28 October 2010

Predictions Plus-Minus Update

Crystal Ball Earth It's time once again for our quarterly accounting for how all the predictions we've made since January 2008 have turned out, where we employ the most brutal, unforgiving metric ever-devised for performance: the plus-minus statistic from hockey and basketball!

Using this system, we gain a point if we're right, lose a point when we're wrong, and score a zero for when the outcome of a prediction either cannot yet be determined or, in the case where we make multiple predictions, where our contrary results cancel each other out.

Ultimately, the plus-minus score counts the number of times our predictions were right more than they were wrong. If we're no better at predicting the future than a coin toss, our plus-minus score will gravitate toward zero. If we're better at predicting the future than that, our score will rise over time. And if we're really bad at prognostication, our score will surely fall.

Three months ago, our plus-minus score was +22. Here's how we stand today:

  • Number of Predictions Made Since January 2008: 101
  • Number of Correct Predictions: 65
  • Number of Incorrect Predictions: 30
  • Number of Outcomes Not Yet Determined: 11

Overall, for the predictions where we've been able to establish the outcomes to date, our plus-minus score has risen to +35. Measured as a percentage, our prediction accuracy rate is 68.4%.

The table below updates the status of all the predictions that we've successfully determined the outcomes during the last three months, as well as those for which we have no determination has yet been made. The blow-by-blow commentary is just a bonus that you'll hopefully find to be entertaining!

