Top
Reference

CPI Quality Adjustments

This page is about the “quality adjustments” in the Consumer Price Index (CPI), and how they affect its application.

Summary

19 Feb 2010

Quality adjustments remove from the CPI the cost of any improvements in quality, even if they become standard. This is, in fact, required in any intellectually consistent approach using the BLS' definition of inflation. However it is, to some extent, at odds with measuring the cost of living as normally experienced.

With the CPI, the BLS attempts to answer the question, “What is the cost, at this month’s market prices, of achieving the standard of living actually attained in the base period?” To do this, improvements in quality are stripped out of price gains. When TVs started to have remote controls, for example, or doctors started routinely testing cholesterol levels, this philosophy required that any price increases due to the “quality improvement” not be counted as part of inflation. The cost of a higher quality level is never counted, even if it becomes standard.

The audio/video portion of the CPI-U increased 0.6%/yr (compounded) from 1993 to 2008, and IT hardware and services decreased 11.3%/yr from 1989 to 2008 (in both cases, this goes back to the first available data). Autos also steadily decreased in cost from 1997 to 2008.

The stated goal of the CPI is intellectually a very natural one. However there is some mismatch between this definition of CPI and one of its main applications, namely cost-of-living adjustments. People not only want today's standard of living rather than that of the base period, but in many cases there is no choice: one cannot obtain today a car without seatbelts, a computer that can only run MS-DOS, or an annual physical with no mention of cholesterol.

Highlights

(Clippings below covered through 2009)

  • Overall goals (19 Feb 2010) The way the BLS defines its goal in calculating the CPI is logically consistent, very pragmatic in terms of calculation, and often appropriate for the main applications of the index. However there is some mismatch with one of its main applications, namely cost-of-living adjustments (COLA).
    • BLS (18 Apr 2008) The BLS Handbook of Methodology has this to say about their overall goal: “A unifying framework for dealing with practical questions that arise in construction of the CPI is provided by the concept of the cost-of-living index (COLI). As it pertains to the CPI, the COLI for the current month is based on the answer to the following question: “What is the cost, at this month’s market prices, of achieving the standard of living actually attained in the base period?” … [for example] Adjustments for quality change in the CPI new car index include structural and engineering changes that affect safety, reliability, performance, durability, economy, carrying capacity, maneuverability, comfort, and convenience. Since 1999, quality adjustments have not been made for changes associated with pollution control mandates.”
    • Those depending on COLA (19 Feb 2010) would naturally prefer a standard of living normal for the times in which they live. This may seem pedantic but in fact it is substantial. The “base period” approach, taken to its natural conclusion, would suggest that someone who has depended upon a CPI-based-COLA for 30 years would have to be satisfied with medical care that did not include statins or tumor-specific oncology, 20-pound 'portable' computers, mobile phones the size of a brick, music without compact discs, no widely available internet, and cars with bench seats and very poor mileage. That is extreme but, by its philosophical basis, a 'base-period' COLA is in fact designed to slowly but relentlessly make technological advances unaffordable, even if they become standard.
  • Gradual introduction of quality adjustments (6 Jan 2010) The methodology of adjusting for quality improvements has gradually been extended to more categories over time. See the 1999 clipping.
  • Related material from PPI (18 Apr 2008) One presumes that the quality adjustment process for the CPI and PPI are closely related. The background material on the PPI has a fairly pithy summary. The three main methods for assigning a price to a quality improvement (which is then taken out of the “real” price increase) are:
    • Producer cost and margin (18 Apr 2008) The most usual method depends on the manufacturer reporting the “the extra product cost and normal margin associated with the addition” of a new feature.
    • Overlap method (18 Apr 2008) “the Bureau may have to assume that any difference in price between the old and the new items is due entirely to differences in quality”. A month is chosen in which both the old item and the new one are both being produced. The price difference is taken as entirely due to quality. Before the chosen month, the old item is in the index, and after that month, the new item.
    • Hedonic regression (18 Apr 2008) “Hedonic regressions estimate the functional relationship between the characteristics embodied in the products in a market and the products’ prices. Such regressions yield estimates of “implicit prices” for specified product characteristics that may be used to value the quality improvement resulting from changes in the various characteristics embodied in a product.”
  • Three specific cases (6 Jan 2010) The audio/video portion of the CPI-U increased 0.6%/yr (compounded) from 1993 to 2008, and IT hardware and services decreased 11.3%/yr from 1989 to 2008 (in both cases, this goes back to the first available data). Autos also steadily decreased in cost from 1997 to 2008.
  • Wide algorithmic variations possible (22 Apr 2008) A survey of quality adjustments performed for household appliances across German city CPIs showed price changes ranging from +31% to –13% between 1980 and 1997, probably from inconsistent quality adjustments.
  • An extreme example (19 Feb 2010) Here is a comparison between the CDC 6600, which was the world's fastest supercomputer in 1963, and my current laptop:
CDC 6600 MacBook Pro 17”
Processor speed 3 MHz 2.8 GHz
Instruction cache 240 B 64 KB
CPU parallel 10 CPU functions dual core
Transistors 400 k 582 m
Memory 1 MB 4 GB
Memory parallel 32 banks -
Size 750 sq ft room 1” x 11” x 16”
Weight 12,000 lbs 7 lbs
Power 150 kW 35 w
Price $10 m $2500
Disk/drum 2 MB 500 GB

