My American Enterprise Institute colleague Nick Eberstadt provides a reminder that the federal government has been a pioneer in producing useful and reliable statistics. It began with the Constitution's unprecedented and, in Eberstadt's words, “truly revolutionary” requirement that a census of population be conducted at regular intervals and that the number of representatives each state would have in the House of Representatives would be based on it.

That’s the good news. The bad news is that the quality and quantity of government statistics has been deteriorating. Eberstadt decries in particular the government’s “hastily devised ‘federal poverty measure’ unveiled in 1965,” which he characterizes as “not merely a bad yardstick” but “an awful one.” It misstates actual poverty by measuring income rather than consumption (most American lower-income people consume more than they earn) and is based on an assumption, not wholly unwarranted in 1965 but antique now, that households spend about one-third of their incomes on food.

He makes this argument more fully in his 2008 book The Poverty of the Poverty Rate. Also, Eberstadt bemoans the defunding, under both Democratic and Republican administrations, of the Census Bureau's Survey of Income and Program Participation.

There's also reason to worry about whether the institutional culture of the government's statistical agencies -- a culture proud of its statistical rigor and freedom from political influence -- is being undermined by the Obama administration. A November 2013 New York Post story about the Census Bureau providing false figures in the weeks before the 2012 general election was disturbing, though AEI's Jim Pethokoukis argues that more evidence is required before concluding that there was wrongdoing.