6 ms·

Note that "inflation" and "inflation adjusted" are not what you'd expect. http://www.financialsense.com/contributors/lance-roberts/inflation-different-than-rea

Note that "inflation" and "inflation adjusted" are not what you'd expect. http://www.financialsense.com/contributors/lance-roberts/inflation-different-than-reality http://www.financialsense.com/contributors/lance-roberts/inf... This is a natural consequence of tying social security (and other benefits) to inflation and Goodhart's Law: https://en.wikipedia.org/wiki/Goodhart%27s_law https://en.wikipedia.org/wiki/Goodhart%27s_law I've been meaning to do an analysis of inflation based on hours worked. It;s hard to redefine hours but I'm sure politicians will find a way!

greggyb 10y ago
> I've been meaning to do an analysis of inflation based on hours worked. It;s hard to redefine hours but I'm sure politicians will find a way! The cynic in me says that they simply record fractional portions of FTE and FTE takes on different values over time and across industries.
conistonwater 10y ago
The financialsense link looks a bit dodgy to me. The first thing that pricks me about it is the way he complains about some change that happened in BLS's CPI calculation in 1997 (which is described here: http://www.bls.gov/cpi/cpigmrp.htm http://www.bls.gov/cpi/cpigmrp.htm). His response: But the manipulation of the data did not stop there. He seems to think that any change, no matter if it's justified on methodological grounds, is "manipulation". I imagine he highlighted this change because the shift after the change in methodology was down. He's entitled to his opinion, of course, but that doesn't make him correct. His discussion of core CPI and the remark that "average consumers who quickly point to the fact the food and energy are big part of their daily lives" is also wrong, and the reason for using core CPI is usually discussed in textbooks (e.g., http://krugman.blogs.nytimes.com/2010/02/26/core-logic/ http://krugman.blogs.nytimes.com/2010/02/26/core-logic/). It has to do with the fact that people need to predict future price levels when signing current contracts, and thus having a volatile measure of inflation would not be a good thing. For an example of how methodological issues can remain for a long time without being fixed, consider that the Dow Jones Industrial Average index is still an arithmetic average of share prices, not weighted by market capitalization like other indices. So fixing methodological issues is something worthwhile, and should be rewarded, not denounced as some kind of political manipulation.
- SilasX 10y ago
  If you're only attenuating inflation signals because of quality improvements, but never amplifying them because of product debasements, then yes, it would be fair to write that off as (politically driven?) manipulation. Btw, here's a particularly egregious use of the logic: "oh, your cost of living isn't going up, because iPads are faster." http://www.reuters.com/article/us-usa-fed-dudley-ipad-idUSTRE72A4D520110311 http://www.reuters.com/article/us-usa-fed-dudley-ipad-idUSTR...
  - rz2k 10y ago
    There's nothing weird about the logic at all. It doesn't take much sophistication to recognize that inflation measuring the cost of a basket of goods is affected when the goods change either through substitution or changes in quality. Incidentally, the larger category encompassing computers is 0.29% of the calculation of CPI, and it has only dropped to about 44% of prices in 2007. There absolutely are problems with policy makers trying to starve the statistical agencies of funding for political reasons, but the actual public servants and economists conducting the work are pretty boring about just trying to model and describe the real economy. It's even worth noting that possible improvements to the accuracy of CPI have been aborted due to political reasons, but the problem was that people's social security checks would have been revised down unless the existing law was changed.
    - SilasX 10y ago
      > It doesn't take much sophistication to recognize that inflation measuring the cost of a basket of goods is affected when the goods change either through substitution or changes in quality. In theory, yes. It takes a little more sophistication to realize that, in practice, "the iPad being faster" doesn't actually cancel the burden of higher food prices.
      - rz2k 10y ago
        I guess I'm not that sophisticated then. :) It seems unavoidable that how much nonfood I get per dollar affects the burden incurred by a dollar spent on food. One could consider how spending on different areas has changed over time. In nominal terms, while we make about 48 times what we did in 1955, we now spend about 17 times what we did on food and on clothing, 56 times what we did on housing, and well over 200 times on healthcare. [1] Considering that we spend about 70 times what we did on large entertainment goods such as televisions and sailboats, does that mean we are able to spend less on other areas? Does it mean that even though we used to have to spend a lot more of our income to own a TV we need to buy more stuff to be entertained? Anyway, I don't really understand what the crowd in the article really wanted. Are they ultimately arguing that he should vote to raise rates, because there is inflation that he's missing by using the wrong indicators? I'm guessing that they're populists since an explanation made them angry, but googling doesn't tell me that populists have a consensus opinion about raising or lowering rates. If they want another indicator to be used as a basis for official cost of living calculations for things like federal wages and social security payment increases, they're getting worked up at the wrong person. [1] https://fred.stlouisfed.org/graph/?g=792u https://fred.stlouisfed.org/graph/?g=792u
dredmorbius 10y ago
I've been thinking of better models for inflation. HH debt as a percentage of income strikes me as a possibly useful metric. It indicates the extent to which expenses are outstripping income, though some of that may be due to demographic effects, e.g., student loans (for a young-skewed population) and medical or nursing home expenses (for an old-skewed one).