My laptop has 1000 times more transistors and runs about 1000 times faster, so is perhaps 1,000,000 times more powerful than the CDC 6600. A millionth of the cost of the laptop is less than a cent. So discounting quality improvements, a 1963 supercomputer costs less than a penny today. Unfortunately for someone who has had their income adjusted by the CPI since 1963, there are no one cent computers for sale today.

See also

Clippings

Below are the clippings, from outside sources, that were used in constructing this page.

Seat belts are a quality adjustment

4 Dec 1996. Social Security Website, full report of Boskin Commission.

http://www.socialsecurity.gov/history/reports/boskinrpt.html

“The Boskin Commission Report”

”[Overall effect:]

The Advisory Commission To Study The Consumer Price Index (aka The Boskin Commission) was appointed by the Senate Finance Committee to study the role of the CPI in government benefit programs and to make recommendations for any needed changes in the CPI. The Commission's December 1996 report recommended downward adjustments in the CPI of 1.1%. The CPI is the basis for Social Security COLAs and this recommendation, if adopted, would reduce future Social Security COLA increases, as well as impact numerous other government programs. …

[Seat belts are a quality adjustment:]

We are persuaded that mandated anti-pollution devices are analogous to an indirect tax. Gasoline taxes may be used to provide a benefit in the form of better highways, but a tax increase is treated correctly by the CPI as an increase in the cost of living. Anti-pollution devices provide a benefit in the form of cleaner air, but by analogy to taxes should be treated as an increase in the price of the car. Using the detailed information given by Gordon, we can calculate an alternative to his index that converts the CPI adjustment for anti-pollution devices from a quality change to a price change, and this results in the finding that the CPI for new cars was downward biased during 1947-83 by 0.94 percent per year. We do not make a similar adjustment for the value of quality change taking the form of safety devices such as seat belts and crash-resistant bumpers, since our feeling is that consumers see the connection between their own safety and the devices more directly than they do between anti-pollution devices and air quality. …

[Quality adjustment for used cars recommended:]

The CPI index for used cars has long been known to be upward biased, simply because no quality adjustments were applied to this category at all. The upward bias over the period 1967-87 is very large, amounting to 2.44 percent per year, if we take the difference in the growth rate between the new car and used car index to represent a measure of the bias.”

