Whoever lays the rails of an age dictates, in some measure, where its commerce shall travel. The rails of our age are not iron; they are silicon — the semiconductor fabrication plants, the photolithography machines, the engineers and the supply chains that produce them — and the United States has spent three decades discovering, with mounting alarm, that it does not own enough of them.

The CHIPS Act, in its essentials, was a correct measure. A national industrial base is not maintained by accident or by the assumption that the market alone, without inducement, will return what it once exported. The Report on Manufactures, which I have the honor of having written, argued in the 1790s that infant industries of strategic significance deserved the public hand for a season. The argument applies now with more force than it did then, because the alternative — dependence on a small island in a contested strait — is not a commercial inconvenience but a strategic peril.

What the legislation got right was the structure: direct investment in fabrication capacity, tax credits aligned to actual domestic production, and conditions on recipients that prevent the immediate offshoring of the very capacity the public has subsidized. What the implementation has gotten wrong, by accounts I find credible, is the pace. Subsidy moves at the rhythm of the appropriations cycle; capital expenditure at the rhythm of the foundry. The two are out of phase, and the result has been delays the strategic argument cannot afford.

Two corrections are needed. First, the appropriations should be multi-year and reliable, so that the foundries can plan their capital deployment without the periodic suspense of whether the Treasury will be at home next quarter. Second, the conditions attached to the subsidy — on labor, on environmental compliance, on supply-chain accounting — should be selected for their relevance to the strategic purpose. A condition that delays a fabrication plant by two years to obtain a marginal labor advantage may, on the strategic ledger, be a poor trade. That is a hard judgment, and reasonable observers will weigh the conditions differently; the public is owed an honest accounting of which trades have, in fact, been made.

The deeper point is that national vigor is not maintained without the will to spend, the discipline to spend usefully, and the patience to spend across more than one election. The semiconductor industry will not be re-shored by any one administration. It will be re-shored, if at all, by a continuity of policy that survives a change of party — which is to say, by the public faith in a strategic purpose that is held in common.

That faith is the asset the public expects of its officers. It is the asset, more than the appropriations, that they have so far been slow to provide.