Developing an Hedonic Regression Model For DVD Players In the U.S. CPI
Paul R. Liegey1
Background
DVD (Digital Versatile Disk) players were introduced into the US market in the spring
of 1997. Monthly sales of DVD players have been rapidly increasing, with calendar year
1999 sales nearly quadrupled from the 1.1 million players sold in 1998 (see attachments 1
and 2).
The CPI began reflecting price change for DVD players with the January 1998 revision.
DVD players are priced in the Other Video Equipment item stratum (RA03) along
with videocassette players/recorders (VCRs), video cameras/camcorders, satellite video
products and other video products/accessories. As of December 1999, the CPI reflected a
26.5 percent decline for the Other Video Equipment item stratum since its
inception in December 1997, with an average monthly decline of 1.1 percent (see attachment
3). Approximately 15 CPI quotes were being priced for DVD players by the end of 1999,
representing about 7 percent of the average monthly sample (n ~ 227) for the item stratum.
DVD players represent a sort of evolved hybrid of VCRs and laser disc players. Since
their introduction to the marketplace, consumers immediately noticed the sharper, clearer
picture and excellent sound quality they provide. All DVD players provide enhanced
features:
- All-digital playback that provides better video and audio detail than either VCRs or
laserdiscs
- Choice of aspect ratios supporting both standard (4:3) or widescreen (16:9) TVs
- Choice of up to 8 tracks of digital audio for multilingual playback
- Choice of up to 9 camera angles for different vantage points during playback
- Menus and simple interactive features including "instant" rewind and fast
forward, search by title, chapter, track, and timecode
- Automatic "seamless" branching of video for multiple movie versions
(theatrical vs. director's cut) or rating choice (G, PG, or PG-13 version of an R-rated
film)
- Choice of up to 32 multilingual subtitle tracks
These enhanced benefits are realized on all DVD players as long as the DVD discs
provide the information (that is, for example, not all DVD discs are coded to playback 8
different languages).
Retail prices for consumer-oriented DVD players currently range from $200 (on sale) to
more than $1,200, depending on the quality of the player and market environment. InfoTech,
Inc. reports that the average retail price for DVD video players has declined from $735 in
the first half of 1997 to $470 in the second half of 1998 (see attachment 4). The mean
price for all DVD players included in this study during the first half of 1999 was
$443.39. The nearly 40 percent decline (from $735 to $443) in the average retail price for
DVD players represents the dramatic reduction in price that is often associated with
relatively new products.
The greatest drawback of the current generation of DVD players is that they do not
record on a blank DVD disc either once (DVD-R) or multiple times (DVD-RW). Once the
"R" and "RW" formats become mainstream for DVD players, they are
likely to replace VCRs and laserdisc players as the video playback machine. DVD
player prices are already competitive with some VCR and laserdisc player prices and offer
a superior quality viewing experience.
Ideally, these three players should be modeled together so the CPI could be able to
quality adjust item substitutes from VCRs and laserdisc players to DVD players as consumer
purchases and preferences for video playback machines change through time. However, some
of the characteristics for each of the players are unique to the individual player, making
it difficult to include all three players in one model.
The 1999 hedonic model developed for DVD players is discussed below.
Data
Since the official CPI sample for DVD players was too small for regression estimation,
an additional sample was selected for regression use only.2 A similar approach
has been used to develop hedonic regression samples for CPI research on VCRs,
camcorders, and other products. This additional sample for DVD players was selected using
a process that mimics the process used to select the official CPI samples. More than 500
DVD player price quotes were sent to the field in April and May 1999 for collection in
existing CPI outlets. Unfortunately, many of the DVD player price quotes could not be
collected because the outlets were not yet carrying the new product. The resulting sample
of 282 DVD players was further pared down by another 23 quotes to exclude non-DVD player
items that had been captured as part of the sample in error. Also included in the 23
quotes that were excluded from the sample were multiple unique item hits (that is, the
same brand and model) for the same outlet. That is, if the same DVD player appeared more
than once in the same outlet, it was excluded so that no outlet had two or more quotes
with the same DVD player. The final sample size used for the DVD player model was 259
observations.
The bulk of the data preparation included review of the CPI checklist specifications
(see last attachment) for consistency, completeness, and accuracy. The 259 observations in
the sample represent 45 unique DVD players with a specific brand and model number (see
attachment 5). Review of the April / May data collected for this product revealed some
inconsistent, inaccurate, or incomplete specification descriptions.
To overcome these data deficiencies, the CPI specification elements (see last
attachment) collected in the sample were overwritten using information primarily from
manufacturers' websites for each of the 45 unique DVD players when such information was
available. Other secondary source information was used to cross check or supplement the
information on the manufacturers' websites. These other sources included Consumer
Digests' 1999 Annual Buying Guide (December 1998) and Consumer Reviews' Video
Buyers' Guide (Summer 1999). A summary of the collected sample and the overwritten
DVD sample is presented in the appendix by specification category and element as they
appear on the CPI checklist for DVD players (see attachment 6).
After the DVD player sample was edited for consistency, completeness, and accuracy,
variables were created for each of the specification elements on the checklist (see last
attachment). Unfortunately, some of the specification elements were useless because of
lack of variation. For example, in the DISC FORMAT specification category only the A2 (DVD
player) and A3 (combination LD and DVD player) specifications were eligible for variable
creation. Since the A2 specification accounted for 100 percent of the sample, however,
DISC FORMAT could not be used as a meaningful variable.
Similarly, in the TYPE OF PLAYER specification category (see last attachment), no
variation occurred in the B specification after the data was overwritten. In other words,
100 percent of the DVD players in the sample were tabletop or shelf
units and no meaningful variable could be created to include in the regression
model. As portable units make their way into the market and our sample, this category will
be more helpful.
In the COMPATIBLE FORMATS specification category (see last attachment), the C1 (plays
compact discs-CD) and G1 (plays recordable digital versatile discs-DVD-R)
specification elements reflected no variation. All DVD players play compact discs (C1),
and none of the 45 unique DVD players in the sample were rated by secondary source
information as being able to play the DVD-R (G1) format. Manufacturer and secondary source
information about whether a DVD player could read the E1 (plays recordable compact
discs CD-R) specification element was usually absent, and this specification
element was also not used because of unreliable data. Only the D1 (plays video
compact discs-CD-V) and F1 (digital video express compatible-DIVX)
could be used as potential variables in the regression model. The DIVX format, a competing
DVD format, is no longer being offered to consumers as of June 1999.