Gradual introduction of quality adjustments in more areas

Jun 1999. Monthly Labor Review.

http://www.bls.gov/OPUB/MLR/1999/06/art4full.pdf

“Consumer Price Index research series using current methods, 1978-98. Kenneth J. Stewart and Stephen B. Reed”

“BLS research indicates that the measured rate of inflation would have been lower since 1978 if methods currently used in calculating the Consumer Price Index for All Urban Consumers had been in place from that year to the present …

In 1967, the Bureau began to adjust new-car prices for changes in the quality of the cars. In 1987, the Bureau began adjusting the used-car index for similar changes by applying, to each model in the used-car sample, the percentage of quality adjustment employed when the model was new. …

In 1988, quality adjustments reflecting the aging of the housing stock sample began. The CPI-U-RS incorporates an estimate of the effect of this change by adjusting the resi- dential rent and owners’ equivalent rent indexes upward by about 0.3 percent per year from 1978 to 1987. This figure represents the average of the adjustment factors used in the CPI from 1988 to 1999. …

In 1991, the Bureau initiated the use of hedonic models to estimate changes in quality for apparel commodities. Using a BLS study that estimated the effect of this improvement over the last 6 months of 1991, the Bureau adjusted all of the CPI-U-RS apparel commodity indexes from 1978 to 1990 upward by ap- proximately 0.4 percent per year. …

In 1998, hedonic regression models were first used to adjust personal- computer prices for changes in quality. Estimates based on an analysis of 1998 data indicate that this change has had the effect of lowering the personal-computer index by about 6.5 percent per year. …

Hedonic tech- niques were used to adjust the television component of the CPI for changes in quality for the first time in 1999.”

[Note that the FT was wrong in saying that downward adjustments in quality are never taken into account.]

Background on methodology for quality adjustments in the PPI

3 Sep 2003. BLS web site.

http://www.bls.gov/opub/hom/homch14_b.htm

“BLS Handbook of Methods. Chapter 14. Producer prices.”

“When changes in physical characteristics of a product cause product cost differences, however, the Bureau attempts to make an accurate assessment of real price change by taking account of quality differences systematically. The explicit quality adjustment method is especially important for automobiles, machinery, and other types of goods that undergo periodic model changes. For these goods, the usual method of quality adjustment involves the collection of data from reporting companies on the costs they have incurred in connection with the quality change. For example, if the price of a new model car is $500 more than the previous model year’s version and $200 of that increase is due to the extra product cost and normal margin associated with the addition of government-mandated safety equipment, then the real price has risen by only $300. The change in the passenger car index will reflect only that amount, not the nominal price rise of $500. Unfortunately, it is not always possible to obtain a value for quality adjustment. If, for instance, the respondent is unable to estimate the production cost difference between an old item and a new one, or if an explicit comparison between an entirely new product and a previous product is not feasible, then no quality adjustment value will be forthcoming. In such cases, the Bureau may have to assume that any difference in price between the old and the new items is due entirely to differences in quality; the Bureau, therefore, employs the “overlap” method (if possible). Under this method, the Bureau collects prices for both the old and the new item over a designated period and chooses a particular month as the overlap month. The difference between the prices of the two items in the overlap month is assumed to represent the value of the difference in quality between the items. For purposes of calculating the official price index, the Bureau uses price changes for the old item through the overlap month, but thereafter follows price changes only for the new item. … “high-tech” industries … frequently develop new products that are technologically superior and cost less. The conventional quality adjustment methodology is suitable for situations in which increased resource costs for producing a product are necessary for improved performance. This is the exact opposite of what typically happens in industries that manufacture sophisticated products composed of electronic components. Such an inverse relationship between cost changes and quality changes requires many different techniques for the construction of an index, especially in the area of quality adjustment. An alternative quality adjustment technique using hedonic regressions has been incorporated into PPI adjustment processes. Hedonic regressions estimate the functional relationship between the characteristics embodied in the products in a market and the products’ prices. Such regressions yield estimates of “implicit prices” for specified product characteristics that may be used to value the quality improvement resulting from changes in the various characteristics embodied in a product. The value of the quality improvement can then be removed from the reported price change to obtain a measure of the pure price change that is appropriate for the PPI.”

