Proponents of the Tax Cut and Jobs Act (TCJA) claimed it would boost the US economy and generate higher incomes for the American people. At first glance, recent estimates from the Congressional Budget Office (CBO) appear to support that  argument. But the CBO analysis includes a subtle, yet extremely important difference: Foreign investors will end up receiving much of the gains and the net income available to Americans will rise barely, if at all.

The CBO analysis implies that TCJA effectively will have no impact US incomes after 10 years. The difference, in econ-speak, is between the estimated effect on Gross Domestic Product (GDP, or the output created within in the US), the effect on Gross National Product (GNP, output created by American workers and American-owned capital), and ultimately on Net National Product (NNP, which is GNP minus depreciation of capital goods, and comes closest of the three measures to American incomes). Hang on while we explain the difference.

CBO estimates that TCJA will increase US GDP by 0.5 percent in 2028. CBO projects that the tax cuts will boost output in 2028 largely because lower tax rates on capital income—such as the 21 percent rate on corporate profits—increases the after-tax rate of return which in turn will boost the stock of productive capital such as computers or factories.

But here’s the kicker:  CBO figures that most of that additional capital will be financed by foreigners—for example, from overseas corporations building factories in the US, or foreign investors buying US stocks and bonds. As a result, net payments of profits, dividends, and interest to foreigners also will rise. Unlike GDP, the GNP subtracts those net payments to foreigners from domestic production. GNP therefore provides a better measure of the impact on US incomes. CBO projects that tax bill will boost GNP by just 0.1 percent in 2028.

But GNP tells only part of the story. Because the increase in output stems mostly from additional investment in capital goods, the nation’s capital stock will be higher relative to output. That’s good because it can raise worker productivity and wages, on average. But to maintain that larger capital stock a larger share of output must be devoted to offsetting depreciation—the wear and tear on those additional capital goods.

It turns out that the rise in depreciation is about 0.1 percent of output in 2028—enough to erase the already meager boost to GNP. Thus, long-run incomes for Americans as measured by NNP will be more or less unchanged by the TCJA.

And that’s the good news about the TCJA. It ignores the negative effects of the tax law: Worsening income inequality, less revenue to finance government services and benefits, and higher federal debt. If the tax cut’s direct benefits on US incomes are non-existent, it is hard to make a case that it is a positive for the US economy in the long term.

Benjamin R. Page is a senior fellow and William G. Gale holds the Arjay and Frances Fearing Miller Chair in Federal Economic Policy at the Brookings Tax Policy Center. This article originally appeared here.