Other examples of specification elements (see last attachment) with little or no
variation after overwriting the sample with secondary source information were the
following:
- The I1 (plays region 1 only) specification in the REGIONAL ENCODING
category: All DVD players have a built-in regional code lockout feature (see definitions
attachment) which can be disabled by the manufacturer or retailer. Information on this
specification is sparse.
- The J1 (single disc) specification in the DISC CAPACITY category:
Forty-four of the 45 unique DVD brand and model numbers in the sample possess a single
disc capacity.
- The S1 (S-video output) specification in the VIDEO OUTPUT CAPABILITY
category: All DVD players have S-video output capability.
- The V1 (coaxial digital output) and W1 (optical digital output)
specifications in the DIGITAL AUDIO OUTPUT CAPABILITY category: All DVD players include at
least one digital audio output (see definitions attachment 8).
- The AD specifications in the COUNTRY OF ORIGIN category: This category was not
overwritten with secondary source information. The category was dominated by AD3-Japan
(~70%) and AD1-not available (~14%).
More variation in characteristics data is better than less when developing a hedonic
model. Systematically gathering information about technical DVD player specifications that
were too difficult to collect on the CPI checklist also proved to be difficult with little
result since not all manufacturers provide this information consistently.
Model
A hedonic model for DVD players estimating the average effect on price of
characteristics has been specified as follows:
Here the bis represent the
effect of the (observable) characteristics on the natural logarithm of price. The CPI
prices that were collected in this sample represent "retail offer" prices, and
approximately 20 percent of these prices were collected "on sale." The mean
price for all DVD players in the sample was $443.39. The mean price for
"regular" priced players in the sample is $467.21 with a standard deviation of
$214.52, and the mean price for "sale" priced players is $343.86 with a standard
deviation of $70.99. Since type of price (that is, regular or sale) is thought to have an
impact on price, a dummy variable for sale price was included in the model to capture this
effect, and its coefficient sign was negative.
A variety of information sources were used to formulate a priori
expectations about what influences price for DVD players. Manufacturers, retailers,
industry analysts, associations, and consumer/marketing organizations frequently focus on
the superior audio and video performance of DVD players in terms of "total viewing
experience" but usually this description is bench-marked or (implicitly) compared to
the current standard for video play back machines. Also, contributing to the superiority
of DVD players'; video playback capabilities are many of the enhanced features that were
mentioned in the background section of this paper. While these features undoubtedly
represent a greater array of consumer choice and presumably increase the viewing
experience utility for the average consumer, they are common to all DVD players and thus
exhibit no variation as potential variables.
The only (physical) characteristic that was available in the sample that is believed to
distinguish video performance among DVD players is component video output
(see definitions attachment 7). Industry and retail information indicates that this
characteristic contributes to the overall capability of a DVD player when connected to a
TV that possesses a component video input. While component video outputs are
often touted as a distinguishing quality feature, they are more commonplace today than
when DVD players were introduced to the market in early 1997. Indeed, this quality
characteristic was relatively common in the sample, occurring in more than 70 percent of
the observations. It was expected that component video output would influence
price positively since it can provide enhanced clarity and color.
Similarly, the only (physical) characteristic that was available in the sample and that
is believed to distinguish audio performance among DVD players is a built-in Dolby
digital decoder (see definitions attachment 7). All DVD players output a digital
audio stream that can produce surround sound like audio capabilities; however, a decoder
(either external or built-in) is needed to process this digital audio stream. While this
feature is expected to contribute to price positively, not all consumers might be
interested in this feature if they have already purchased an external Dolby digital
decoder. About 38 percent of the DVD players in the sample possess a built-in Dolby
digital decoder.
Both group and individual dummy variables were created for each of the 14 DVD player brands
represented in the sample. A priori expectations about how these brands
should influence price were based on retail and industry information. While no explicit
information was available about brand rankings and DVD player quality, some brands (and
models) are targeted toward more sophisticated users of consumer electronics while most
other brands (and models) are advertised simply by comparing the enhanced performance of
DVD players to VCRs. Since DVD is currently perceived to be the video playback
standard of the future, many manufacturers are racing to put their DVD player products on
the market now to establish a reputation for future sales. DVD player manufacturers such
as Marantz, Onkyo, Sony, Denon, Yamaha, and Mitsubishi were thought to cater to more
sophisticated users while other brands such as RCA, Proscan, Samsung, Panasonic, and
Toshiba seem to represent a more standard class of DVD players. Of course, individual or
unique models of players within a brand, especially in successive generations, might vary
in quality from the family of players offered by any given manufacturer. Brand
effects were included in preliminary regressions, excluding one to serve as a base.
Other characteristics data that were collected on the DVD checklist and were thought to
possibly influence price included: plays the compact video disc format, plays
the DIVX format, compatible with Digital Theater Systems (DTS), remote
control typestandard versus universal, and warranty coverage—1
year parts and labor warranty versus 1 year parts and 3 months labor.
In addition, control variables were created for city size, region,
and type of business as collected by the CPI. The use of these control
variables is to minimize any of the unexplained variation that might remain after the
model has been fitted with price determining characteristics.
Iterative regressions were performed until the remaining parameter estimates in the
model below exhibited relative robustness to the inclusion and deletion of other variables
not included. The direction and magnitude of the parameter estimates seem reasonable, and
the statistics pertaining to fit, significance, and collinearity are within generally
accepted limits.
