CPI and ECONOMICS CLASS > If you or others from sci econ would be willing to >check out my page it would be appreciated. It is elementary >compared to your fine page, but have put quite a bit of work >into it and it could use more sanity checks for possible >errors either in facts or in concepts. >B. Burkart >http://www.netcom.com/~bhb1 Hi, Looking over your web page I found this about CPI and Boskin: QUOTE: "COLA Reduction Efforts History The desire for reduced COLAs for Social Security Federal programs etc. was mentioned in several of the "doom" books in 1993 or before. Those books were recommending cuts simply to try to reduce those programs and long before the so called "accuracy" issue came up." END OF QUOTE: When I studied economics in college back in the mid 1950's it was pointed out that the CPI and any other price index (there were several of them) will necesasarily over state inflation. They are based on a fixed "basket" of goods: when the prices change, the basket is not changed to reflect the purchase changes in response to the new prices. The only real issue is "by how much is inflation overestimated?". I thought all economists knew this. Since price and the corresponding purchase changes take place over such a long time frame that most people don't notice, you can see the process better if you travel to another country where the relative prices are different. See my web page files on PPP and PPP & CPI in the MONEY & INFLATION part of the econ section. I will close with a true story about how, by paying attention in my college econ class, I saved thousands of dollars for my mother-in-law. LISTEN UP ALL YOU STUDENTS My mother-in-law had a big business/rental housing building in downtown Tel Aviv, Israel, that her husband left when he died in 1945. She could barely make a living from it because of rent control laws, and was getting too old to keep trying. So in 1964 she was offered a deal by the Keren Kayemet (a sort of land corporation/government agency): they would take the property and give her a two bedroom house in Haifa plus a monthly pension. She said yes, and I happened to be there, and was invited to be present when they finalized the deal. They spoke in English for my benefit, and I raised the issue of correcting the pension (to be paid in shekls, the local currency) for inflation. They said sure, and it could either be linked to the US dollar or to the Israel Consumer Price Index. Linking payments to the dollar is common for contracts in Israel, but I told her to go for the CPI, since I remembered from econ class that this would overcorrect for inflation. The original amount was the equivalent of $200 some dollars per month. But the dollar equivalent rose to over $500 by the time she died this year. And the shekl in the interval was replaced by the New Israeli Shekl (NIS), worth 1000 old ones. Had she had NO inflation correction, her pension would have been almost nothing in 30 years. A dollar link would have had it fall with the US inflation (especially during the 1970's). But the CPI correction had it rise in dollars and also in purchasing power. Another lesson in this story is the "protection" that rent control has. The Tel Aviv building is located in a busy part of the city but was worth little to her, due to rent control: she could not raise rents or evict the tennents. The day Keren Kayemet took over, the rents went up and some of the tennants were out in the street. It helps to know the right people. I hope that this will not reduce my chances of making the coveted "Burkart Hall of Fame" :-( ,,,,,,, _______________ooo___(_O O_)___ooo_______________ (_) jim blair (jeblair@facstaff.wisc.edu) Madison Wisconsin USA. This message was brought to you using biodegradable binary bits, and 100% recycled bandwidth. Subject: Re: CPI, SS , Boskin & Hall of Fame Date: 4 Feb 1998 23:12:05 GMT From: rvien@see.sig.com (Robert Vienneau) Organization:Dreamscape Online Newsgroups: alt.politics.economics, sci.econ Jim Blair writes: When I studied economics in college back in the mid 1950's it was pointed out that the CPI and any other price index (there were several of them) will necesasarily over state inflation. They are based on a fixed "basket" of goods: when the prices change, the basket is not changed to reflect the purchase changes in response to the new prices. The only real issue is "by how much is inflation overestimated?". I thought all economists knew this. >Robert Vienneau: >It's a fallacy to appeal to anonymous authority. Hi, Sorry if I was not clear: what I intended to say is that this is the sort of thing that is commonly taught in a basic college economics course. As for example Economics 101 at Cal Berkeley. See below. >As usual, Mr. Blair doesn't know what he is talking about. A Laspeyres >price index generally overstates inflation, but a Paasche price index >generally understates inflation. Both indices are based on a fixed basket >of goods. (What happens if all commodities are perfect complements?) Anyone interested in the terms here can look at Econ 100b Lecture 3 by Brad De Long, Associate Professor of Economics, University of California Berkeley, CA found at: http://econ161.berkeley.edu/Teaching_Folder/Econ_100b_S96/Lectures/lecturethree.html QUOTE: (on index types) A price index with a fixed basket of goods: it is sometimes called a "Laspeyres" index. A price index dual to a fixed-prices quantities index: it is sometimes called a "Paasche" index. An index with changing baskets of weights chained together is--no surprise--a "Chain" index. I don't want you to worry about the names "Paasche" and "Laspeyres". I want you to, instead, worry about the problems of making index numbers. Four key lessons: A fixed-weight index takes no account of people's ability to substitute to achieve the same utility when commodities become more expensive... A deflator-like index does not take account of the utility lost as changing prices push you away from your original basket of consumption choices. For years closer to the present than the base year, real GDP tends to overstate rates of economic