As an economist, I don’t like the adage that “figures don’t lie, but liars can figure” because it implies that no quantitative data can be trusted. But, having worked with economic data for some three decades, I know that even very reliable data sets have quirks or nuances make them easy to misunderstand.
Unfortunately, that also means they can be misused, through ignorance or malice, to create a false impression.
This happens a lot during political campaigns. Candidates make assertions about their own accomplishments or their opponents’ failures that are not necessarily true. Outright lies are not common, because being caught in a blatant falsehood can carry a political penalty. Misrepresentations are extremely common, however.
All this has led to a proliferation of media “fact checkers” who examine campaign assertions. In most cases this is a healthy corrective. But in some cases of economic indicators, the fact checkers don’t understand the data involved, and their explanation muddies the waters rather than clarifying them. And there are many cases in which even experts would disagree on the interpretation.
Two caveats are important: Most political claims assume that officeholders can control what happens in the economy. What a president or governor actually can influence in an economy amounts to much less than is commonly thought. What an individual congressman or legislator can do is virtually nil.
Moreover, most campaign claims are flagrant examples of the “post hoc fallacy” that college students are taught to avoid, the erroneous assumption that because one thing happened before another, the first thing necessarily caused the second thing. An example would be “Ed Lotterman went to work for the Federal Reserve in 1992 and the economy began an eight-year run of good growth.” Of course, a minor functionary at a regional bank has nothing to do with national economic growth, but some claims made by politicians are nearly as ridiculous.
All that said, candidates are quick to claim successes using weak reasoning, so they cannot complain when weak reasoning is used against them. It is the game that everyone plays. Live by the sword and die by the sword.
For an example, consider a recent report by FactCheck.org that dinged the Obama campaign claiming that 5.2 million new jobs were added during his term. An on-screen graphic does note that this is starting in March 2010 rather than the day he took office. There is no explanation, however, that this refers only to private-sector jobs. It ignores that government employment has fallen steadily for nearly four years. Overall, the employment increase since Obama’s inauguration is only 325,000.
FactCheck did the public a service by pointing out the discrepancies in the Obama ad, particularly its lack of transparency about the details of the data cited. Good job.
However, it raises the question of when one can reasonably expect the policies of a new president to have an effect on some economic variable such as employment. If a president is inaugurated on Jan. 20, can he do anything to boost the February job numbers, which will be measured three weeks after he ends his oath? Of course not. What about March, April, May, etc.?
Over time, George W. Bush’s responsibility for labor market conditions, however limited or great that may be, had to fall, and that of Barack Obama had to rise. But is there a clear milestone when the baton of responsibility passes? Unfortunately there is not.
The whole issue here is what economists call “lags.” There always is some lapse of time between when a change in some policy involving taxes, spending, interest rates or whatever, begins to have any effect on the real economy.
Unfortunately, these lags vary from one policy to another and from one historical situation to another. Economists don’t agree on what they are exactly.
I think economists, regardless of political affiliation, would agree that starting a year into a new president’s term gives a more accurate picture of responsibility than starting on inauguration day. But is 12 months the best transition point? No one really knows.
I’m sure I’ll get emails asserting Obama should rightly be charged with anything after January, 2009. But be aware: If you apply that criteria, you also have to accept, for example, that employment growth was stronger under Jimmy Carter than it was under Ronald Reagan, and gross domestic product growth just as good. At least that is what the unlagged numbers indicate. Lagging them 12 months changes things.
Indeed, one can go through a whole series of economic indicators for presidents from Truman through Obama and lag them by zero, six, 12, 18 or 24 months and see how their calculated performance often flops back and forth from sterling to abysmal.
The fact obscured by all of this is that, in all cases, which president was in office and what policies he pursued is only one factor among many in how the economy performed. And in many cases it was a small factor.
Unfortunately, that is not what many want to hear. Many people want villains to blame and champions on white horses to save us. And some want quantitative proof of whom they are. Perhaps that is why we keep getting disappointing results.
For anyone who wants to play with how different lags affect the “performance” of different presidents, myriad economic indicators are available for easy download in spreadsheet format from the FRED database at the St. Louis Federal Reserve, at http://research.stlouisfed.org/fred2/. If you are not familiar with such indicators, Gary Clayton and Martin Giesbrechts’s “A Guide to Everyday Economic Statistics” is a wonderful, compact reference.