Variable Category
|
Variable Name |
Parameter Estimate |
T Statistic |
Tolerance |
|
Intercept |
5.695137 |
282.426 |
|
Type of Price: |
Sale Price |
-0.101076 |
-5.006 |
0.8538723 |
Brand: |
Onkyo |
0.879601 |
10.035 |
0.9211067 |
Sony (high quality models) |
0.699876 |
24.037 |
0.5472271 |
Denon |
0.377169 |
5.235 |
0.9123200 |
Yamaha |
0.364231 |
3.017 |
0.9670469 |
Mitsubishi |
0.361707 |
5.083 |
0.9351258 |
Zenith |
0.296203 |
2.464 |
0.9756267 |
JVC |
0.269904 |
5.142 |
0.8696479 |
Proscan |
0.134653 |
3.546 |
0.5277970 |
Panasonic |
0.092632 |
3.407 |
0.7376674 |
Sony (standard models) |
Base |
|
|
Toshiba |
-0.067112 |
-3.144 |
0.6396037 |
Samsung |
-0.163521 |
-2.266 |
0.9093194 |
Audio Capability: |
Built-in Dolby Digital Decoder
|
0.265096 |
12.995 |
0.5561771 |
Compatible with DTS |
0.139823 |
6.935 |
0.6384225 |
Compatible with Dolby Digital
|
Base |
|
|
Warranty: |
One year parts and labor |
0.123135 |
4.617 |
0.5048170 |
One year parts/three months labor |
Base |
|
|
Type of Outlet: |
Full price video/audio outlet |
0.086722 |
4.408 |
0.7963235 |
Discount video/audio outlet |
-0.074741 |
-2.831 |
0.8613919 |
Control Variables: |
B size city/Midwest region |
-0.093542 |
-3.682 |
0.8994561 |
R2=0.9004 |
Adjusted R2=0.8930 |
F value=120.591 |
Number of observations=259 |
Liegey and Shepler (1999) as well as other researchers have acknowledged that in
"the (consumer) electronics market, brand plays an important role as a
price factor" when developing a hedonic model for a consumer electronic good. Given
the data limitations in terms of variation of collected characteristics and the difficulty
associated with collecting (technical) supplemental characteristics consistently across
brands, the final model above represents a snapshot of how the average consumer values
quality for DVD players in today's embryonic market.
Representing brand with individual dummy variables, an approach similar to
Ioannidis and Silver (1997) provides a better fit with less multicollinearity than
grouping brands by quality into low, medium, and high categories, an approach taken by
Moulton, Moses, and LaFleur (1999).
Initial parameter values seemed to be a bit low for the Sony brand compared to other
brand variables. Sony, an established leader in the consumer electronics market in terms
of reputation and market share, seemed to span such a large portion of the quality
spectrum, offering both standard and high end models. To compensate, variables for standard
and high quality models were created for this brand (see attachment 5 for
which model numbers were assigned to each group). Models included in Sony's high
quality group are currently rated by many industry analysts as state-of-the-art DVD
players. They include features such as dual-discrete optical pickups (one type of laser is
used to read DVD software and another is used to read CD software), a built-in Dolby
digital decoder, and a DVD "Navigator" remote control which provides the user
with enhanced search and scan capabilities.
Inclusion of the Sony high quality variable provided a considerably better
fit for the model, with the R2 statistic increasing from 80 percent to 89
percent. The Sony standard variable was used as a base variable for this
category since it represents a fair percentage of the sample (14 percent) and embodies the
usual features common to other players that consumers have come to expect from these video
players compared to VCRs. Other brand parameter estimates in the model above
appeared robust to the inclusion and deletion of other variables and were consistently
(statistically) significant in preliminary regressions.
While the audio variable parameter estimates represented in the model, built-in
Dolby digital decoder and DTS output, performed as expected in terms
of magnitude and direction, the video variable, component video output,
consistently reflected a (marginally) negative value that was sometimes statistically
insignificant depending on the particular regression being estimated. It was decided that
this video variable should be dropped from the model since it did not conform to a
priori expectations. (That is, its parameter estimate was consistently negative,
and contrary to market information this feature was not as rare—at least in this
sample—as is sometimes advertised.) In particular, more than 70 percent of the DVD
players in the sample possessed the component video output variable.
Furthermore, many consumers in today's market may not value or realize the enhanced
picture quality offered by component video output unless their monitor or TV
possesses a component video input. Relatively few TVs in today's households have component
video input.
Other variables that were tested for use in the model were the two (useable) format
variables, plays video compact discs and plays DIVX, as well as
variables for remote control type, standard versus universal, and manufacturer
warranty. The plays video compact discs feature is generally
advertised as an enhanced feature not available on all players and allows the user to
watch recordings in the older format. A little more than half the sample possessed this
feature but its sign consistently was negative when included in the model, and this
variable was highly correlated with built-in Dolby digital decoder (see
attachment 8). Since there was no practical or sensible way to create a new variable for built-in
Dolby digital decoder so as to alleviate the correlation with the plays video
compact discs variable, the plays video compact discs variable was
excluded from the model. Similarly, the plays DIVX format variable was
correlated with the few brands/models (RCA and Proscan) that offered this feature. In
addition, since this format was discontinued in June 1999, it would not be a useful
variable to leave in the final regression model.
While all DVD players come equipped with a remote control, some models try
to distinguish themselves from others by advertising a universal remote control.
Universal remotes give the user the enhanced benefit of being able to operate both the DVD
player and the TV with one device instead of two. A priori expectations were
that the presence of a universal remote control would add value to the
overall price of a DVD player since this feature represents added convenience to the user.
About 45 percent of the DVD players in the sample are sold with a standard remote
control while the remaining 55 percent were offered with a universal remote
control. When the universal remote control variable was included in
the model, the parameter estimate was consistently negative and (statistically)
insignificant. In addition, this variable is somewhat correlated with built-in Dolby
digital decoder (see attachment 9). Removing the universal remote control
variable and including a variable for standard remote contol produced
a similar multicollinearity problem. The model would benefit more from the inclusion of a built-in
Dolby digital decoder variable than one of the remote control
variables. Therefore, the variable for remote control was not included in the
model.
The last category of collected characteristics that seemed to have some potential in
explaining DVD player prices was manufacturers' warranty. The standard
manufacturer warranty offered with most DVD players today is one year parts and
three months labor. This type of warranty was offered on approximately 80 percent
of the DVD players in this sample. Most of the remaining players had warranties of one
year parts and labor. A longer warranty for labor (one year versus the base of
three months) might have a positive impact on price. When a variable for one year
parts and labor was included in the model, it was positive and (statistically)
significant. This variable remained relatively robust to the inclusion and deletion of
other variables and was included in the final model.
Various control variables were tested representing region of
the country, city size, and type of business as defined in the
CPI. The most consistently performing and statistically significant control variables that
were included in the final DVD model were full-priced video and audio equipment
stores, discount video and audio electronics stores, and B size
cities (populations 500,000-1,000,000) in the Midwest region. The purpose of control
variables is to minimize the unexplained variation that might remain after the
model has been fitted with price determining characteristics.
