Liberals like to think they have no bias, that they base their positions on evidence, following the data wherever it leads. But the real world is so complicated that they often trick themselves into discovering evidence for what they already believe, with the smug bonus of convincing themselves that it’s a “reality-based” position superior to their ideological opponents.
Betsey Stevenson and Justin Wolfers claim to have evidence that Republicans are more fiscally irresponsible than Democrats. I think it’s just evidence of their own biases.
It’s important to note that Betsey and Justin, while appearing to be objective academic economists, consistently display a liberal leaning. For example, Justin frequently tweets complaints about Romney’s vague and contradictory budget plans, while never criticizing or even defending Obama’s actual budgets which have repeatedly been unanimously rejected by both parties in Congress.
Justin also tweeted about 9 times in the last month regarding Obama’s InTrade price when it was rising (or about Romney’s when it was falling), but never when the trends were reversed, even though the overall price has been flat over the time period. Justin’s market focus appears to be more “reality-based” than conservative commentators who tweet only the good individual Romney polls, but his selectiveness is just as subjective.
A few months ago Betsey and Justin claimed to have evidence that the GOP’s debt ceiling fight last year hurt the economy. I believe their evidence was weak, cherry-picked, and contradictory. Alas, it looks like they’re still manufacturing evidence for their pre-existing bias that Republicans are worse than Democrats.
What is the evidence this time?
In this piece, Betsey and Justin compare the national debt when the presidency is held by Republicans and Democrats. I disagree with the entire premise, since Congress arguably has more control over the budget than the president, but let’s go with it for now. They look at market responses to changing expectations about election outcomes. It’s clever enough to convince Ezra Klein (who already shares the same pre-existing bias), but it doesn’t make much sense to me.
First they claim that interest rates go down when Democrats are expected to win, and vice versa:
Let’s take the 2004 election as a particularly stark case study… flawed exit-poll numbers suggested that John Kerry would win in a landslide. For the next few hours, financial markets believed that there would be a Democrat in the White House. Then, by late evening, the votes were counted, and it became clear that President George W. Bush, the Republican, had won re-election…
Yields on government bonds were lower during the brief period in which a Democrat was expected to be president, suggesting investors believed the Republican would increase the national debt…
This reaction has been typical in recent decades: A study of similar events… found that since 1980, bond yields have tended to rise on news that a Republican will be elected. The pattern held last week, when interest rates on government bonds increased slightly after Romney’s strong performance in the first presidential debate.
Never mind that Justin assured us after the debate that markets were not changing expectations about Obama’s victory. I can’t find intraday information for the 2004 election in question, but here’s a graph of government bonds from the last few months:
Can you spot the slight increase from “Romney’s strong performance”? I think the rise from 1.6% to 1.7% on the right end of the graph is supposed to be it, but just like with the consumer confidence charts a few months ago, Betsey and Justin are extremely confident they can identify a specific reason for one little squiggle on a big graph of squiggles. Taleb calls it the Narrative Fallacy; I also call it arrogance.
Nevertheless, Betsey and Justin conclude that “markets believe the modern Republican Party has abandoned its historical commitment to fiscal responsibility,” and furthermore, “they have been right.”
Now for the stock markets
Alright, now that we’ve identified precisely how bond markets respond to Republicans and concluded what that means about Republican fiscal responsibility, let’s look at how stock markets respond:
The same study found that stocks typically rise a percentage point or two on news that a Republican will be elected…
…investors’ preference for Republicans generally hasn’t been rewarded after presidents take office… the stock market rises more strongly under Democratic presidents. True to this pattern, the Standard & Poor’s 500 Index has gained more than 80 percent during the tenure of President Barack Obama.
Let’s ignore the completely-out-of-context jab at the end, unless you believe that George W. Bush was personally responsible for the collapse of global markets at the end of his term that allowed for the stock market to rise 80% the next four years and still be below levels five years prior. The overall point is that markets believe Republicans will be better for the stock market, but they’re actually wrong.
Now wait a minute! I thought we just got done paying attention to expectations in the bond markets because it told us something about Republicans. This is a classic example of trying to have it both ways: When investors expect Republicans to increase the debt, it’s evidence that Republicans are bad because the investors were right. When investors expect Republicans to increase the stock market, it’s evidence that Republicans are bad because the investors were wrong! If we’re just going to look at the final data to decide if Republicans are good or bad, why are we bothering to look at investor expectations at all?
More fundamentally, the whole discussion implies that bond markets and stock markets are separate things, and that we can discern reasons for movements in either one. But don’t stocks and bonds often move inversely as investors switch from one to the other? It looks like the Dow Jones dropped 100 points from its high on Election Day 2004, and then added 100 points the next day. How can we know if the movement in bonds was a direct reaction to expectations that a Republican president would increase the national debt, or an indirect reaction to expectations that a Republican president would increase stocks? We can’t.
But what about the debt data?
Still, even if we throw out the weak, cherry-picked, and contradictory evidence about market expectations, we are still left with the actual data that shows debt as a percentage of GDP expanding under almost all Republican presidents and shrinking under almost all Democratic presidents. I mean, just look at the way the green numbers line up with the blue presidents on this chart – well, at least until Obama.
Betsey and Justin note that the pattern only holds “before the last recession.” Here is another example of trying to have it both ways. When the stock market rises under Obama, they credit that as being “true to the pattern.” But when debt as a percentage of GDP rises under Obama, well, let’s just ignore that because it’s not true to the pattern.
But our sample size is already pretty small, and once we start throwing out outliers, it gets even smaller. Here’s another: debt as a percentage of GDP was flat during Clinton’s first four years and only started dropping in the second half, during which Republicans also controlled Congress. Like all history, the details of that time period are complex, but I think any fair study would conclude that the Republicans were at least as responsible for the reduction as Clinton, if not more; some paint Gingrich and company as the leading heroes.
If we throw out the Obama years because they break the pattern, and the Clinton years because Republicans were jointly responsible, now we’re really running thin on evidence that Democrats are better for the national debt than Republicans – especially for any time period recent enough to matter.
This doesn’t mean that Republicans are any better. As Doug Mataconis says,
The last time we had Republicans in charge of the Executive and Legislative Branches, we ended up with a trillion dollar unfunded entitlement, a massive increase in the Federal education bureaucracy, and two wars that we not only didn’t pay for but which coincided with tax cuts which is about the fiscally dumbest thing you can do.
But it doesn’t mean Republicans are objectively any worse, either. I think the evidence shows that both parties have been fiscally irresponsible since 2000. Considering the mock horror Republicans show at merely reducing the rate of increase in military spending, and the mock horror Democrats show at reducing the rate of increase in spending on anything else, I don’t think either party looks very fiscally responsible, though I’m optimistic that one or both of them will prove me wrong.
You can always torture numbers from the past to prove that your team is better than the other team. But the only “reality” I’m interested in is the reality of mathematics, and one way or another, it’s going to win in the end.
Follow @postlibertarian on Twitter to keep up with new posts...