The Department of Energy yesterday reported a sharp drop last year in emissions of carbon dioxide, and a steeper decline than what was anticipated through the impacts of the recession alone. The graph above shows one particularly notable disconnect — a drop in emissions far steeper than the drop in gross domestic product. The report uses well-designed graphics to break down the trend sector by sector and every page is worth exploring. Another notable finding is the influence of a big switch from coal to natural gas for electricity generation, as gas prices fell nearly 50 percent while coal prices rose 6.8 percent relative to 2008. For anyone who cares about the climate, the bottom line there — because natural gas emits nearly half the carbon dioxide as coal for the same amount of produced heat — is finding a way to manage risks from harvesting vast deposits of gas without rejecting that resource altogether. The findings have cheered environmentalists and climate campaigners, who see signs that the country could hit proposed targets for emissions in 2020 without too much cost or disruption. It’s conceivable, some seasoned experts on energy and the economy say, but way too soon to celebrate. One reason simply is that the underlying data tracked by the government are notoriously “foggy,” according to Lee Schipper, a specialist in energy use with appointments at Stanford University and the University of California, Berkeley. Dr. Schipper, who has spent many years sifting data on home size, heating bills, appliance purchases, driving habits, freight shipments and other activities that indirectly gauge the use of different fuels, sent me a note about the challenge of gauging trends during turbulent economic times. One big issue, he said, is that the Department of Energy has stopped tracking many lines of data that matter. Here’s his note: If there is one lesson, it is that in times of rapid growth or recession, different parts of the economy change at different rates, and that differential alone can cause significant changes in the ratio of energy to G.D.P. Since a big recession might hit coal-burning utilities’ customers more than other utility customers (to name one example) or hit coal-using industries like cement and steel more than others, one has to look carefully not only at CO2 emissions changes but at underlying economic activity or personal activity changes and how those are tied to emissions in a disaggregated way. Some countries can do this roughly 18 months to two years after the end of each year. We can’t. We don’t even maintain regular energy accounts by major manufacturing branches. We last surveyed household vehicle fuel use in 1985, and our trucking inventory and use survey died in 2002. We stopped trying to estimate household appliance electricity use in the late 1990s…. I call this the blind leading the blind. Like “ Cash for Clunkers Is a Lemon,” as I wrote in the Washington Post, we seem to like to make policies (or pronouncements) whose outcomes cannot be measured for years. I remember when high-level clowns in the Bush administration were pointing to the decline in carbon emissions in the mid-2000s, but of course not taking credit (or blame) for the higher oil and gas prices that most agreed lay behind those declines. It’s hard to imagine how the U.S. will enact any sensible policies in this foggy atmosphere.