Cost of base period living; new car example

Jun 2007. BLS Handbook of Methodology Chapt 17.

http://www.bls.gov/opub/hom/pdf/homch17.pdf

“Chapter 17. The Consumer Price Index (Updated 06/2007)”

“A unifying framework for dealing with practical questions that arise in construction of the CPI is provided by the concept of the cost-of-living index (COLI). As it pertains to the CPI, the COLI for the current month is based on the answer to the following question: “What is the cost, at this month’s market prices, of achieving the standard of living actually attained in the base period?” … Seasonal adjustment is done only at the national level for the U.S. city average CPI-U and CPI-W. Presently, the C-CPI-U does not have sufficient historical data to permit calculation of stable seasonal factors. … One of the more difficult problems faced in compiling a price index is the accurate measurement and treatment of quality change due to changing product specifications and consumption patterns. The concept of the CPI requires a measurement through time of the cost of purchasing an unchanging, constant-quality set of goods and services. … Direct quality adjustment. The most explicit method for dealing with a replacement item is to estimate the value of the differences. The estimate of this value is called a quality adjustment amount QAj,t-1. In this case: Pv+1,j,t-1 = Pvj,t-1 + QAj,t-1. Chief sources of direct quality adjustment information are manufacturers’ cost data and hedonic regression. …Adjustments for quality change in the CPI new car index include structural and engineering changes that affect safety, reliability, performance, durability, economy, carrying capacity, maneuverability, comfort, and convenience. Since 1999, quality adjustments have not been made for changes associated with pollution control mandates.”

Background on price index quality adjustment

8 Oct 2007. OECD handbook

” HANDBOOK ON HEDONIC INDEXES AND QUALITY ADJUSTMENTS IN PRICE INDEXES: SPECIAL APPLICATION TO INFORMATION TECHNOLOGY PRODUCTS by Jack Triplett”

http://puck.sourceoecd.org/vl=2363135/cl=21/nw=1/rpsv/cgi-bin/wppdf?file=5lgsjhvj74xs.pdf

“Different quality adjustment methodologies are employed for ICT [information and communication technology] products across OECD countries, and they seemingly make large differences in the trends of price movements for these products. Wyckoff (1995) reported that changes in computer equipment deflators among OECD countries ranged from plus 80% to minus 72% for the decade of the 1980s; the largest decline occurred in the US hedonic price indexes for computer equipment.

This handbook reviews the methods employed in price indexes to adjust for quality change. A natural division is between “conventional” methods typically employed by the statistical agencies of many OECD countries, which are discussed in Chapter II, and hedonic methods for adjusting for quality change (alternatively known as hedonic price indexes). The latter have a prominent place in price indexes for ICT products in several OECD countries. Hedonic methods for producing quality-adjusted price indexes are reviewed in Chapter III.

The handbook brings together material that is now scattered in a wide number of places, but goes beyond the economic literature in significant respects, particularly in the comparison of conventional and hedonic methods in Chapter IV. This handbook compares and contrasts the logic of hedonic methods and conventional methods and the results of employing them in different circumstances. Although most of the examples in the handbook are drawn from ICT products, the principles in it are very general and apply as well to price indexes for non-ICT products that experience rapid quality change, and also to price indexes for services, which are affected by quality changes fully as much as price indexes for goods, though sometimes that has not sufficiently been recognised. …

And finally, Hoffmann (1998) reports on the quality adjustments performed for household appliances across German city CPIs. In the German statistical system, the country-wide CPI is an aggregation of city indexes constructed independently by each Land. Hoffmann showed that appliance indexes were not very correlated across German cities. They showed price changes ranging from +31% to –13% between 1980 and 1997 (see Figure 2.4). He concluded that the wide differences in these indexes were caused by inconsistent quality adjustments, not by real differences in price changes for appliances in German cities.”

[This handbook is 250 pp long. It seems objective and sensible. Clearly the quality adjustment problem is a real one, and the complaints that governments are trying to understate inflation, while possibly true, are too simple.]

