Different Types of Incorrect Predictions
-- Posted by Neil H. Buchanan
My first order of business today is to concede -- happily -- that my prediction in last Friday's Dorf on Law post turned out to be wrong: "There will be at least one state (probably Ohio) in which there is a mysterious difference in the official count and the best forecasts before, and on, election day." As it turned out, either the armed standoff between the two parties that I discussed later in that post has now reached equilibrium (wasteful though it is), or the election was not close enough to steal. In the latter case, it would have been unwise for anyone to try to intimidate enough voters, hack enough electronic machines, and so on to try make a difference (and to be caught doing it), because the election was really out of reach.
We do know, of course, that there were still instances in which Democratic-leaning precincts ended up with extremely long waiting times, that various dirty tricks were employed (robo-calls to elderly voters saying that they could vote over the phone, or the next day), and so on. This must change. But what I called "not a bold prediction, by any means," turned out to be wrong. Good news.
[As a minor secondary matter, I also wanted to mention that, as I re-read last Friday's post, I was stunned to see that I had sincerely and accurately penned the following sentence: "Everything in downtown Toledo appeared to be running smoothly." I guarantee you that I have never written or said anything like that in my life, nor do I expect to do so again! Toledo continues to struggle, not quite coming back from its rust-belt status in the way that some other Midwestern cities have -- although it is doing much better than, say, Detroit and Buffalo, as far as I can tell. Especially because of that, it was impressive to see the early voting apparatus apparently working so well two weeks ago. I'm a Toledo kid at heart, but my love for the place is not blind.]
As far as conceding incorrect predictions goes, I can also compare my completely incorrect prediction about this election with another incorrect prediction that I made two-and-a-half years ago. In "Double Dip a Done Deal. D'Oh!," after conceding that macroeconomic forecasting is a dangerous business, I nonetheless dove in head-first and stated that "all of the pieces now seem to be in place for what we have feared all along: a policy-induced relapse into a second bout of recession, without ever having emerged from the Great Recession. In this post, I will explain why this sorry outcome now seems inevitable, what could be done to stop it, and what should be done after we fail to stop it."
Two-plus years is more than enough time to test such a prediction. Again, I was wrong. And again, I was happy to be wrong. The unemployment rate has slowly declined in the ensuing months and years, job growth has been positive, and the weak economy was not quite weak enough to sink a sitting President's re-election bid. My prediction, to repeat it one more time, was wrong. In this case, however, the nature of the error is different.
My prediction on Tuesday was essentially that something really big was going to happen on Election Day, along the lines of what happened in Florida in 2000 or Ohio in 2004, where the evidence of what people intended to do (and, often, what they thought they had done) simply did not comport with the official vote counts. Even though I did not predict that the election itself would be stolen, I did foresee at least one state's outcome being obviously "rigged" in some meaningful sense. There is no evidence that this happened, even to a middling degree.
By contrast, the difference between a double-dip recession and what we have experienced over the 2009-12 period is a matter of relatively minor degree. The imperfect-but-workable definition of a recession is that the economy (as measured by inflation-adjusted GDP) shrinks for two consecutive quarters. That did not happen in the last two years, which is good. The difference between declining GDP and barely-rising GDP is, however, not especially meaningful, because it imposes an arbitrary dividing line that no one really cares about. For example, two consecutive quarters of zero percent growth is not really different from two consecutive quarters of 0.1% declines in GDP. This is, therefore, more a matter of labeling than anything else. Either way, the things that I focused on in 2010 -- problems in Europe (affecting our exports), the end of stimulus spending, consumer worries, and lack of business confidence -- were very real, leading to two years of lost economic output, along with the human misery that is so often ignored in the sanitized world of GDP analysis.
In short, if I was going to be wrong about my double-dip prediction, I would have preferred to be really, really wrong. The small difference between the prediction and the reality was welcome, but for the believers in the Confidence Fairy to have been right, we should have seen a lot of trickling down over the last two years, which we have not.
Now, we face another economic moment of truth. As soon as the election ended, attention turned to the so-called "Fiscal Cliff," which amounts to a large collection of spending cuts and tax increases that are hard-wired to hit on January 1, 2013, unless the law is changed before then. (Professor Dorf's Verdict column next week will discuss this in more detail.) Changing the law, however, will require compromise between Republicans in the House and Democrats in the Senate and the White House.
Amazingly, the Democrats have already blinked. Senator Schumer of New York, who is considered to be one of the Senate's leading liberals, just announced that Democrats will give up on the President's central talking point for the last five years, which is that the income tax rates on the upper brackets must return to their levels during the 1990's. Now, apparently, the Democrats are saying that they could be fine sticking with the current (supposedly temporary) 35% top rate, and then doing a Romney-like (but non-imaginary) base broadening move to increase revenues from upper-income taxpayers.
Maybe this is good politics, but it is bad economics. The one thing that we could do to appease the Confidence Fairy at this point would be to raise more revenues from wealthy people, whose spending would be largely unaffected by tax increases (as a CBO report this week confirmed). Moreover, the Congressional Research Service's recent report (that became a campaign issue, very briefly) suggests that raising rates will not reduce economic growth So, giving up on truly progressive tax changes at the start of negotiations strikes me as quite foolish. Again, however, I have tenure, and do not have to run for re-election.
Three years ago, as part of the Bush tax cut package, we were nearing the point where the estate tax was going to go away for the 2010 calendar year, and then to return in 2011. I -- along with every tax expert I knew -- said that this would never, ever happen. We were wrong, and the politics have become even crazier since then. No one knows what will happen. If early indications are any guide, however, the outcome is going to be very disappointing to liberals, and very congenial to conservatives. Is that a prediction? Absolutely not.
