Innumeracy

Robert Waldmann

Doug Elmendorf says that the costs of cap and trade will be miniscule and does not try to estimate the benefits. Juliet Eilperin of the Washington Post demonstrates that she is the second most innumerate person on the planat, the most innumerate is the** guy who wrote the headline “Cap-and-Trade Would Slow Economy, CBO Chief Says”

The headline supposed to be justified by

Elmendorf testified before the Senate Energy and Natural Resources Committee that the cap-and-trade provisions of the House bill — in which emitters of greenhouse gases would be able to buy and sell pollution credits — would cut the nation’s gross domestic product by 0.25 to 0.75 percent in 2020 compared with “what it would otherwise have been,” and by 1 to 3.5 percent in 2050.

Now the standard way of describing GDP growth is the annual rate. As the cost of health insurance subsidies and medicaid advantage is the 10 year cost not the yearly cost, so the effect of cap and trade on growth is the effect after 10 years or 40 years not the effect on the growth rate. Elperin doesn’t say when the CBO assumes cap and trade will be introduced (the bill they scored did but I haven’t read it relying as I do on newspapers to report the news). I guess it would start in 2010, so Elmendorf said the growth rate of GDP would be reduced by 0.025 % to 0.075 %percent for ten years and 0.025 % to 0.0875 % percent for 40 years.

These are tiny tiny numbers compared to the average growth rate and compared to the standard deviation of the annual growth rate and compared to anything.

In a standard Solow type growth model with perfect competition and no spillovers, you can’t get something for nothing. The CBO calculates that we will get a cooler climate for as close to nothing as the result of any such calculation has ever been.

But numbers don’t matter at the Washington Post — they only care about the sign of the number which follows directly and immediately from standard (and false) assumptions.

Scheduled flight WP0085* lowered it’s cruising speed from 600 to 599.975 mph and therefore stalled, crashed and burned.

*The flight number 85 is not to be interpreted as extremely rude and unfounded speculation about anyone’s IQ.

** update: typo corrected.

Letters matter too. The calculation is of GDP. We might care more about GNP. Other things equal GNP -GDP will be reduced by the interest on the money we send overseas to buy petroleum. This is not a miniscule sum.

in 2005 US oil imports were around $ 200 billion. Ignore the fact that US petroleum production will decline and assume that the relative price of petroleum will be equal to that in 2005. That is, very roughly, 1.4% of GNP. Over 10 years the cumulated value of petroleum imports will amount to very very roughly 13% of GNP (here I assume we pay interest lower than the growth rate of GNP as foreign suckers buy T-bills). According to the CBO, cap and trade would cause higher GNP in 2020 if we pay 2% on the debt, that cuts 0.026 % of GDP from GNP-GDP.

With totally absurd assumptions about how little we will import and how cheap it will be and that foreign suckers will keep loaning to us at low rates, I get a cost of interest on petroleum imports of the same order as the estimated effect of cap and trade on GDP. With semi realistic assumptions about the volume of imports and the relative price of petroleum (note the dollar value of imports roughly doubled from 2000 to 2005) I can easily get that cap and trade will pay for itself if it significantly reduces importation of petroleum.

That is not counting the value of coal still in the ground let alone the value of reduced global warming which is, you know, the point of the whole exercize.

update: Coal still in the ground. GNP and GDP count the value added by mining as the value of the product, not the value added by digging it up. This is a measurement error. US coal production is about half a billion tons. The price differs *a lot* depending on quality and bounces around a lot, but is roughly around maybe $ 40/ ton so coal “production” is about 0.15% of GDP (why can’t I find this number on the web). Just guessing really, I’d say that amounts to measurement error of well over 0.1% of GDP (0.1% corresonds to say that one third of the value of coal is created by digging it up which is absurdly high). OK so measurement error in coal mining value added is smaller than the effect of cap and trade on measured GDP. The reduction in the over estimate of production is only about 10% of the estimated reduction in measured production.