Hedonically-Adjusted, Well-Spun, Nominal Misery Index

Numerous pundits have taken to claiming that the Misery Index — a combination of Inflation measures and Unemployment — is still relatively low when compared with the 1970s.

They are full of shit. I have to call shenanigans on that foolishness. The way Inflation and Unemployment are measured today versus 30 years prior makes this an apples & oranges comparison. Merely showing 2008 versus 1973 is  nonsense (see the chart of the misery index, below via the WSJ).

Why? If we were measuring Inflation & Unemployment the way we did in the 1970s, we would see unemployment much closer to U6 Unemployment levels of 9.7% (versus the popularized headline inflation level, U3 now at 5.5%); the inflation measures would see an greater differential — CPI might be closer to 10+.

That would put the Misery Index somewhere between 17 and 21 — pretty close to the 1970s highs.

(If any of you chart wizards can figure out how to create a more accurate Misery Index, please tag me. I figure we can use U6 for Unemployment, but I am not sure where to get more accurate data history for CPI)


Miserable Index:

Chart courtesy of WSJ


Compare the above chart above with the two below:

Inflation, Pre-1983 Measures

Unemployment, Various Measures


courtesy of Harpers


Recession Fears Reignited
WSJ, June 7, 2008; Page A1

Print Friendly, PDF & Email

What's been said:

Discussions found on the web:
  1. Tim commented on Jun 9

    I can’t speak to changes in the data for unemployment; but the CPI, despite its shortcomings, is a more accurate index now than it it was in the 1970s. Consequently, I’m more likely to doubt the accuracy of the older data than the newer.

  2. bluestatedon commented on Jun 9

    “the CPI, despite its shortcomings, is a more accurate index now than it it was in the 1970s. Consequently, I’m more likely to doubt the accuracy of the older data than the newer.”

    Which has to mean, logically, that things were much less horrible under Carter than we’d all been led to believe.

  3. DojiStar commented on Jun 9

    shadowstats.com (and perhaps other places) continues to publish CPI calculated prior to various adjustments, like the ’90s revision. A subscription may be required, though, for access to the full data series.

  4. ames hubbard commented on Jun 9

    split the diff btwn shadow goverment statistics and the feds, i.e., 16+ and 3+. the truth is always between the extremes. that woud give approx 10%.

  5. Marcus Aurelius commented on Jun 9

    Posted by: bluestatedon | Jun 9, 2008 11:55:17 AM


    Great reply, and a gold star to you.

  6. Ed H commented on Jun 9

    Tim said:
    “can’t speak to changes in the data for unemployment; but the CPI, despite its shortcomings, is a more accurate index now than it it was in the 1970s. Consequently, I’m more likely to doubt the accuracy of the older data than the newer.”

    But, if you want apples to apples you need to make the stats the same across time periods. If we don’t have the datasets to retroactively adjust the data from earlier periods to today’s new formulation, we should at least add a ‘comparables’ number to today’s statistical formulations. That way, it’s an apples to apples comparison.

  7. cm commented on Jun 9

    As for propaganda in general, after 9/11 Bush delivered as big speech I believe in front of Congress. The way it was (apparently) scripted and the way it was televised — the format, not the content — eerily reminded me of Communist Party convention reporting back home. I talked to some coworkers also hailing from ex “communist” states, and they thought the same.

    We are also starting to see supply volatility in retail — empty or understocked store shelves, less fresh merchandise, increased bargain hunting (shelves cleaned of sale items sometimes the first sale day), and generally more disaffection and clock-watching in the workforce.

    With all due respect, I think those are things than many “Americans” cannot fully appreciate. You need to have lived there to see the parallels. (Not to claim a big insight of course.)

  8. bluestatedon commented on Jun 9

    MA, I have great confidence that somebody will argue very soon that the old CPI was accurate for Carter but the current one is accurate for Bush. Probably somebody holding down a desk at Fox. After all, consistency is the hobgoblin of small minds, right?

  9. Jim Haygood commented on Jun 9

    The source of the chart “Inflation, pre-1983 measures” is given in the Tampa Bay article as John Williams of Shadowstats.

    A smoothed index of y-o-y energy and food price increases (the very items excluded from the core-blimey CPI) would be a good substitute to measure consumer misery. These measures are in the same ballpark now as 1973-74 and 1980-81.

  10. dano commented on Jun 9

    Recession: When your neighbor loses his job.

    Depression: When you lose your job.

  11. Jack commented on Jun 9

    Please don’t make me pull out my old text books from the 70s, your explanation of major changes in weightings, variables, etc.? TIP pricing misleading?


  12. Markz commented on Jun 9

    I really despise comments that make broad assertions without any support. So, how exactly, do you support your claim that the CPI today better reflects true cost of living increases?

    Here’s the counter-argument:
    1) From everything I’ve read, the 1990s changes to incorporate hedonic substitution* and geometric vs. arithmetic weighting now severely understate the actual cost of living increases an average person feels.

    2) The increased focus on “core” inflation completely falls apart when there are sustained increases in food and energy prices.

    Here’s an interesting article with some more on these CPI issues:

    Shadow stats also has interesting data here:

    So, everything suggests the true CPI (i.e. cost of living changes as felt by an average US consumer) are probably understated by at least 1 or 2 points.

  13. RW commented on Jun 9

    I’ve seen some fairly strong arguments at Mark Thoma’s site that attempting to measure both inflation and cost-of-living with the same instrument is part of the problem but as far as I can tell (being a non-economist) a bigger problem may be that the dominant model of inflation emphasizes wages (and wage spiralling) as fundamental properties.

    “…somebody will argue very soon that the old CPI was accurate for Carter but the current one is accurate for Bush.”