DVD Player Characteristics Excluded from the Model
Alexandre Balkanski (1998), president and CEO of C-Cube Microsystems, states that
"(DVD) players will be among the most complicated consumer electronics products ever
produced. Simply, the players consist of the following major components (see figure
below):
- Disc Reader Mechanism consists of a motor which spins the disc and the laser which reads
the information from it. The laser uses red light (as opposed to an audio CD player's
infrared laser).
- The DVD-DSP (digital signal processor) is an integrated circuit that translates the
laser pulses back into electrical form that other parts of the decoder can use.
- The Digital Audio/Video Decoder reconstitutes the compressed data on the disc,
converting it into studio-quality video and CD-quality audio for output to TVs and stereo
systems.
- Microcontroller controls the operation of the player, translating user inputs from the
remote control or front panel into commands for the audio/video decoder and the disc
reader mechanism. The microcontroller also implements parental lockout, dials distributors
for access codes, and controls decryption."
DVD Player Architecture
When the average consumer is faced with limited information about what (varying)
features are available on complicated electronic goods, like DVD players, they rely on brand
reputation for an aggregate measure of product quality, reliability, and service. Our
ability to collect specification data that may influence price is currently limited to how
manufacturers and retailers advertise features for DVD players.
Information about the component pieces referenced in the "DVD Player
Architecture" diagram above is central to how DVD players perform and could be used
to develop a better hedonic model for this product. For example, laser (pickups) referred
to in the "disc reader mechanism" above are used in optical storage systems
(such as DVDs) and are classified by the color that corresponds to the lasers
particular wavelength, which in turn depends on the characteristics of the materials used
to generate the laser light. Some DVD players employ only one laser to read DVDs and CDs
while more sophisticated players, like the high quality Sony players, use two different
lasers to read these formats more precisely. Industry reports also note that the use of
lasers in the blue and violet, that has a shorter wavelength than a blue laser, end of the
light spectrum will further enhance DVDs performance.3 This kind of
technical information related to DVD player performance is difficult for manufacturers and
retailers to communicate to the average consumer but may be essential in determining what
influences price.
Specification data that is less technical may serve as a proxy for quality if
sufficient variation exists in the sample being collected. This year's review in the Consumer
Digests' 2000 Annual Buying Guide (December 1999) notes that
"Now that the (DVD) format has proved itself, hardware manufacturers are readying
a raft of new features to help distinguish the entry-level $200 models from the deluxe
machines they're pitching to the videophile market. Deluxe DVD players come with jog/shuttle
dials on the player and the remote, and some offer clearer picture search,
variable speed picture search or a wider range of search speeds. A remote control is de
rigeur, even on the budget-priced models, but step-up models offer programmable
keys to allow adjustments of your system's volume, turning the video monitor on and off
and many other functions.
Other deluxe features include video black level expansion, which provides
another 7 percent of dynamic range; direct-digital outputs,
allowing connection to sophisticated audio gear, digital recorders and high-end surround
decoders; enhanced on-screen menus, making it easier to set up the player
and navigate through a disc with special features; smooth motion scanning at high
search speeds, usually bolstered by complex computer chips to store pieces of
video in RAM; marker functions to return to specific scenes; picture
zoom, which lets you blow up the image and reposition it, letting you closely
examine details at the press of a button; and just as there are with CD players, there are
multidisc changers, for stacking up to five movies at a time for all-day
entertainment."
Also potentially important in distinguishing DVD player performance is whether or not a
player is THX certified. Under this certification
process, developed by George Lucas' THX Division, a DVD player submitted by the
manufacturer first undergoes a comprehensive battery of tests using computer-controlled
test equipment programmed with proprietary THX routines. After a player finishes this
exhaustive ordeal on the test bench, it then "goes home" for extensive hands-on
use in a variety of systems and listening/viewing environments. The first four THX certified players were announced at the 1998
Winter Consumer Electronics Show and ranged in price from $1500 to $15,000
Hedonic Quality Adjustments and Price Index Simulation
Of the three different approaches or methods that use the results of hedonic regression
models to quality adjust price indexes, BLS employs the 'matched model' method in its
official indexes.4 This method controls for quality changes based on the
difference in product specifications between two items when a substitute observation, or
quote, occurs in the price index sample. It is important to note that under the 'matched
model' approach only substitution price changes, or quotes, are eligible for
hedonic quality adjustments.
DVD players, as noted above, are priced in the Other Video Equipment item
stratum (RA03) along with video cassette players/recorders (VCRs), video
cameras/camcorders, satellite video products, and other video products/accessories. Review
of attachment 10 reveals that in the second half of 1999 on average 11 price quotes
represented DVD players each month in the CPI. In addition, of these 11 DVD player price
quotes, only (less than) one price quote, on average, represented a substitution price
change. Since the Other Video Equipment item stratum (RA03) averaged about
227 price quotes per month in the second half of 1999 (see attachment 9), little reason
exists to believe that a quality adjusted index would differ from the published index for
this item stratum index.
Of the 6 DVD player substitution price quotes included in the 9906 through 9912 CPI
samples, 3 represented "model number only" specification changes which would
yield no quality adjustments for these substitute price changes. Although the remaining 3
substitution price quotes reflected bona fide specification changes, one of
these three price quotes reflected a "reinitiation" from a VCR player to a DVD
player and could not be adjusted since the hedonic model for this item only includes DVD
player prices and specifications. The remaining two substitution price quotes could have
been adjusted with the DVD player hedonic model using the parameter estimates for brand
and built-in Dolby digital decoder.
Quality adjusting substitution price changes for DVD players in the CPI will have only
a small index impact at the US level until a greater number of these products are
represented in the Other Video Equipment item stratum (RA03). The
introduction of new products that undergo rapid price and quality change is widely
recognized as a problem for price index estimation. BLS currently is testing a variety of
ways to bring new goods into the CPI more quickly than would occur through the normal
sample rotation process.
Conclusion
DVD players are an important product for the CPI to continue to represent in its sample
since they are an example of a new product that has become very popular with consumers of
electronic goods. It has taken "just two years from DVD's national launch in April
1997 for the format to ship 1 million units, compared with four years for compact discs
and 11 years for VHS. Also, DVD players have outsold the VCR by 13 to one and the CD
player by four to one" in terms of their first two years of unit sales.5
Almost five and a half million DVD players have been sold to consumers by U.S. retailers
as of January 2000.