Political Calculations' Plus-Minus Score Update, 29 October 2010
Date Prediction Outcome +/- Score
15 January 2008 An unexpected surge of site traffic in the week before the 2010 elections led us to this long-forgotten prediction from January 2008, where we presented a tool that projected that the U.S. National Average Wage Index in 2007 would be $39,273, but we anticipated that it would be much closer to $40,000. According to Social Security, the average income in the U.S. in 2007 was $40,405. Side Note: Our tool's forecast for 2008's average income was $40,523 (actual was $41,334, an error of 2%) and for 2009, we forecast $41,794 (actual was $40,935, also an error of 2%). We anticipate that the average income in 2010 will be more than 2% lower than our tool's forecast of $43,085. +1
13 August 2009 We make fun of 47 economists for predicting that the U.S. recession would end in the third quarter, based on the "Cash for Clunkers" program. We point out that the dividend futures data for the S&P 500 has been saying the recession would be over in 2009Q3 for months, long before C4C even became legislation! Too soon to tell. This prediction looks pretty likely, but we'll have to wait for the NBER to get around to declaring the date of the end of the recession. Update: Looking more and more likely. Update 2: They finally got around to it, saying the bottom of the recession occurred in June 2009, so all of 2009Q3 was recession free! +1
10 December 2009 We said: "If the Obama administration holds to the recent practice of U.S. politicians of keeping near the lower margin, the expected national debt burden for 2011 of 2.9 indicates that the top marginal income tax rate will increase to roughly 44-45%." Alan Viard did the tax rate math analysis of the Obama administration's legislative accomplishments, finding that "virtually all of top earners’ ordinary income will be taxed at 44.6 percent, starting in 2013." +1
21 December 2009 Using incomplete data for the month of December 2009, economy would dip in the second quarter of 2010, with a slow recovery afterward. We anticipate that meaningful growth in the number of jobs would likely begin with the third and fourth quarters of 2010. We anticipate that the NBER will declare the recession they found to have begun in December 2007 to have ended in the third quarter of 2009, but we make a case for 2010Q2 as a more realistic alternate. Too soon to tell. It will be a while before we get a full confirmation for these predictions. On the potential plus side for us, different branches of the Federal Reserve have used their own models for predicting what the NBER will do to find that July 2009 is the month they will most likely declare to be the ending date for the recession. Update: We were a bit off in calling July 2009 as the end of the recession, since the NBER marked June 2009 as the bottom, but we nailed 2010Q2 as being exceptionally slow, making this batch of predictions a split decision so far. We're still waiting to see if meaningful growth in the number of jobs happens through the fourth quarter, but so far, we're on track with that last part of the prediction as the number of people counted as having jobs has risen since July 2010. +0
14 July 2010 We make two predictions for the 2010 major league baseball season: The Arizona Diamondbacks will lose around 100 games and the New York Yankees will end the season with about 105 wins. We'll call these predictions correct if we're within 2 games of those numbers of wins or losses in the final regular season standings. What can we say? We do some predictions just for fun! Update: But clearly, we shouldn't! As it happened, the D-Backs lost 97 games, just one less than what it would have taken to score a point for that prediction. Meanwhile, we finally understand the reason why those second-half of the season underperforming, pin-striped weenies are called "Damn Yankees!" -2
30 July 2010 When they announced the advance GDP figure for 2010Q2, the BEA also revised GDP data going back to January 2007. Since our GDP forecasting technique uses some of that data, we updated it to incorporate that data, now predicting that real GDP for 2010Q2 will finalize at 13,281.6 billion in constant 2005 U.S. dollars! The advance data figure was within 0.5% of the target based on the BEA's revised data, so things are looking very good. But then, we already scored it as a hit. That we got much closer than the 1.2% we calculated based on the pre-revised data forecast just makes it taste sweeter!.... We'll update this entry when the GDP data for 2010Q2 is finalized in September 2010. Update: The final estimate of GDP in 2010-Q3 was $13,194.9, just 0.7% away from a perfect bulls-eye! +1
30 July 2010 When we updated our forecast for where the finalized GDP for 2010Q2, we snuck in a prediction for 2010Q3 in our chart. Although we expect our greatest accuracy when we used the newest, finalized GDP data for the preceding quarter, we're testing out how well projecting the next quarter's GDP based on the advance release data works out! We'll see if we're anywhere close in October 2010, and won't know for sure until December 2010. We'll also offer a more routine prediction when the 2010Q2 data is finalized in September 2010. Update: The clock is still ticking with the advance release GDP data coming out on Friday, 29 October 2010. +0
25 August 2010 Can we predict what average health insurance premiums will turn out to have been in 2010? If single coverage falls between $5,098 and $5,265 and family coverage falls between $14,166 and $14,452, then yes, we can! We won't know until next year when the Kaiser Family Foundation releases its 2010 Annual Survey of Employer Health Benefits. +0
1 September 2010 We anticipate that the average of the S&P 500's daily closing prices throughout the month of September 2010 will fall between 1017 and 1071. The market changed directions on us as the noise that began in April 2010 finally began dissipating. It might look like the S&P's closing the month with an average of 1122.08 is bad news for us, but it's actually great news because we can finally go back to using our direct method for forecasting stock prices! -1
2 September 2010 Maybe our most daring prediction ever. We revisit those ever "unexpected" weekly jobless claims numbers and offer a running prediction - one that continues week after week - anticipating that these new unemployment insurance claims filings will run betwen 415,000 and 515,000 indefinitely. (Note: These weekly figures have been both higher and lower than this range during the last four years.) Eight weeks at this writing. Eight hits. Although we might tweak our forecast in the days ahead, in our view, it will likely take an actual act of Congress to break the stagnant trend in jobs that we've identified. +8
28 September 2010 We create a tool that predicts that the average cost of tuition and required fees at a four-year institution of higher learning (aka "college") will rise to $14,541 in 2009, $15,394 in 2010, $15,869 in 2011, $15,538 in 2012 and $16,212 in 2013. We modify these predictions for 2009, 2010 and 2011, taking the "under" for each of these years. We would anticipate being within 2% of the values for 2012 and 2013. These predictions will take some time to play out. The confirmation will be provided by the annual Digest of Education Statistics produced by the U.S. Education Department. +0
5 October 2010 Using the finalized GDP data for the second quarter of 2010, we project GDP for the third quarter of 2010 will be within 2% of $13,284.3 billion chained 2005 U.S. dollars, giving nearly 70% odds that real GDP will actually be between $13,115 and $13,424. This is our "official" prediction for where real GDP will be in 2010-Q3, which we'll "officially" find out on 22 December 2010. Our first indication of how close we are will come as early as Friday, 29 October 2010 when the advance estimate of GDP is released by the BEA. +0
8 October 2010 Applying our tool that relates the rate of economic growth to the rate of unemployment in the U.S., we project that real GDP will be about $13,272.3 billion for 2010-Q3. Assuming that figure is correct (or nearly so), we use that result with our "official" method to project that the GDP growth rate for 2010-Q4 will fall from about 2.4% in the third quarter to 2.0%. To be determined. It's pretty interesting to have an entirely different method of projecting future GDP give such close results to our "official" approach, where we've consistently been within 2% of our target value. +0
25 October 2010 A week before the month to which it will apply even arrives, we forecast that the average of daily closing stock prices in November 2010 will be in a range between 1182 and 1218. With October 2010 very much on track to hit our target range for that month, this change would mark an upward move in stock prices. We went back over the data for the S&P 500 since January 1871 and found that on a month-to-month basis, "up" has happened some 56.1% of the time. Of course, that figure also means there's a 43.9% chance stocks will fall, so there's plenty of opportunity to be wrong. Right now, we're at the cusp for what direction the market will have moved a month from now, so here's hoping for a not-so-noisy month!... +0

Previously on Political Calculations

The following links will take you to our previous prediction outcome reports, which we've presented below in the order they've appeared here approximately every three months beginning with April 2009. You can get the most recent status updates by clicking the "track record" tag at the bottom of the post.

at