Hedonic quality adjustments

27 Nov 2007. FTUSA p26.

“The statistical authorities have also been “adjusting” the inflation numbers downwards to take account of improvements in quality. Most people would accept the logic of making quality adjustments for computers. But once they started going down this route, the government's statisticians have found it hard to stop. The US government is now making “hedonic quality adjustments” to a whole range of items: clothes dryers, microwave ovens, camcorders, DVD players, even college textbooks. The problem is that no one is making any negative adjustments to take account of declines in quality. Because the data are being adjusted only for improvements, never for declines, it is arguably biasing the published inflation numbers downwards. … Excess liquidity created by the Federal Reserve and other central banks during 2001-2005 did indeed result in faster inflation - but in asset prices rather than goods and services. It also created distortions - but in property prices, subprime mortgages and securitisations.”

Audio/Video

22 Apr 2008. BLS web site.

(From the “Get detailed statistics” go to the row for CPI-u and choose the cell under the column “Create customized tables, one screen”)

12 Months Percent Change
Series Id:    CUUR0000SERA,CUUS0000SERA
Not Seasonally Adjusted
Area:         U.S. city average
Item:         Video and audio
Base Period:  DECEMBER 1997=100

Year	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec	Annual	HALF1	HALF2
1994	2.4	1.2	-0.4	-0.9	0.2	-0.7	-1.8	-2.5	-2.7	-3.1	-2.9	-2.5	-1.1	0.3	-2.6
1995	-1.8	-1.4	-1.4	-0.6	-0.2	-0.7	-0.1	0.2	0.4	0.5	0.5	0.4	-0.3	-1.0	0.3
1996	-0.1	0.3	0.5	0.5	0.1	1.5	2.5	2.5	2.5	3.1	3.1	3.3	1.6	0.4	2.9
1997	3.7	3.6	3.3	2.2	2.9	3.2	2.7	2.8	2.9	2.5	2.8	2.4	2.9	3.2	2.7
1998	2.5	2.2	2.3	3.0	2.3	1.6	1.2	1.2	1.5	1.2	0.5	0.7	1.7	 	 
1999	0.8	0.4	-0.2	-0.4	-0.3	-0.5	-0.5	-0.3	-1.3	-1.0	-0.7	-0.6	-0.4	 	 
2000	-0.9	-0.8	-0.3	-0.7	0.4	0.8	0.7	0.7	1.4	0.9	0.8	0.6	0.3	 	 
2001	0.7	0.8	0.7	1.4	0.3	-0.2	0.4	0.1	-0.2	0.3	0.5	0.5	0.5	 	 
2002	0.9	1.3	1.3	1.2	1.5	1.7	0.9	0.7	1.0	1.3	1.6	2.0	1.3	 	 
2003	1.3	0.9	0.8	0.9	0.7	0.7	1.1	1.3	1.2	0.9	0.8	0.1	0.8	 	 
2004	0.2	0.3	0.6	0.9	0.8	0.7	0.7	0.4	0.5	0.7	0.2	0.6	0.6	 	 
2005	0.6	0.2	0.3	0.1	0.0	-1.2	-1.2	0.2	0.4	0.2	0.2	0.0	0.0	 	 
2006	-0.1	0.0	0.6	1.0	0.9	2.0	1.8	0.4	0.1	-0.3	-0.5	-1.1	0.4	 	 
2007	-1.3	-1.1	-2.2	-2.5	-1.8	-1.7	-2.1	-2.3	-1.7	-0.9	-0.9	-0.1	-1.6	 	 
2008	0.2	0.0	0.6

[Levels Jan 1993 94.1; Jan 2008 103.0; ratio 1.095; 
up 9.5% in 15 years, or 0.6% annual compounded.]