As noted earlier, once the recording capability for DVD players becomes mainstream,
they should replace VCRs and laserdisc players as a dominant type of video playback
machine. DVD player prices are already competitive with some VCR and laserdisc player
prices and offer a superior quality viewing experience. Many of the enhanced features that
are advertised and make possible this superior quality viewing experience are common to
all DVD players.
The DVD hedonic model developed in this study represents a snapshot of how the average
consumer values quality for DVD players in todays embryonic market. Ideally, the
final regression would have contained more variables representing objective product
characteristics, and fewer brand variables that act as proxies for perceived product
quality. The direction and magnitudes of the parameter estimates seem reasonable, however,
and the statistics pertaining to fit, significance and collinearity are within expected
limits. The parameter estimates in the model exhibit relative robustness to the inclusion
and deletion of other variables not included in the final specification. The model should
be used in conjunction with commodity analyst judgment to quality adjust CPI quote level
substitution data when possible.
The CPI checklist for this product (see last attachment) is expected to be revised
prior to the planned collection of the FY 2000 hedonic DVD sample as a means of capturing
some of the deluxe non-technical specifications referenced above.
Notes
- The author wishes to thank Mary Kokoski, Lynn Reese, Nicole Shepler, Bill Thompson,
Beverly Heneghan, and John Greenlees for helpful suggestions.
- For additional information about the hedonics project on quality change in the CPI see
Charles Fortuna (1999), "Hedonics Project on Quality Change," Quality Quarterly;
Internal BLS Newsletter, Winter 1998-99, pp. 1-3. http://www.dcppi.bls.gov/CPIProg/docs/qq/win98.doc
(accessed November 1999).
- Murray Slovick, "Roses Are Red, Lasers Are Blue," E-Town, The home electronics
guide @ http://community.etown.com/news/articles/bluelasermsa62899.html,
(accessed November 1999). June 1999, pp. 1-5.
- For an informative description and evaluation of each of these hedonic approaches to
quality adjust price indexes, see Silver (1998), "An evaluation of the use of hedonic
regressions for basic components of consumer price indices."
- Yahoo! Business Wire, November 8, "DVD Player Pricing Is At All-Time Low For
Holiday Selling Season," pp1.
References
Consumer Digest's 1999 Annual Buying Guide (December 1998), pp. 84-87.
Consumer Digest's 2000 Annual Buying Guide (December 1999), pp. 84-86.
Consumer Review's Video Buyers Guide (Summer 1999), pp. 50, 56-57,
70.
Ioannidis, C. and Silver, M.S. (1997), "Estimating the Worth of Product
Characteristics," Journal of the Market Research Society, 3, 4, pp.
559-570.
Silver, M.S. (1998), "An evaluation of the use of hedonic regressions for basic
components of consumer price indices," Third Meeting of the International Working
Group on Price Indices, Statistics Netherlands, Voorburg: Netherlands (1998) 263-274. http://www.statcan.ca/secure/english/ottawagroup/pdf/23SIL3.pdf
(accessed November 1999).
Paul R. Liegey and Nicole Shepler, "Adjusting VCR prices for quality change: A
study using hedonic methods," Monthly Labor Review, September 1999, vol.
122, no.9, pp. 38-45.http://www.bls.gov/opub/mlr/1999/09/art3full.pdf
(accessed November 1999).
Brent R. Moulton, Timothy J. LaFleur, and Karin E. Moses, "Research on Improved
Quality Adjustment in the CPI: The Case of Televisions," in Walter Lane, ed., Proceedings
of the Fourth Meeting of the International Working Group on Price Indices, Bureau
of Labor Statistics, Washington, DC, January 1999, pp. 77-99. http://www.statcan.ca/secure/english/ottawagroup/pdf/18moul98.pdf
(accessed November 1999).
Alexandre Balkanski President and CEO of C-Cube Microsystems, "Digital Video Disc:
The Coming Revolution in Consumer Electronics," @ http://www.c-cube.com/technology/dvd.html,
(accessed November 1999). 1998, pp. 1-15.
Attachment 1
Attachment 2
DVD Player Sales
The following chart tracks monthly DVD player sales, back to the format's launch in
March 1997. Note that sales data from the current month is not included in this chart.
Please note that numbers indicate sales of players from manufacturers to U.S. retailers
only (Canadian sales are estimated at end of graph). LD Combo players are included in
these figures, but DVD-ROM drives are not. The numbers also include DIVX players — CEA
does not break them out. This chart is updated monthly. All sales data is courtesy of the Consumer Electronics Association (formerly CEMA), and
is used with permission.
DVD Player Sales
History (since format introduction) |
|
2000 |
1999 |
1998 |
1997 |
JAN |
N/A |
125,536 |
34,027 |
N/A |
FEB |
N/A |
109,399 |
34,236 |
N/A |
MAR |
N/A |
123,466 |
38,336 |
N/A (Format Launch) |
APR |
N/A |
269,107 |
42,889 |
34,601 |
MAY |
N/A |
279,756 |
47,805 |
27,051 |
JUN |
N/A |
326,668 |
79,044 |
29,037 |
JUL |
N/A |
325,151 |
84,709 |
19,416 |
AUG |
N/A |
260,225 |
81,170 |
34,021 |
SEP |
N/A |
501,501 |
113,558 |
34,371 |
OCT |
N/A |
603,048 |
163,074 |
56,407 |
NOV |
N/A |
449,242 |
136,908 |
37,657 |
DEC |
N/A |
646,290 |
233,505 |
42,575 |
Yearly Total |
N/A |
4,019,389
(as of 12/31) |
1,089,261
|
315,136 |
U.S. Grand Total |
5,423,786
(3/97 to 12/31/99) |
|
|
Source: http://www.thedigitalbits.com/articles/cemadvdsales.html
(accessed January 2000).
