Looking at government data is kind of like staring into the abyss. If you look long enough, you begin to believe in it. It's important to understand that a measurement is only as good as it's ability to accurately reflect what exists. The main idea behind objectivity, and measurement is precisely this.
How wide is the width of a thumb? Do we use your thumb or mine? Mine is kind of grisly, and has an unsightly wart on one side. But still, that wart counts. That 2mm is important to me. It reminds me how important it is every time I bump it into something, actually. Then I spend about 5 minutes cursing its painful authority and looking for the nail clippers, Compound W, and all sorts of home remedies to rid my beleaguered existence of it's totalitarian rule.
But suppose we used my thumb and your thumb, and her thumb, and his thumb, and my kid's thumb, and your kid's thumb, and their kid's thumb. Individual differences would be averaged out, and we'd be left with a reasonable idea of how wide a thumb really is without using any one particular thumb, warts and all.
The Bureau of Labor Statistics does this with over 200 categories of goods and services, and this makes a lot of sense. When we're talking about quality of life, we're also talking about judgments and tastes. I know, for example, that my wife hates sushi. She despises the little seaweed wrapped rice balls, and can't stand the thought of eating raw fish while I, on the other hand, adore the delicate interplay of flavors and fresh, clean sensation of well-prepared, and artfully presented delectables. She, on the other hand, much prefers deep-fried anything, as long as it comes with a mess of mashed potatoes and corn, or, so long as it's not okra.
While it may be tempting to correlate our individual tastes with where we each of us were born, you might be surprised.
But these individual differences are why there are so many categories. If we wanted a simple measure of food-stuffs, necessary goods and services to insure survival, we'd create one. It would look kind of like a measure of poverty though, because it wouldn't account for individual tastes and how those tastes contribute to any meaningful statement about one's quality of life. Sushi, if you ask me, is only marginally related to a big bucket of fried chicken, and in no way resembles a big plate of pasta, or, even worse, a can of store-brand ravioli. Even the little mini ones with the extra cheese.
But here's where it gets really interesting. In the late 1990's politicians like Newt Gingrich made it their business to induce weighted measures into the CPI (Consumer Price Index). Weighting is an important concept because it allows for more accurate statistical analysis when we're talking about the probability of a certain event. To understand this concept we can look back at my wife's, and my own tastes in food. When making a decision on where to go to eat when we do actually get to go out to eat (kind of rare these days) whose tastes will be considered more?
Looks like it's KFC again.
In the same way that our family system is weighted toward making Mommy happy, consumers act, to a certain degree, rationally. Having a place to live is more important than buying your prescriptions. Buying your prescriptions is more important than having a varied and healthy diet. Having a varied and healthy diet is more important than a full tank of gas. And having a full tank of gas is more important than being able to go to the doctor when you're ill.
Unfortunately weighting economic data in this way is kind of like lying. Each of these services are necessary, and only during price pressures (where the cost of goods are going up, and income is stagnant) is it meaningful. Building weighted measures into the CPI -U, which is supposed to be a measure of the cost maintaining one's standard of living (whether it's going up or going down, or going nowhere which is actually preferable in this day and age) is akin to firing a starting gun in a race to the bottom.
In order to explain this, we have to understand a few core concepts regarding budgets. A budget is a reasonable guess at what one plans to spend. It's kind of like a plan. I plan to eat sushi next month. In order to do that, I have to put aside the cost of that sushi, the gas it took to get there, and the tip I expect to leave for their always impeccable service. They bring hot towels out for their customers before the meal, present the food artfully, and excel at being unobtrusive but available. Did I say I love sushi?
All of those costs have to be accounted for in order to make it happen. If I don't have the money for the bill, or the cost of the gas, or enough left over to reward the service commensurate with their skill and attention, I can't afford the trip. So, in making the plan, all of those costs have to be accounted for. But what if I got there and their prices had gone up? Or the price of gas had gone up?
I'd have to make different choices. Lunch at a sushi bar would turn into lunch at a sandwich bar. This rational choice, because of actual restraints at the time of purchase rather than a reasonable guess bused on historical data drives the actual goods and services purchased and the quality of that transaction. It is in this way that the CPI-U encourages a race to the bottom. The price of goods and services is not represented, as only the cost of the eventual purchase made is recorded in aggregation. The CPI-U runs behind the times, and it keeps on running as consumers make more and more adjustments to their standard of living which was what the CPI-U is meant to measure.
The “fixed-basket” model of the standard of living is in no way represented in current economic data, and it hasn't been for a long time. The same goods and services you bought and paid for ten years ago have changed dramatically as people have moved toward cheaper and cheaper options to fulfill the same need serviced by that “fixed-basket.” Economic substitution, as it's known, has allowed BLS data to be corrupted by price pressures, considerably reducing the apparent cost of living while substantially hiding the inflation of that fixed-basket of goods and services.
Yet analysts say the nature of sushi and fried chicken are inherently different, and the difference in price reflects that difference in quality. This seems rational. It is rational to say that sushi is worth more to me than fried chicken. The same can't be said for my wife. I'm willing to pay more for sushi than fried chicken. However, if I was able to afford that choice 10 times last year but only 6 times this year, the value of my income, whether the amount went up or down, actually went down. In effect, I would have to rationalize a rise in value to continue to choose sushi over fried chicken, or even ravioli. In the same way that a $10 an hour wage 20 years ago is not the same as a $10 an hour wage today.
The economic substitution model works to hide the actual inflation of goods and services. We can take, for example, the cost of a cell-phone and show this trend. While having a phone is necessary for employment, the quality of said phone is variable. It makes perfect sense that people will only buy the goods and services they find valuable. Having a phone is not having the newest phone, or the smartest phone, although I'm not sure why we ever picked that term. I, for one, would not want a phone that made rude, juvenile, and obvious comments about my wittiness or lack thereof throughout the day. Nor would I prefer to be wired in to a network of people that want to know if I'm thinking about what I'm going to get our friend Bob for his birthday next week while I'm trying to work. In other words, some services aren't valuable to some people even if they are available.
The problem is that a smart phone is packaged with a phone, facebook, twitter, text messaging, data usage, and picture messages. Services are packaged together and sold in blocks; even if you don't use video conferencing, you still pay for video conferencing, or at least a small portion toward it. It's kind of like buying a car. You have to have a road to drive it on. In this way, the economic substitution model doesn't really reflect a substitution at all, but merely a reduction.
What this means: if you were already paying for a doctor visit inside a market that normalized the price of that service it didn't pay you to switch doctors to one that cost less at the office but more at the pharmacy because this one doctor doesn't give out free samples. The quality of the visit is eroded while the cost of the visit remains the same. It's a system, and each part of the system represents a cost and a benefit that is part of the whole price paid.
The Chained CPI, or C-CPI-U, is even more esoteric, convoluted, and unrepresentative of real life as people live it.
Hopefully I've added something to your understanding of the CPI-U and how it's built. The relative costs of this “measurement” are broad considering that it's used in calculating a great deal of government data regarding our economy, not just including retirement, benefits, and compensation. Government data drives labor prices in more than merely the minimum wage. Fallacies in the CPI-U can drive prices of goods up as it fragments the quality of markets while driving overall compensation down.
Just like understanding the differences between buying a condo and renting a walk-up, it's important to understand the way we formalize concepts about the economy. Small differences can mean big changes. Just like the wart on my thumb; it wouldn't hurt if I could afford to have it removed.