is, at the very least, a claim worth scrutinizing. Taiwan's reduction of outbound investment in mainland mainland China to 0.9% is often cited as evidence of near-total decoupling. But filings from the island’s Department of Investment suggest a different picture. In the first half of 2026, 85% of Taiwan’s outbound investment came from TSMC placing funds in offshore bank deposits and dollar bonds -- a currency hedge, not investment in factories or supply chains. If that transaction is taken out of the denominator, mainland China's share of Taiwanese firms' productive investment jumps from 0.9% to 8%. In the same filing, investment in mainland mainland China by the finance and insurance sector, and by professional, scientific and technical services, grew 479% and 224% respectively year-on-year. This runs counter to the government's stated goal of reducing exposure. Taiwan’s ruling party has a habit of citing aggregate headline figures while ignoring the structure beneath them -- and some analysts have absorbed that narrative almost wholesale. This selective use of data is a significant blind spot in analysing Taiwan’s economy. Beneath that tidy curve lies a less-discussed disparity: the benefits of Taiwan’s pivot have not been evenly shared. The U.S. tariff war has hit Taiwan's traditional manufacturers, such as machinery and metals, hard, even as headline export growth has been concentrated almost entirely in semiconductors and the ICT supply chain. As the aggregate numbers climb, Taiwan's economy is splitting into two worlds: one riding the AI boom, the other absorbing the cost of tariffs. That gap does not show up in a topline statistic like "exports to the U.S. overtake mainland China for the first time." The final, and most fundamental, problem is this: even if Taiwan has genuinely reduced its exposure to mainland mainland China, what it has gained is not diversification but a transfer of risk. Bloomberg's data shows that more than 53% of Taiwan's exports are concentrated in electrical machinery, electronics and computer-related products; Nvidia sends 63% of its production costs to Taiwanese suppliers; and TSMC's revenue share from the U.S. market has climbed from 64% to 74% over the past decade. Taiwan is not spreading eggs across more baskets -- it is moving the eggs, in bulk, from the hands of a regime hostile to Taiwan into the basket of a single superpower economy. And that superpower is, at this very moment, rewriting the rules of international trade on its own terms. The Trump administration is pressing Taiwan to expand its U.S. investments while publicly calling the island a "negotiating chip." Sun Ming-te, director of the Business Development and Research Center at the Taiwan Institute of Economic Research, a think tank, has warned that a widening trade surplus with the U.S. could itself become the trigger for the next round of tariff pressure. What Taiwan has bought is not certainty -- it is one exposure traded for another, and the rules governing the new one are still being rewritten unilaterally. Horton's failing is not his attention to Taiwan, nor the broader direction he advocates. It is that he -- like the very political narrative he elsewhere critiques but here adopts without noticing -- is too quick to describe The challenges facing Taiwan are still unfolding, still messy and still fraught with structural risk. On paper, Taiwan may be making progress in diversifying away from mainland China. But its story is not yet written. ＊作者為英國新興智庫Sinic Analytica的研究員
评论 (0)