My first order of business today is to concede -- happily -- that my prediction in last Friday's Dorf on Law post turned out to be wrong: "There will be at least one state (probably Ohio) in which there is a mysterious difference in the official count and the best forecasts before, and on, election day." As it turned out, either the armed standoff between the two parties that I discussed later in that post has now reached equilibrium (wasteful though it is), or the election was not close enough to steal. In the latter case, it would have been unwise for anyone to try to intimidate enough voters, hack enough electronic machines, and so on to try make a difference (and to be caught doing it), because the election was really out of reach.
We do know, of course, that there were still instances in which Democratic-leaning precincts ended up with extremely long waiting times, that various dirty tricks were employed (robo-calls to elderly voters saying that they could vote over the phone, or the next day), and so on. This must change. But what I called "not a bold prediction, by any means," turned out to be wrong. Good news.
[As a minor secondary matter, I also wanted to mention that, as I re-read last Friday's post, I was stunned to see that I had sincerely and accurately penned the following sentence: "Everything in downtown Toledo appeared to be running smoothly." I guarantee you that I have never written or said anything like that in my life, nor do I expect to do so again! Toledo continues to struggle, not quite coming back from its rust-belt status in the way that some other Midwestern cities have -- although it is doing much better than, say, Detroit and Buffalo, as far as I can tell. Especially because of that, it was impressive to see the early voting apparatus apparently working so well two weeks ago. I'm a Toledo kid at heart, but my love for the place is not blind.]
As far as conceding incorrect predictions goes, I can also compare my completely incorrect prediction about this election with another incorrect prediction that I made two-and-a-half years ago. In "Double Dip a Done Deal. D'Oh!," after conceding that macroeconomic forecasting is a dangerous business, I nonetheless dove in head-first and stated that "all of the pieces now seem to be in place for what we have feared all along: a policy-induced relapse into a second bout of recession, without ever having emerged from the Great Recession. In this post, I will explain why this sorry outcome now seems inevitable, what could be done to stop it, and what should be done after we fail to stop it."
Two-plus years is more than enough time to test such a prediction. Again, I was wrong. And again, I was happy to be wrong. The unemployment rate has slowly declined in the ensuing months and years, job growth has been positive, and the weak economy was not quite weak enough to sink a sitting President's re-election bid. My prediction, to repeat it one more time, was wrong. In this case, however, the nature of the error is different.
My prediction on Tuesday was essentially that something really big was going to happen on Election Day, along the lines of what happened in Florida in 2000 or Ohio in 2004, where the evidence of what people intended to do (and, often, what they thought they had done) simply did not comport with the official vote counts. Even though I did not predict that the election itself would be stolen, I did foresee at least one state's outcome being obviously "rigged" in some meaningful sense. There is no evidence that this happened, even to a middling degree.
By contrast, the difference between a double-dip recession and what we have experienced over the 2009-12 period is a matter of relatively minor degree. The imperfect-but-workable definition of a recession is that the economy (as measured by inflation-adjusted GDP) shrinks for two consecutive quarters. That did not happen in the last two years, which is good. The difference between declining GDP and barely-rising GDP is, however, not especially meaningful, because it imposes an arbitrary dividing line that no one really cares about. For example, two consecutive quarters of zero percent growth is not really different from two consecutive quarters of 0.1% declines in GDP. This is, therefore, more a matter of labeling than anything else. Either way, the things that I focused on in 2010 -- problems in Europe (affecting our exports), the end of stimulus spending, consumer worries, and lack of business confidence -- were very real, leading to two years of lost economic output, along with the human misery that is so often ignored in the sanitized world of GDP analysis.
In short, if I was going to be wrong about my double-dip prediction, I would have preferred to be really, really wrong. The small difference between the prediction and the reality was welcome, but for the believers in the Confidence Fairy to have been right, we should have seen a lot of trickling down over the last two years, which we have not.
Now, we face another economic moment of truth. As soon as the election ended, attention turned to the so-called "Fiscal Cliff," which amounts to a large collection of spending cuts and tax increases that are hard-wired to hit on January 1, 2013, unless the law is changed before then. (Professor Dorf's Verdict column next week will discuss this in more detail.) Changing the law, however, will require compromise between Republicans in the House and Democrats in the Senate and the White House.
Amazingly, the Democrats have already blinked. Senator Schumer of New York, who is considered to be one of the Senate's leading liberals, just announced that Democrats will give up on the President's central talking point for the last five years, which is that the income tax rates on the upper brackets must return to their levels during the 1990's. Now, apparently, the Democrats are saying that they could be fine sticking with the current (supposedly temporary) 35% top rate, and then doing a Romney-like (but non-imaginary) base broadening move to increase revenues from upper-income taxpayers.
Maybe this is good politics, but it is bad economics. The one thing that we could do to appease the Confidence Fairy at this point would be to raise more revenues from wealthy people, whose spending would be largely unaffected by tax increases (as a CBO report this week confirmed). Moreover, the Congressional Research Service's recent report (that became a campaign issue, very briefly) suggests that raising rates will not reduce economic growth So, giving up on truly progressive tax changes at the start of negotiations strikes me as quite foolish. Again, however, I have tenure, and do not have to run for re-election.
Three years ago, as part of the Bush tax cut package, we were nearing the point where the estate tax was going to go away for the 2010 calendar year, and then to return in 2011. I -- along with every tax expert I knew -- said that this would never, ever happen. We were wrong, and the politics have become even crazier since then. No one knows what will happen. If early indications are any guide, however, the outcome is going to be very disappointing to liberals, and very congenial to conservatives. Is that a prediction? Absolutely not.