All sales data is courtesy of the Consumer
Electronics Association (formerly CEMA)
Attachment 3
Attachment 4
DVD Pricing Trends
Source: InfoTech, Incorporated http://www.infotechresearch.com/
(accessed November 1999)
Attachment 5
Brand / model number composition for
DVD hedonic sample |
|
|
|
|
|
|
|
|
|
|
|
|
|
Brand
Name |
|
Model # |
|
N (Model
#) |
N
(Brand) |
|
Mean Price |
% of Sample |
|
|
|
|
|
|
|
|
|
|
Denon |
|
DVD2500 |
|
1 |
|
|
|
|
|
|
|
DVD3000 |
|
2 |
3 |
|
682.98 |
1.2 |
|
|
|
|
|
|
|
|
|
|
|
RCA |
|
RC5200P |
|
2 |
|
|
|
|
|
|
|
RC5210P |
|
17 |
|
|
|
|
|
|
|
RC5220P |
|
1 |
|
|
|
|
|
|
|
RC5231Z |
|
18 |
|
|
|
|
|
|
|
RC5510P |
|
1 |
39 |
|
310.81 |
15.1 |
|
|
|
|
|
|
|
|
|
|
|
Samsung |
|
DVD709 |
|
1 |
|
|
|
|
|
|
|
DVD907 |
|
2 |
3 |
|
336.31 |
1.2 |
|
|
|
|
|
|
|
|
|
|
|
Sony |
|
DVPS300 |
|
4 |
|
|
|
|
|
|
|
DVPS330 |
|
1 |
|
|
|
|
|
Standard |
|
DVPS500D |
3 |
|
|
|
|
|
|
|
DVPS530D |
21 |
|
|
|
|
|
|
|
DVPS550D |
8 |
37 |
|
426.17 |
14.3 |
|
|
|
DVPC600D |
21 |
|
|
|
|
|
High quality |
|
DVPS7000 |
5 |
|
|
|
|
|
|
|
DVPS7700 |
9 |
35 |
|
848.84 |
13.5 |
|
|
|
|
|
|
|
|
|
|
|
Toshiba |
|
SD2008 |
|
8 |
|
|
|
|
|
|
|
SD2107 |
|
1 |
|
|
|
|
|
|
|
SD2108 |
|
2 |
|
|
|
|
|
|
|
SD2109 |
|
39 |
|
|
|
|
|
|
|
SD3108 |
|
1 |
|
|
|
|
|
|
|
SD3109 |
|
13 |
64 |
|
325 |
24.7 |
|
|
|
|
|
|
|
|
|
|
|
Yamaha |
|
DVDS700 |
|
1 |
1 |
|
699.95 |
0.4 |
|
|
|
|
|
|
|
|
|
|
|
Zenith |
|
DVD2000 |
|
1 |
1 |
|
399.95 |
0.4 |
|
|
|
|
|
|
|
|
|
|
|
JVC |
|
XVD701BK |
3 |
|
|
|
|
|
|
|
XV501BK |
|
3 |
6 |
|
454.98 |
2.3 |
|
|
|
|
|
|
|
|
|
|
|
Mitsubishi |
|
DD2000 |
|
1 |
|
|
|
|
|
|
|
DD3000 |
|
2 |
3 |
|
526.31 |
1.2 |
|
|
|
|
|
|
|
|
|
|
|
Onkyo |
|
DVS717 |
|
2 |
2 |
|
899 |
0.8 |
|
|
|
|
|
|
|
|
|
|
|
Panasonic |
|
DVDA105 |
|
2 |
|
|
|
|
|
|
|
DVDA110 |
|
3 |
|
|
|
|
|
|
|
DVDA112 |
|
2 |
|
|
|
|
|
|
|
DVDA115 |
|
1 |
|
|
|
|
|
|
|
DVDA120 |
|
9 |
|
|
|
|
|
|
|
DVDA310 |
|
4 |
|
|
|
|
|
|
|
DVDA320 |
|
1 |
|
|
|
|
|
|
|
DVDX410 |
|
7 |
29 |
|
396.64 |
11.2 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Phillips/Magnavox |
|
DVD420AT |
3 |
|
|
|
|
|
|
|
DVD815AT |
2 |
|
|
|
|
|
|
|
DVD825AT |
2 |
|
|
|
|
|
|
|
DVD850AT |
1 |
8 |
|
346.08 |
3.1 |
|
|
|
|
|
|
|
|
|
|
|
Pioneer |
|
DV414 |
|
7 |
|
|
|
|
|
|
|
DV606D |
|
1 |
8 |
|
428.8 |
3.1 |
|
|
|
|
|
|
|
|
|
|
|
Proscan |
|
PS8680Z |
|
20 |
20 |
|
423.69 |
7.7 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
All brands |
|
|
|
259 |
259 |
|
443.39 |
100 |
|
Attachment 6
Hedonic Specification Data for DVD Players (RA031-02)
Specification Category |
C&S Collected
Data(n=282) |
Overwritten Data
(n=259) |
Specification Element |
frequency |
/ % of Category |
frequency |
/ % of category |
|
|
|
|
|
|
|
|
|
|
Disc Format |
|
|
|
|
|
A1 Laser Disc
(LD) |
6 |
/ 2.2 |
0 |
/ 0 |
|
A2 Digital Versatile Disc
(DVD) |
254 |
/ 91.4 |
259 |
/ 100 |
|
A3 Combination LD and DVD |
8 |
/ 2.9 |
0 |
/ 0 |
A99 Other |
10 |
/ 3.6 |
|
|
|
Frequency missing
= 4 |
|
|
|
Type of Player |
|
|
|
|
|
B1 Portable |
12 |
/ 4.3 |
0 |
/ 0 |
|
B2 Portable with monitor |
1 |
/ 0.4 |
0 |
/ 0 |
|
B2 Tabletop or shelf unit |
261 |
/ 94.2 |
259 |
/ 100 |
B99 Other |
3 |
/ 1.1 |
|
|
|
Frequency missing
= 5 |
|
|
|
Compatible Formats |
|
|
|
|
|
C1 Plays compact discs (CD) |
239 |
/ 84.8 (239/282) |
259 |
/ 100 |
|
D1 Plays video CDs (CD-V) |
184 |
/ 77.0 (184/282) |
133 |
/ 51.4 (133/259) |
|
E1 Plays recordable CDs (CD-R) |
63 |
/ 22.3 ( 63/282) |
No
reliable data available |
|
F1 Plays DIVX |
47 |
/ 16.7 ( 47/282) |
45 |
/ 17.4 ( 45/259) |
|
G1 Plays Recordable DVDs (DVD-R) |
72 |
/ 25.5 ( 72/282) |
0 |
/ 0 |
H99 Other |
18 |
/ 6.3 ( 18/282) |
No
reliable data available |
|
|
|
|
Regional Encoding |
|
|
|
|
|
I1 Plays Region 1 (US and Canada) only |
236 |
/ 94.8 |
No
reliable data available |
I99 Other |
13 |
/ 5.2 |
|
|
|
Frequency missing
= 33 |
|
|
|
Disc Capacity |
|
|
|
|
|
J1 Single Disc |
239 |
/ 87.5 |
236 |
/ 91.1 |
J99 Other |
34 |
/ 12.5 |
|
|
|
Frequency missing
= 9 |
