The following brief article was attached in a quick note from my better half, who commented Wonder how this comports with your pricing. The quick piece was a snarky bit from a writer who noted that we should all feel wealthier for the low food inflation as noted by the government’s CPI. As I was several days away from doing the pricing, I thought that it would be worthwhile to compare how the PracticalDad Price Index does actually compare with the monthly results of the CPI? More importantly, what might account for any differences?
First things first, however. The Total Index for the 47 item marketbasket rose slightly from September’s 111.07 to October’s 111.11 (November 2010 = 100.00); the Food-Only Index covering the 37 items within the basket dropped by .28 from September’s 114.73 to October’s 114.45 (November 2010 = 100.00). While butter and hot dogs continued to demonstrate the effect of inflation by rising 11.4 and 6% respectively, these were offset by significant declines in white bread, sugar and rice (15.3, 8.1 and 6.4% respectively). In the case of a 20 ounce loaf of store brand white bread, each of the three grocery stores has now dropped the standard price to $1 or less, whereas the prices for the loaf had been on the order of $1.25 or more in past months. As usually occurs, these declines occurred in different months so that the price impact is muted over a longer period of time. What this does show however, is that the grocers are responding to the decreasing buying power of the typical consumer by finding ways to offer a lower priced basic good; I expect that this is because there is use of lower quality flour or some other ingredient than before (stealth deflation).
So how does the CPI figure issued by the US Government differ from what I’m capturing with the PracticalDad Price Index? Let’s examine the six months of results, ending with the most recent comparable month of August 2014. Understand that because the CPI is calculated for multiple geographic areas, there’s roughly a two month window difference between the most recent PracticalDad and CPI results. For comparison purposes, I’m looking at the 37 foodstuff items of the PracticalDad basket and comparing that sub-index result with the Food-at-home line from the CPI monthly report; the full Food Index from the BLS CPI report also includes the change in prices for food purchased in restaurants so I’m reviewing against the sub-category for similarity. In the table below, I’m showing the monthly Food-Only Index from the PracticalDad site as well as the monthly change from the previous month (M-o-M) for both the PracticalDad and CPI Indices.
| Month/Yr | PD Index | M-o-M Change | M-o-M CPI |
| 2/14 | 112.46 | .07 | .5 |
| 3/14 | 112.06 | (.4) | .5 |
| 4/14 | 113.59 | 1.53 | .4 |
| 5/14 | 111.54 | (2.05) | .7 |
| 6/14 | 112.65 | 1.11 | 0 |
| 7/14 | 113.22 | .57 | .4 |
| 8/14 | 113.71 | .49 | .2 |
What’s most noticeable about the Month-over-Month results between the two is that the PracticalDad Index has a much greater variability from one month to the next than the CPI. This makes sense given that the PracticalDad Index is calculated at only three stores in a single geographic vicinity while the CPI Food-at-home is priced nationally from multiple geographically disparate markets across the country. Likewise, the CPI basket is much larger than the 37 foodstuff items in the PracticalDad basket so that there’s a larger number of sampled items; the larger the sample, the less the variability. For a look at the foodstuffs sampled by the Bureau of Labor Statistics, see here, especially down at page 8. Even with this roster from the BLS, it doesn’t list the individual items sampled each month. When I began to sample prices in November 2010, it was a purposeful decision to keep the basket manageable, using the most common items that a family would purchase each month, based upon my own experience as shopper, cook and chief bottle-washer through the years.
The monthly pricing for the PracticalDad Index is done solely by me and even with four years of familiarity, it’s not uncommon to have to return to a store more than once in order to double-check a particular item. This is particularly the case when there’s an incident of stealth inflation and a package size changes. What sometimes makes it difficult is that the change is insidious so that one foodstuff package changes across the stores over a period of months. That said however, the PracticalDad Index has captured instances of stealth inflation in the following products: sugar, orange juice, hot dogs, women’s pads, diapers, store brand tomato sauce, cereal, and coffee. In the case of the BLS, the pricing is done by “economic assistants” hired by the individual metropolitan field offices of the BLS; these individuals are trained in what to look for and then set loose with the results forwarded to the DC office of BLS. The data is then inspected by analysts and any questions are returned to the field if there are issues. The analysts do adjust for any qualitative and quantitative changes, but these changes are never reported in the actual report. So even if the changes are noted in the calculations, they are never individually reported in the monthly results. While I actually believe that BLS process is fundamentally sound, the presence of so many different assistants with so many different products to price can lead to some missed prices.
The fact that the pricing in the PracticalDad Index is limited to a small area means that there can be a greater examination of the inflation versus deflation scenario as well. The index has captured price increases that are attributable to monetary policy, most especially in the pricing of commodity products when hot money flows into, and then out of, that product. Coffee, Canola oil and sugar are prime examples as their prices exploded several years ago with the hot money flows into the commodity markets; these prices have since dropped from previous highs, even accounting for stealth inflation packaging changes in coffee and sugar. I’m also able to see the deflationary side as individual stores are having to adjust for the increasing lack of consumer buying power. One store in particular has had to simply cut prices of some basic foodstuffs since their customer base doesn’t have the funding. The practice has shifted to the other stores in multiple cases as well. Ultimately, it’s this micro-scale that helps explain some of the distrust of the CPI figures by the general public. No one person shops in a wide variety of stores and their viewpoint is limited. But not all stores price in the same way and have the same bargaining power with their suppliers; a regional grocery chain simply can’t bargain with suppliers as well as a national or international chain such as Royal Ahold. Likewise, different stores serve different customer levels and a customer at Trader Joes won’t feel the same pain as one in rural Indiana. When a store raises prices or passes along steath inflationary packaging, you might see it in your store but on a macro-level, it might not be fully processed through the entire system and hence not fully reflected in the CPI.
Here’s my own perspective on the issue of CPI reliability. I have come to believe that the mechanism is fundamentally sound, even if the presence of so many “fingers in the pot” can make pricing problematic at times. But if there’s issue to be taken, it’s to be taken at the upper levels of the government, who split out the more critical sector of food from the CPI component when they determine what increases to put through for benefit recipients.