Depending on who you talk to (the federal government or a like-minded individual), the Consumer Price Index (CPI) is either an accurate or inaccurate measure of price inflation, respectively. What is important to understand is that each side stands to benefit from its position. The government tends to understate inflation while a like-minded individual tends to overstate inflation.
In regards to the government, the dollar amount of entitlements doled out by programs such as social security are affected by cost-of-living adjustments, which supposedly reflect an increase (or decrease) in the cost-of-living. As such, the government tends to understate price inflation to save itself money, thereby reducing the cost-of-living adjustment.
In regards to the individual (at least to those who visit this website), price inflation is a constant concern. After all, individuals who understand sound money and free markets know prices should come down over time. However, due to the Federal Reserve (primarily), this is not the case. Thus, we warn against the coming price inflation, and in my opinion, tend to overstate price inflation (without empirical evidence) as a means of reinforcing our argument.
Thus, I believe it is in our best interest to understand (to the best of our ability) the CPI, which is a measure of price inflation calculated by the federal government. Until we truly understand how their statistics are computed, our concerns will not be taken seriously.
On April 14, 2010, the monthly CPI was released. The headline numbers were as follows:
|CPI – M/M change||0.0%||0.1%|
|CPI – Y/Y change||2.2%||2.4%|
|CPI less food & energy – M/M change||0.1%||0.0%|
|CPI less food & energy – Y/Y change||1.3%||1.2%|
In the table above, M/M refers to month-over-month while Y/Y refers to year-over-year. The prior column refers to the CPI published the month before. To me, these numbers seem ridiculously row (and run counter to real life experience), but why?
An explanation may be derived from the CPI report itself. If you go to the last page of the report, you will see different groups of items (e.g., food and beverages, housing) that are weighted differently. Each group is then divided into categories that are weighted differently. These categories are further divided into subcategories that are weighted differently. So on and so forth.
The category, subcategory, and sub-subcategory I would like to focus on is housing, shelter, and owners’ equivalent rent of residences, respectively; this can be found on page 9. In regards to the CPI, the relative importance of housing (in December 2009) was 41.960. If you sum the relative importance of each group, the sum is 100. Thus, housing is considered (by the BLS) to be a large factor in price inflation. Fair enough.
The sum of the relative importance of each subcategory of housing is equal to the relative importance of housing (i.e., 32.289+5.081+4.590=41.960), as illustrated below:
|Subcategory||Relative Importance (December 2009)|
|Fuels and utilities||5.081|
|Household furnishings and operations||4.590|
Additionally, the sum of the relative importance of each sub-subcategory of shelter is equal to the relative importance of the subcategory of shelter (i.e.,5.966+0.769+25.206+0.347=32.288) , as illustrated below:
|Sub-subcategory||Relative Importance (December 2009)|
|Rent of primary residence||5.966|
|Lodging away from home||0.769|
|Owners’ equivalent rent of residences||25.206|
|Tenants’ and household insurance||0.347|
Careful calculation reveals a discrepancy of 0.001; this is likely due to rounding error.
Owners’ equivalent rent of residences (OER) has a relative importance of 25.206 in regards to the total relative importance (i.e., 100). But, what is owners’ equivalent rent of residences? According to Investopedia, OER is “the amount of rent that could be paid to substitute a currently owned house for an equivalent rental property … [it] is the amount a homeowner would pay to rent or would earn from renting his or her home in a competitive market.”
Contrary to popular belief, rents are not primarily a function of housing prices. Instead, rents are primarily a function of wages. Peter Schiff and others have argued this in the past (personally, I believe they are correct). This is illustrated by the figure below (as well as common sense):
The National Case-Shiller Index measures pricing in the residential housing market. Please see the figure below for wages:
As evidenced by the two figures, rents and wages have acted similarly while rents and housing prices have not.
One notable aspect of the OER vs. National Case-Shiller Index figure is the divergence of the latter from the former. From 1999 to 2006, housing prices increased on average over 10% a year while wages increased on average only 1% a year.
Herein lies one of the many problems with the CPI. The CPI essentially understates price inflation for housing during booms (since price inflation isn’t meaningfully incorporated into the OER). However, at this point in time, the decline in OER is assisting the federal government in underreporting price inflation.
Additionally, since the relative importance of housing in the CPI is so large, inflation appears to be more subdued than it really is. A closer inspection of the other groups yields the following information:
|Group||Relative Importance||M/M Change||Y/Y Change|
|Food and beverages||14.795||0.2%||0.3%|
|Education and Communication||6.434||0.1%||2.4%|
|Other Goods and Services||3.483||0.2%||4.9%|
If you look closely at page 10 of the report under the heading Special indexes, you will see an index entitled All items less shelter. This index shows that price inflation is much higher than the CPI numbers (including food and energy, unadjusted) of 0.4% and 2.3% for M/M and Y/Y, respectively. If shelter is excluded, the actual M/M and Y/Y rates are 0.6% and 3.8%, which are hardly subdued, considering the state of our economy.
In conclusion, individuals who are concerned about price inflation should pay close attention to both the headline numbers and the All items less shelter index.Published in