|
|
|
Audio Capability |
|
|
|
|
|
K1 Built-in Dolby Digital decoder |
161 |
/ 57.1 (161/282) |
97 |
/ 37.5 |
|
L1 Compatible with DTS |
184 |
/ 65.2 (184/282) |
182 |
/ 70.2 |
|
M1 Karaoke Function |
10 |
/ 3.5 ( 10/282) |
0 |
/ 0 |
N99 Other |
|
|
|
|
|
|
|
|
|
|
Video Output Capability |
|
|
|
|
|
R1 Component video output |
191 |
/ 67.7 (191/282) |
185 |
/ 71.4 |
|
S1 S-video output |
219 |
/ 77.7 (219/282) |
259 |
/ 100 |
T99 Other |
73 |
/ 25.9 ( 73/282) |
|
|
U99 Lines of horizontal output |
178 |
/ 63.1 (178/282) |
|
|
|
|
|
|
|
|
Digital Audio Output Capability |
|
|
|
|
|
V1 Coaxial digital output |
170 |
/ 60.2 (170/282) |
210 |
/ 81.1 |
|
W1 Optical digital output |
168 |
/ 59.6 (168/282) |
220 |
/ 84.9 |
X99 Other |
51 |
/ 18.1 ( 51/282) |
|
|
|
|
|
|
|
|
Remote Control |
|
|
|
|
|
Y1 Standard remote |
135 |
/ 47.9 (135/282) |
116 |
/ 44.8 |
|
Y2 Universal remote |
104 |
/ 36.9 (104/282) |
143 |
/ 55.2 |
AA1 Illuminated remote |
42 |
/ 14.9 ( 42/282) |
90 |
/ 34.7 |
AB99 |
|
|
|
|
|
|
|
|
|
|
Warranty |
|
|
|
|
|
AC1 One year parts and labor |
109 |
/ 39.8 |
48 |
/ 18.5 |
|
AC2 Two years parts and labor |
9 |
/ 3.3 |
3 |
/ 1.2 |
|
AC3 Three years parts and labor |
0 |
/ 0.0 |
|
|
AC99 Other |
156 |
/ 56.9 |
|
|
|
Frequency missing
= 8 |
|
|
|
Country of Origin |
|
|
|
|
|
AD1 Not Available |
26 |
/ 9.6 |
35 |
/ 13.5 |
|
AD2 United States |
18 |
/ 6.6 |
18 |
/ 6.9 |
|
AD3 Japan |
197 |
/ 72.4 |
182 |
/ 70.3 |
AD99 Other |
31 |
/ 11.4 |
|
|
|
Frequency missing
= 10 |
|
|
|
Attachment 7
Definitions of Select DVD Characteristics
Component video output : A video signal in which the brightness
(luminance) and color (chrominance) portions of the signal are processed separately. Where
S-video separates the luminance and chrominance portions of the signal, a component video
output goes a step further and splits the chrominance portion into two components.
Benefits include improved color accuracy and less color bleeding.
Digital audio output : All DVD players include at least one digital
audio output for sending the Dolby Digital bitstream to a Dolby Digital decoder (either a
stand-alone decoder or one built into an A/V receiver). The two most common types of
digital output are coaxial and optical. Both types
require a special cable to connect to the digital input of a Dolby Digital receiver or
decoder.
Dolby® Digital : A discrete multichannel digital audio standard
offering enhanced sonic realism. DVD can contain 5.1 channels of Dolby Digital sound; six
sound channels (left, center, right, left surround, right surround, plus a low-frequency
subwoofer channel, hence "5.1" not "6.0"). The "default"
sound format for DVD is Dolby Digital. Some DVD players come with a built-in Dolby
Digital decoder which can take the Dolby digital bitstream from
the disc and convert it into 6 channels of audio. DVD players without built-in decoders
need to connect to a receiver or processor that can take the Dolby digital bitstream from
the disc and convert it into 6 channels of audio to realize the enhanced audio effects.
DTS (acronym for Digital Theater Systems) : One of the rival sound
formats to Dolby Digital is called DTS (Digital Theater Systems). DTS is another six
channel (5.1)format. DTS uses much more storage space on a disc than Dolby Digital.
Generally higher-priced decoders and DTS-enabled DVD players are needed to hear DTS sound.
A DTS DVD MUST contain a Dolby Digital soundtrack in addition to the DTS soundtrack to
maintain compatibility, in accordance with the DVD specification.
Regional codes : DVD players have a built-in regional code lockout
feature, while the DVD discs may or may not contain a code (regional codes are optional on
DVD software). A player will be unable to play a disc that has a different regional code.
Discs may contain codes for more than one region or may not have any code, a feature which
allows them to be played on any player in any country. The regional code for USA/Canada is
"1."