IT hardware and services

22 Apr 2008. BLS website.

(From the “Get detailed statistics” go to the row for CPI-u and choose the cell under the column “Create customized tables, one screen”)

Series Id:    CUUR0000SEEE
Not Seasonally Adjusted
Area:         U.S. city average
Item:         Information technology, hardware and services
Base Period:  DECEMBER 1988=100

Year	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec	Annual	HALF1	HALF2
1989	 	 	 	 	 	 	 	 	 	 	 	-7.3	 	 	 
1990	-6.5	-6.5	-5.4	-4.1	-3.2	-2.1	-1.5	-0.5	1.1	0.1	-2.9	-2.6	-2.9	 	 
1991	-2.6	-2.5	-3.7	-3.7	-4.1	-5.9	-8.8	-7.4	-8.2	-6.4	-4.5	-4.1	-5.2	 	 
1992	-7.5	-6.4	-4.8	-5.9	-6.0	-5.7	-3.3	-5.6	-5.0	-4.9	-5.7	-6.1	-5.5	 	 
1993	-3.3	-3.9	-6.0	-4.4	-3.7	-4.7	-6.3	-6.3	-6.6	-9.4	-7.6	-7.6	-5.9	 	 
1994	-6.9	-10.0	-9.0	-9.1	-9.4	-8.6	-8.0	-7.1	-9.7	-8.5	-9.0	-9.1	-8.6	 	 
1995	-11.8	-10.2	-12.8	-12.9	-11.7	-11.4	-12.0	-11.9	-9.3	-9.8	-11.0	-10.7	-11.4	 	 
1996	-8.7	-7.7	-5.9	-8.0	-10.4	-9.9	-12.4	-13.7	-12.9	-11.7	-11.4	-11.6	-10.3	 	 
1997	-11.9	-12.8	-13.2	-12.7	-12.4	-14.3	-12.0	-12.4	-12.3	-11.4	-12.5	-12.1	-12.4	 	 
1998	-13.6	-16.7	-17.2	-16.7	-18.3	-18.6	-20.4	-22.0	-24.3	-26.2	-25.8	-26.6	-20.4	 	 
1999	-26.8	-24.8	-25.3	-25.0	-25.5	-26.6	-23.3	-20.7	-20.2	-20.5	-20.1	-19.0	-23.6	 	 
2000	-17.2	-17.1	-16.0	-16.8	-13.9	-12.8	-14.3	-15.4	-14.7	-13.9	-14.2	-15.6	-15.1	 	 
2001	-17.1	-17.0	-17.3	-17.2	-18.4	-17.7	-17.1	-17.9	-18.8	-18.2	-17.4	-16.8	-17.8	 	 
2002	-16.4	-17.0	-16.4	-15.8	-14.7	-14.0	-13.6	-11.6	-12.3	-12.4	-13.5	-13.1	-14.1	 	 
2003	-11.9	-11.1	-10.6	-10.2	-11.4	-12.0	-13.0	-14.2	-12.4	-11.9	-11.0	-11.0	-12.0	 	 
2004	-10.5	-10.1	-9.5	-10.2	-9.1	-8.0	-7.5	-6.4	-5.8	-7.1	-7.1	-7.2	-8.1	 	 
2005	-7.2	-7.9	-7.9	-7.3	-7.4	-8.7	-8.1	-8.8	-9.5	-8.3	-8.4	-7.7	-8.1	 	 
2006	-8.5	-7.1	-7.1	-7.2	-7.2	-6.6	-6.6	-6.7	-7.5	-10.5	-13.0	-14.5	-8.1	 	 
2007	-16.2	-16.5	-16.5	-15.7	-15.7	-16.6	-17.1	-16.1	-14.8	-12.7	-10.5	-8.8	-15.2	 	 
2008	-6.2	-5.5	-5.7	

[Level Jan 1989 99.5, Jan 2008 10.2; ratio 0.10, 
down 90%, or down 11.3% annual compounded.]