    Could probably take that to the bank just like you could bank on arguments that the attempt to impeach Clinton was justified while impeaching Bush/Cheney is not but maybe, just maybe, now that that everyone has seen what a real failed presidency looks like we won’t have to hear about the “failed presidency of Jimmy Carter” as often.

  14. Nick commented on Jun 9

    Have to disagree with the CPI accuracy comment; the CPI is much less accurate today, and intentionally so. There are two major things which changed: Owner’s Equivalent Rent and Hedonic Regression. OER was a good change (IMHO); it flattens out the data somewhat, but generally doesn’t skew the CPI one way or the other over the long term. Hedonic Regression, on the other hand, allows the government to arbitrarily decrease the value of the CPI, in a purely bogus way.

    Hopefully at some point there will be enough public frustration with the obvious inaccuracy of the CPI to force a change, but I’m not holding my breath. Until then, a good rule of thumb seems to be adding 4% to the official number.

  15. Ed H commented on Jun 9


    I agree with you 100%. There has been a lot on the fake CPI stats since Bill Grooss came out against them.


    I also have a post that says similar things:


    The point being: if the stats today are better, then adjust the previous stats to reflect an apple to apples comparison. 11% CPI in 1981 is not the same as it is today.

  16. Michael Donnelly commented on Jun 9

    I remember the changes to the CPI, but I don’t recall any changes to unemployment. So why would that number change?

    U6 back in the day just would have been a lot higher than 9.7%

  17. Jim Haygood commented on Jun 9

    bart — Bingo! Exactly the chart Barry proposed, already done. And it shows that only 1973-74 and 1980-81 had higher misery spikes than today.

    If the chart extended back in time before 1970, a similar spike might be found in the inflationary, recessionary late 1940s. That’s the one my parents used to complain about. It whacked ’em when they were college students. So my mom’s sharply-tuned inflation antennae started detecting “deja vu” in 1971 and 1972, well before the crude hit the fan in October 1973.

  18. Some Jerk commented on Jun 9

    Dang, in the back-to-1900 chart, it’s clear why the 20s were roaring. Negative misery!

  19. bart commented on Jun 9


    That negative pain misery index in the ’20s blew me away too, I had to double check the data. In the 1921 recession, CPI bottomed at a monthly value of -15.8% in June and didn’t go positive until early 1923.

    Same here on the ’70s, my father picked it up due to having experienced Weimar Germany personally.


  20. Blissex commented on Jun 9

    UK inflation pressures are very very high:


    Note this beautiful quote:

    “To cut interest rates further, we suspect that the Bank will want clear evidence that wage moderation is continuing and that reduced demand is increasingly undermining companies’ pricing power,” said Howard Archer of Global Insight.

  21. ScottB commented on Jun 9

    Kevin Phillips writes: “The story starts after the inauguration of John F. Kennedy in 1961, when high jobless numbers marred the image of Camelot-on-the-Potomac and the new administration appointed a committee to weigh changes. The result, implemented a few years later, was that out-of-work Americans who had stopped looking for jobs — even if this was because none could be found — were labeled “discouraged workers” and excluded from the ranks of the unemployed, where many, if not most, of them had been previously classified.”

    That’s the basis for your claim that current unemployment is actually closer to U6. Have you verified Phillips’ assertion? I have looked through BLS methodology. While the definition of discouraged worker has changed over time, discouraged workers have never been counted as part of the unemployed. I’ll call BLS tomorrow to verify, but I’m pretty sure he is dead wrong on this.

  22. Don commented on Jun 9

    Anyone (including Barry) know where to get U-6 data before 1994? BLS doesn’t have it. What’s funny/ironic is that the employment rate we have now that omits “marginally attached” and “discouraged workers” was defined away during the Clinton Administration in October 1995.

    One other thing. U-6 in 1994 averaged 10.9. It averaged 9.7 in 1996. It’s 9.4 now. I was surprised that the employment situation was that bad during Clinton’s first administration and that as bad as things seem to be, they’re actually better now than at anytime during his first term.

    Not judging, or being a Fox shill. Just shocked that the U-6 is better now than it was from 1994-1996.

    If no source for U-6, I’ll try to recreate it for prior years if I can find data to reconstruct it.

  23. ken H. commented on Jun 9

    The other issue is that this is just getting going, early innings of pain to come (IMHO). This index is going to get worse.

    I appreciate CM’s post as well because in the 70’s journalists had some integrity. Today it’s all propaganda and pump. CNBC is embarrasing and I mostly watch for the information with the sound down. Unless BR’s on!, or Santelli. Watching Santelli brow beat Kneale is pure joy. I can just see Kneale stomping off camera crying.

  24. dave54 commented on Jun 10

    TIP’s are not trading based on inflation expectations; It’s oil/gold/dollar…TIP’s trade relative to low yielding [Goldilocks] stocks & bonds.

  25. bart commented on Jun 10

    Mea cupla Don, I blew it.

    That U6 data was not from economagic. It was from a proprietary reconstruction that was done by John Williams of shadowstats.com.

  26. Tim commented on Jun 10


    I have read the articles you suggest. However, I seem to remember the Blinder report indicating that the previously constructed index was overstated by more than 1 – 2% (3% seems to stick in my mind).

    As to the issue of core inflation – I didn’t bring it up and while I feel there’s some validity in tossing out the more volatile elements, that’s based on assumption that the volatility swings both ways. And, as I’m sure you’ll agree, we’ve not seen that in energy or food prices. Consequently, the focus on core seems, at least in this instance, misplaced.

    Thanks for your views.

  27. The Big Picture commented on Jun 10

    Enhanced Misery Index

    Yesterday, we asked what the Misery Index would look like if the CPI and Unemployment were reported honestly. A colleague at Credit-Suisse suggested adding in the annual change in house prices to the misery index. Suddenly, we can see that the consumer…

Posted Under