greggyb 10y ago

> I've been meaning to do an analysis of inflation based on hours worked. It;s hard to redefine hours but I'm sure politicians will find a way! The cynic in me says that they simply record fractional portions of FTE and FTE takes on different values over time and across industries.

conistonwater 10y ago

The financialsense link looks a bit dodgy to me. The first thing that pricks me about it is the way he complains about some change that happened in BLS's CPI calculation in 1997 (which is described here: http://www.bls.gov/cpi/cpigmrp.htm http://www.bls.gov/cpi/cpigmrp.htm). His response: But the manipulation of the data did not stop there. He seems to think that any change, no matter if it's justified on methodological grounds, is "manipulation". I imagine he highlighted this change because the shift after the change in methodology was down. He's entitled to his opinion, of course, but that doesn't make him correct. His discussion of core CPI and the remark that "average consumers who quickly point to the fact the food and energy are big part of their daily lives" is also wrong, and the reason for using core CPI is usually discussed in textbooks (e.g., http://krugman.blogs.nytimes.com/2010/02/26/core-logic/ http://krugman.blogs.nytimes.com/2010/02/26/core-logic/). It has to do with the fact that people need to predict future price levels when signing current contracts, and thus having a volatile measure of inflation would not be a good thing. For an example of how methodological issues can remain for a long time without being fixed, consider that the Dow Jones Industrial Average index is still an arithmetic average of share prices, not weighted by market capitalization like other indices. So fixing methodological issues is something worthwhile, and should be rewarded, not denounced as some kind of political manipulation.

SilasX 10y ago
If you're only attenuating inflation signals because of quality improvements, but never amplifying them because of product debasements, then yes, it would be fair to write that off as (politically driven?) manipulation. Btw, here's a particularly egregious use of the logic: "oh, your cost of living isn't going up, because iPads are faster." http://www.reuters.com/article/us-usa-fed-dudley-ipad-idUSTRE72A4D520110311 http://www.reuters.com/article/us-usa-fed-dudley-ipad-idUSTR...
- rz2k 10y ago
  There's nothing weird about the logic at all. It doesn't take much sophistication to recognize that inflation measuring the cost of a basket of goods is affected when the goods change either through substitution or changes in quality. Incidentally, the larger category encompassing computers is 0.29% of the calculation of CPI, and it has only dropped to about 44% of prices in 2007. There absolutely are problems with policy makers trying to starve the statistical agencies of funding for political reasons, but the actual public servants and economists conducting the work are pretty boring about just trying to model and describe the real economy. It's even worth noting that possible improvements to the accuracy of CPI have been aborted due to political reasons, but the problem was that people's social security checks would have been revised down unless the existing law was changed.
  - SilasX 10y ago
    > It doesn't take much sophistication to recognize that inflation measuring the cost of a basket of goods is affected when the goods change either through substitution or changes in quality. In theory, yes. It takes a little more sophistication to realize that, in practice, "the iPad being faster" doesn't actually cancel the burden of higher food prices.
    - rz2k 10y ago
      I guess I'm not that sophisticated then. :) It seems unavoidable that how much nonfood I get per dollar affects the burden incurred by a dollar spent on food. One could consider how spending on different areas has changed over time. In nominal terms, while we make about 48 times what we did in 1955, we now spend about 17 times what we did on food and on clothing, 56 times what we did on housing, and well over 200 times on healthcare. [1] Considering that we spend about 70 times what we did on large entertainment goods such as televisions and sailboats, does that mean we are able to spend less on other areas? Does it mean that even though we used to have to spend a lot more of our income to own a TV we need to buy more stuff to be entertained? Anyway, I don't really understand what the crowd in the article really wanted. Are they ultimately arguing that he should vote to raise rates, because there is inflation that he's missing by using the wrong indicators? I'm guessing that they're populists since an explanation made them angry, but googling doesn't tell me that populists have a consensus opinion about raising or lowering rates. If they want another indicator to be used as a basis for official cost of living calculations for things like federal wages and social security payment increases, they're getting worked up at the wrong person. [1] https://fred.stlouisfed.org/graph/?g=792u https://fred.stlouisfed.org/graph/?g=792u

dredmorbius 10y ago

I've been thinking of better models for inflation. HH debt as a percentage of income strikes me as a possibly useful metric. It indicates the extent to which expenses are outstripping income, though some of that may be due to demographic effects, e.g., student loans (for a young-skewed population) and medical or nursing home expenses (for an old-skewed one).