New vehicles

22 Apr 2008. BLS website.

(From the “Get detailed statistics” go to the row for CPI-U and choose the cell under the column “Create customized tables, one screen”)

Absolute values

Series Id:    CUUR0000SETA01,CUUS0000SETA01
Not Seasonally Adjusted
Area:         U.S. city average
Item:         New vehicles
Base Period:  1982-84=100

Year	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec	Annual	HALF1	HALF2
1988	116.1	116.0	115.7	115.6	115.9	116.1	116.1	115.9	116.2	117.2	118.4	119.0	116.5	115.9	117.1
1989	119.4	119.5	119.4	119.2	119.2	118.9	118.5	117.7	117.1	118.5	120.6	121.9	119.2	119.3	119.1
1990	122.4	122.2	121.6	121.1	121.0	120.6	120.2	119.9	119.6	121.1	122.8	124.3	121.4	121.5	121.3
1991	125.3	125.9	126.1	125.9	125.9	125.8	125.6	125.1	124.8	125.8	127.3	128.3	126.0	125.8	126.2
1992	128.7	128.9	129.1	129.1	129.2	129.1	128.6	128.5	128.3	129.1	130.6	131.3	129.2	129.0	129.4
1993	131.8	132.0	132.0	132.2	132.4	132.2	132.2	132.2	132.1	133.4	134.8	135.6	132.7	132.1	133.4
1994	136.1	136.5	136.8	136.9	137.2	137.4	137.4	137.3	137.5	138.4	139.4	140.1	137.6	136.8	138.4
1995	140.6	140.7	140.7	141.1	141.1	141.0	140.3	140.0	140.0	140.9	142.2	142.8	141.0	140.9	141.0
1996	143.2	143.5	143.6	143.5	143.4	143.5	143.2	142.9	143.2	143.8	144.8	145.4	143.7	143.5	143.9
1997	145.4	145.4	145.4	145.2	144.6	144.2	143.7	143.0	142.7	143.3	144.0	144.1	144.3	145.0	143.5
1998	144.4	144.4	144.4	144.3	143.3	142.6	142.7	142.8	142.3	142.5	143.5	144.1	143.4	143.9	143.0
1999	144.4	143.8	143.4	143.3	142.9	142.5	142.0	141.4	141.6	142.3	143.1	143.6	142.9	143.4	142.3
2000	143.3	143.0	143.3	143.5	143.3	142.9	142.5	141.9	141.4	141.6	142.7	143.6	142.8	143.2	142.3
2001	143.7	143.3	142.8	142.7	142.3	141.7	141.2	140.3	140.2	141.0	142.6	143.5	142.1	142.8	141.5
2002	142.7	141.2	140.7	140.4	139.8	139.2	138.7	138.1	138.7	139.5	140.4	140.6	140.0	140.7	139.3
2003	139.7	139.2	139.3	138.7	138.1	137.3	136.7	136.8	136.4	136.5	137.5	138.0	137.9	138.7	137.0
2004	138.0	138.3	137.9	137.6	137.4	137.2	135.9	134.9	134.9	135.9	137.9	138.8	137.1	137.7	136.4
2005	139.8	139.9	139.1	138.8	138.7	138.1	136.3	135.0	135.8	137.1	138.0	138.3	137.9	139.1	136.8
2006	139.3	139.3	138.8	138.4	137.7	137.2	136.9	136.4	136.3	136.8	136.8	137.1	137.6	138.5	136.7
2007	137.603	137.340	137.228	136.963	136.295	135.820	135.415	135.204	134.927	135.344	136.250	136.664	136.254	136.875	135.634
2008	136.827	136.279	135.727	
YOY change

12 Months Percent Change
Series Id:    CUUR0000SETA01,CUUS0000SETA01
Not Seasonally Adjusted
Area:         U.S. city average
Item:         New vehicles
Base Period:  1982-84=100