Attachment 8
Select Pearson Correlation Coefficients for DVD Characteristics / N =
259
|
Play V-
CD |
Dolby
decoder |
DTS |
Comp-
onent |
Coaxial |
Optical |
S-Remote |
U-Remote |
I-Remote |
Play V-CD |
1.00000 |
|
|
|
|
|
|
|
|
Dolby decoder |
0.73720 |
1.00000 |
|
|
|
|
|
|
|
DTS |
-0.00777 |
-0.09010 |
1.00000 |
|
|
|
|
|
|
Component |
0.37619 |
0.48939 |
0.31788 |
1.00000 |
|
|
|
|
|
Coaxial |
0.49628 |
0.37378 |
-0.22792 |
0.26186 |
1.00000 |
|
|
|
|
Optical |
0.32458 |
0.30349 |
0.52921 |
0.57013 |
-0.17582 |
1.00000 |
|
|
|
S-Remote |
-0.36611 |
-0.44026 |
0.04224 |
-0.18660 |
0.27647 |
-0.27209 |
1.00000 |
|
|
U-Remote |
0.36611 |
0.44026 |
-0.04224 |
0.18660 |
-0.27647 |
0.27209 |
-1.00000 |
1.00000 |
|
I-Remote |
0.33717 |
0.55776 |
-0.00432 |
0.46154 |
-0.12366 |
0.30725 |
-0.44529 |
0.44529 |
1.00000 |
Attachment 9
Number and Distribution of CPI Price Changes for Other Video Equipment
(RA03) and DVD Players (RA031 - 02)
Other Video Equipment (RA03)
Month |
Total Prices Collected
|
Substitutions # / (% of Total Prices)
|
Comparable # / (% of Subs)
|
Non-comparable # / (% of Subs)
|
9906 |
233 |
53 / (23) |
28 / (53) |
25 / (47) |
9907 |
191 |
38 / (20) |
15 / (39) |
23 / (61) |
9908 |
233 |
34 / (15) |
15 / (44) |
19 / (56) |
9909 |
201 |
32 / (16) |
16 / (50) |
16 / (50) |
9910 |
247 |
26 / (11) |
11 / (42) |
15 / (58) |
9911 |
220 |
27 / (12) |
12 / (44) |
15 / (56) |
9912 |
264 |
14 / ( 5) |
2 / (14) |
12 / (86) |
|
|
|
|
|
Average |
227 |
32 / (14) |
14 / (44) |
18 / (56) |
DVD Players (RA031 - 02)
Month |
Total
Prices Collected
|
Substitutions # / (% of Total Prices)
|
Comparable # / (% of Subs)
|
Non-comparable # / (% of Subs)
|
9906 |
9 |
1 / (11) |
0 |
1 / (100) |
9907 |
11 |
0 |
0 |
0 |
9908 |
8 |
0 |
0 |
0 |
9909 |
10 |
1 / (10) |
1 / (100) |
0 |
9910 |
9 |
0 |
0 |
0 |
9911 |
15 |
3 / (20) |
3 / (100) |
0 |
9912 |
16 |
1 / ( 6) |
0 |
1 / (100) |
|
|
|
|
|
Average |
11 |
0.9 / ( 8) |
0.6 / (67) |
0.3 / (33) |
BUREAU OF LABOR STATISTICS U.S. DEPARTMENT OF LABOR
CONSUMER PRICE INDEX - ELI CHECKLIST______________________________________________________
collection outlet quote arranging
period: __ __ __ __ number: __ __ __ __ __ __ __ code: __ __ __ code: __ __ __ __
_________________________________________________________________________________________
ELI No./ cluster
title RA031 OTHER VIDEO EQUIPMENT code 02_
item availability: 1-AVAILABLE 2-ELI NOT SOLD 3-INIT INCOMPLETE
purpose of checklist: 1-INIT 2-INIT COMPL 3-SPEC CORR 4-SUB 5-REINIT 6-CHECK REV
_________________________________________________________________________________________
CURRENT PERIOD | SALES TAX
|
price _ _ _ _ _ _ . _ _ _ | included: YES NO
|
type of price: REG SALE |
|
|____________________________________________
|
YEAR-ROUND | in-season: JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
____________|____________________________________________________________________________
respondent: location:
_________________________________________________________________________________________
field message:
_________________________________________________________________________________________
CLUSTER 02 - VIDEO DISC PLAYERS
DISC FORMAT DISC CAPACITY
A1 Laser Disc (LD) J1 Single disc
A2 Digital Versatile Disc (DVD) J99 Holds multiple discs (CD or DVD),
A3 Combination LD and DVD
A99 Other, number, ________________________
______________________________ AUDIO CAPABILITY
K1 Built-in Dolby Digital Decoder
TYPE OF PLAYER (DD or AC-3)
B1 Portable L1 Compatible with Digital Theater
B2 Portable with monitor Systems (DTS)
B3 Tabletop or shelf unit M1 Karaoke Function
B99 Other type, N99 Other audio capability,
______________________________ ______________________________
COMPATIBLE FORMATS BRAND
C1 Plays Compact Discs (CD) P1 Denon
D1 Plays Video Compact Discs (CD-V) P2 JVC
E1 Plays Recordable Compact P3 Marantz
Discs (CD-R) P4 Mitsubishi
F1 Digital Video Express P5 Onkyo
(DIVX) Compatible P6 Panasonic
G1 Plays Recordable Digital P7 Philips/Magnavox
Versatile Discs (DVD-R) P8 Pioneer
H99 Other compatible format, P9 Proscan
P10 RCA
______________________________ P11 Samsung
P12 Sony
REGIONAL ENCODING P13 Toshiba
I1 Plays region one (United States P14 Yamaha
and Canada) only P15 Zenith
I99 Plays other regional encoding, P99 Other,
type, ________________________ ______________________________
MANUFACTURER'S MODEL NUMBER
Q99 ______________________________
SPECIFY ADDITIONAL INFORMATION ON RA031 PAGE 5 OF 16
ZZ99
_________________________________________________________________________________________
BLS 3400B (Rev. February 1995) RA031 page 4 of 16
Revised April 1999
RA031-02 - VIDEO DISC PLAYERS - CONTINUED
VIDEO OUTPUT CAPABILITY COUNTRY OF ORIGIN
R1 Component Video Output AD1 Not Available
S1 S-Video Output AD2 United States
T99 Other video output capability, AD3 Japan
AD99 Other,
______________________________
U99 Lines of Horizontal Output, ______________________________
______________________________ OTHER FEATURES
DIGITAL AUDIO OUTPUT CAPABILITY BA99 ______________________________
V1 Coaxial Digital Output
W1 Optical Digital Output BB99 ______________________________
X99 Other digital audio output, type
BC99 ______________________________
______________________________
OTHER PRICE FACTORS
REMOTE CONTROL
Y1 Standard Remote BD99 ______________________________
Y2 Universal/Learning Remote
AA1 Illuminated Remote BE99 ______________________________
AB99 Other remote control feature,
BF99 ______________________________
______________________________
** OTHER CLARIFYING INFORMATION
WARRANTY
AC1 One year parts and labor BG99 ______________________________
AC2 Two years parts and labor
AC3 Three years parts and labor BH99 ______________________________
AC99 Warranty, other type,
BI99 ______________________________
______________________________
_________________________________________________________________________________________
BLS 3400B (Rev. February 1995) RA031 page 5 of 16
Revised April 1999
Last Modified Date: October 16, 2001