growth. For years further in the past than the base year, real GDP tends to understate rates of economic growth. .....and later... Biases in measurement? Yes: The CPI overstates rates of price increase by 1/2 to 1.5 percent per year (and the GDP deflator has similar biases). We are getting richer and more productive faster than the official numbers report. Additional bias: For years closer to the present than the base year, real GDP tends to overstate rates of economic growth (and the implicit price deflator tends to understate inflation). Additional bias: For years further in the past than the base year, real GDP tends to understate rates of economic growth (and the implicit price deflator tends to overstate inflation). UNQUOTE My comments: I agree with Professor De Long that knowing names like "Paasche" and "Laspeyres" is less important than understanding the problems of constructing an inflation correction, and knowing the sources of the errors in them. And what you call a Paasche index is what have called a price deflator in my posts and on my web page. Yes, they underestimate inflation. Readers may also want to look at: This from Why Are People So Indifferent to Indexation? by Robert J. Shiller at: http://www.econ.yale.edu/~shiller/indxpap6.html to find this quote: It is well known among economists that LaSpeyres price indices, like the US Consumer Price Index, are biased towards overestimating the increase in the cost of obtaining a fixed level of utility, they overestimate the increase in the cost of living, so long as relative prices change during the process of inflation. The reason for this bias is just that when relative prices change, people can substitute for the items that become relatively less expensive. A contract that promises that a person will receive a fixed amount in real dollars as measured by the consumer price index allows that person to consume the same market basket as before, and so the person can be no worse off, he or she will generally be better off since relative price changes create opportunities for substitution. END OF QUOTE >In this context, the "true" rate of inflation is the rate of change in >money incomes that would leave consumers as well off in their own >judgement (that is on the same indifference curve), given unchanging >tastes. At least the CPI is based on the actual selling price of actual goods. Weighted by how actual people actually buy them. "In their own judgement" sounds subjective and hard to measure to me. And who has "unchanging tastes"? Not me. Do you buy the same things that you did 5 years ago? >The CPI is not a pure Laspeyres index. Care to explain what this means? I suppose that nothing is PURE anything, but the BLS CPI was intended to be a Laspeyres index. >So one could ask whether corrections >to the CPI, some of which have been made by the BLS, lead to a CPI >that understates some components of inflation. Back to Boskin. This is where I came in. Since price and the corresponding purchase changes take place over such a long time frame that most people don't notice, you can see the process better if you travel to another country where the relative prices are different.---jeb >Since tastes vary over time and space, a precise definition of inflation >does not exist here. >Robert Vienneau By "here" do you mean in space and time? But where else IS there? Again I recommend my web page file PPP & CPI under MONEY & INFLATION. AND: jim blair: It has been economic dogma for decades, and taught in basic college Econ classes since the 1950's (at least) that comsumer price indices over-state the effect of inflation, probably by about one percent. The 1% figure is widely accepted, but of course controversial. "Dennis Wallick" : >The committee did no original research. mwitte@merle.acns.nwu.edu (Mark Patrick Witte): In fairness to the members of the committee, they were chose largely on the basis of their considerable contributions to the research on the measurement of quality, productivity, and other issues that directly affect the computation of price indexes. >Very few agree with Boskin's guess of 1%. > So, if your "claim" of the overestimate is 1%, please let me know how >you derived that figure..... It was this: Bias in CPI Inflation Estimates of Bias in the U.S. Consumer Price Index Author(s) Point estimate Advisory Comm. to Study the CPI (1996) 1.1 Michael Boskin (1995) 1.5 Congressional Budget Office (1994) .5 Michael R. Darby (1995) 1.5 W. Erwin Diewert (1995) 1.5 Robert J. Gordon (1995) 1.7 Alan Greenspan (1995) 1.0 Zvi Griliches (1995) 1.0 Dale W. Jorgenson (1995) 1.0 Jim Klumpner (1996) .4 Lebow, Roberts and Stockton (1994) 1.0 Ariel Pakes (1995) .8 Shapiro and Wilcox (1996) 1.0 Wynne and Sigalla (1994) 1.0 Equally weighted average 1.1 SOURCE: Moulton (1996). from: http://www.dallasfed.org/whatsnew/ar_96.html I think economists are more likely to agree with it than politicians or those with a vested interest in a large CPI based COLA "correction". The idea that the CPI overstates inflation was not seriously challenged until there was a movement to actually reduce the COLA that powerful labor unions and special interest groups receive by contract or law . Groups like AARP and those receiving Social Security. Subject: Re: CPI, SS , Boskin & Hall of Fame Date: Sun, 08 Feb 1998 19:29:56 GMT From: oldnasty@mindspring.com (Grinch) Newsgroups: alt.politics.economics, sci.econ References: 1 , 2 , 3 georgeb@p085.aone.net.au (george blahusiak) wrote: >Actually I looked into this and found that both scenarios are correct, >and both are incorrect, depending on whether you think your glass is >half full or half empty. > >Over here we have a CPI which is based on household spending data >that is 10 years old. Your right, substitution makes a lot of sense. >But, you also have to take into account that it does not include all >new bank charges over the same period, nor does it take into account >new mobile phone charges, nor does it take into account new taxes and >other imposts from the govt, not does it take into account, say, new >internet charges, not does it take