Year	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec	Annual	HALF1	HALF2
1988	1.2	2.2	2.3	1.9	1.8	1.8	1.5	1.7	2.1	1.9	1.8	2.2	1.8	1.9	1.8
1989	2.8	3.0	3.2	3.1	2.8	2.4	2.1	1.6	0.8	1.1	1.9	2.4	2.3	2.9	1.7
1990	2.5	2.3	1.8	1.6	1.5	1.4	1.4	1.9	2.1	2.2	1.8	2.0	1.8	1.8	1.8
1991	2.4	3.0	3.7	4.0	4.0	4.3	4.5	4.3	4.3	3.9	3.7	3.2	3.8	3.5	4.0
1992	2.7	2.4	2.4	2.5	2.6	2.6	2.4	2.7	2.8	2.6	2.6	2.3	2.5	2.5	2.5
1993	2.4	2.4	2.2	2.4	2.5	2.4	2.8	2.9	3.0	3.3	3.2	3.3	2.7	2.4	3.1
1994	3.3	3.4	3.6	3.6	3.6	3.9	3.9	3.9	4.1	3.7	3.4	3.3	3.7	3.6	3.7
1995	3.3	3.1	2.9	3.1	2.8	2.6	2.1	2.0	1.8	1.8	2.0	1.9	2.5	3.0	1.9
1996	1.8	2.0	2.1	1.7	1.6	1.8	2.1	2.1	2.3	2.1	1.8	1.8	1.9	1.8	2.1
1997	1.5	1.3	1.3	1.2	0.8	0.5	0.3	0.1	-0.3	-0.3	-0.6	-0.9	0.4	1.0	-0.3
1998	-0.7	-0.7	-0.7	-0.6	-0.9	-1.1	-0.7	-0.1	-0.3	-0.6	-0.3	0.0	-0.6	-0.8	-0.3
1999	0.0	-0.4	-0.7	-0.7	-0.3	-0.1	-0.5	-1.0	-0.5	-0.1	-0.3	-0.3	-0.3	-0.3	-0.5
2000	-0.8	-0.6	-0.1	0.1	0.3	0.3	0.4	0.4	-0.1	-0.5	-0.3	0.0	-0.1	-0.1	0.0
2001	0.3	0.2	-0.3	-0.6	-0.7	-0.8	-0.9	-1.1	-0.8	-0.4	-0.1	-0.1	-0.5	-0.3	-0.6
2002	-0.7	-1.5	-1.5	-1.6	-1.8	-1.8	-1.8	-1.6	-1.1	-1.1	-1.5	-2.0	-1.5	-1.5	-1.6
2003	-2.1	-1.4	-1.0	-1.2	-1.2	-1.4	-1.4	-0.9	-1.7	-2.2	-2.1	-1.8	-1.5	-1.4	-1.7
2004	-1.2	-0.6	-1.0	-0.8	-0.5	-0.1	-0.6	-1.4	-1.1	-0.4	0.3	0.6	-0.6	-0.7	-0.4
2005	1.3	1.2	0.9	0.9	0.9	0.7	0.3	0.1	0.7	0.9	0.1	-0.4	0.6	1.0	0.3
2006	-0.4	-0.4	-0.2	-0.3	-0.7	-0.7	0.4	1.0	0.4	-0.2	-0.9	-0.9	-0.2	-0.4	-0.1
2007	-1.2	-1.4	-1.1	-1.0	-1.0	-1.0	-1.1	-0.9	-1.0	-1.1	-0.4	-0.3	-1.0	-1.2	-0.8
2008	-0.6	-0.8	-1.1	

Comparing the CDC 6600 and my MacBook Pro

1963. Computer History Museum

“Some facts about the Control Data 6600”

http://archive.computerhistory.org/resources/text/CDC/CDC.6600.1963.102621029.pdf

and

12 Aug 2000. Doctor Dobb's Journal

http://www.ddj.com/184404102

“Control Data 6600: The Supercomputer Arrives. Dag Spicer”

and

2009. Apple Web site.

http://www.apple.com/macbookpro/specs-17inch.html (not a permalink)

[The original pamphlet advertising 1963's fastest supercomputer (first source above) makes interesting reading, especially with the photographs.]

CDC 6600 MacBook Pro 17”
Processor speed 3 MHz 2.8 GHz
Instruction cache 240 B 64 KB
CPU parallel 10 CPU functions dual core
Transistors 400 k 582 m
Memory 1 MB 4 GB
Memory parallel 32 banks -
Size 750 sq ft room 1” x 11” x 16”
Weight 12,000 lbs 7 lbs
Power 150 kW 35 w
Price $10 m $2500
Disk/drum 2 MB 500 GB

Image courtesy of the Computer History Museum.