into account.... > >In addition it does not take into acount changing conditions, part of >which I have described above. For example, at one time a telephone was >a pure luxury, today it is a necessity. At one time a motor car was a >luxury, today it is nto only a necessity, but with greater demands on >out time it would appear that two cars are the norm for families with >choildren. When a prior generation's luxuries becomes everyday possessions, this is a sign of improving living standards and a lower cost of living rather than the reverse. > I have noted the internet as being a new product/service. >Will the internet become an essential, and will it replace some other >product/service? If it does, it won't raise the cost of living and will improve welfare since consumers will have demonstrated a preference for it over the alternatives it replaced. The introduction of new products and services improves welfare while having no effect on CPI. This is the largest single cause of bias in the CPI causing it to understate improvements in consumer welfare. A simple example illustrates: Suppose you shop in a store stocked with 100 goods that are all included in the CPI index. Your "cost of shopping" necessarily will reflect the CPI. Now imagine the store adds to its line of goods many more items that aren't included in the CPI index. Your welfare as a shopper goes up because you have all the choices you had before *plus* many more. If you choose to purchase any of the new items rather than any of the original 100 that you would have purchased without the extra choice, you will be *measurably* better off. Yet your improvement in welfare will not be reflected in the CPI because as far as it is concerned the new products don't exist. In fact, if you choose to buy any of the new products you'll be better off even if they cost *more* than the old products you'd have bought absent the extra choice. This is true or you wouldn’t buy the new products, you'd buy the old products instead (or save your money). Consumers don't have to buy internet connections or other new products or services. They have free choice, and even if a consumer who buys a newly-innovated item views it as an extra expense, he must think he's better off paying for it or he'd choose not to and continue living as he did before. The "substitution effect" gets a lot of publicity in the CPI debates and is almost always explained in terms of existing goods: "when the price of apples goes up people buy more oranges". But the largest single factor that causes the CPI to overstate the real cost of living and understate consumer welfare is the substitution of *new products for old ones* First, the CPI is completely blind to the welfare benefits that result from newly-innovated products and services that are not in its index (and none are in its index when they are new). Second, the standard pricing behavior for newly-innovated products is for them to be introduced at a high price and then decline in price -- and the CPI completely misses these price drops until the new products are included in the index, if they ever are. An excellent example of this is cellular telephones. Cell phone services are not included in the CPI at all, even though they now account for almost 20% of all phone service. Obviously, cell phones must provide a substantial benefit to consumers or they wouldn't be such a fast- growing part of such a major market. Also, the price of cell phones has plummeted since their introduction. They cost $2,000 each when introduced -- now gas stations give them away free as premiums. (I short while back when I was in the local supermarket the check-out clerks were comparing their cell phones and talking about where to get the best rates for service). Both the welfare impact and price-drop impact of cell phones has been omitted from the CPI, but they have been studied by the index makers. The conclusion is that if cell phones had been included in the CPI for the last ten years the "telecoms" component of the CPI would be about 20% lower than it is now -- instead of showing a 10% increase over the last 10 years it would show as much as a 10% decrease. [See _Cellular Telephones, New Products and the CPI_, by Jerry Hausman, MIT, National Bureau of Economic Research Working Paper #5982] So those "new mobile phone charges" would *lower* the CPI if they were included in it. And so would Internet charges and all those other new products and services that we find ourselves paying for voluntarily. >By the same token it (the CPI) doesn't take into account reductions in >various charges (hah!) or the loss of products and services from the >economy i.e., basket of goods. When products and services disappear because people don’t want to buy them anymore, what's the cost to the consumers? Did the disappearance of the Edsel raise the cost of living? Did the "loss" of the TRS-80 computer? (I had one and loved it in its day, but ....) >Before passing on to more important topics, if ask anyone who really >knows and they will tell you that the CPI is not a measure of infla- >tion, it is not and never was intended to be a measure of inflation. Exactly true. Neither is the CPI a measure of consumer welfare. But people use it as a measure of inflation and welfare all the time. The CPI is a price index, no more. >I myself made that same mistake, thinking the CPI was a measure of >inflation, but when I telephoned the Bureau of Stats they set me right >with a very few words indeed. > >Because of the way it is calculated and the way it is used it will >always lag one way or the other. The only way to measure inflation is >to measure the changing price of all goods and services in the economy >in ablolute terms. Unfortunately some people (politicians and other >people who would manipulate public opinion amongst them) find that >distasteful. It can't be manipulated. > >Cheers, > >George Regards ========= "In all countries the usual price index numbers exaggerate the increase in the cost of living and understate the laborer's real reward per unit of work". -- John Maynard Keynes, 1921