Audio-visual and computing equipment in the CPI

The audio-visual and computing equipment subgroup of the consumers price index (CPI) had a combined expenditure weight of 1.82 percent at the June 2008 quarter. The sources and methods used to compile the audio-visual and computing equipment subgroup are explained in this article.

Position in the CPI structure

The recreation and culture group of the New Zealand Household Expenditure Classification represented 9.54 percent of the expenditure in the CPI at the June 2008 quarter. The audio-visual and computing equipment subgroup makes up about a fifth of the expenditure in this group. Table 1 outlines the expenditure weights of the audio-visual and computing equipment subgroup and its classes.

Table 1

Expenditure weight estimation

Information from the 2006/07 Household Economic Survey (HES) and other sources was used to determine the relative importance of the recreation and culture group. Over 70 percent of expenditure on goods and services in the recreation and culture group came from the HES. The HES collected detailed information on the spending patterns of about 2,600 households. Market research information was also an important source of information.

To obtain an estimate of expenditure at the June 2008 quarter, HES and other information for the year to June 2007 was 'price updated' to the June 2008 quarter. The International Labour Office recommends that CPIs are price updated and Statistics New Zealand follows this common international practice. The effect of price updating is that the underlying 2006/07 quantities are expressed in the prices of the June 2008 quarter. In general, the indicators used to price update the weights were the lower-level CPI indexes corresponding to the goods and services in the new basket. However, because of the time gap between the 2006/07 HES and when the new basket of goods and services was first used in the June 2008 quarter, there was the possibility of relative quantity shifts for some goods and services.

To reflect significant volume shifts in goods and services, 'volume adjustments' were made to some items to improve the relevance of the expenditure weights. It should be noted that volume change is considered to encompass not only changes in physical quantities, such as the number of desktop computers purchased, but also changes in quality.

The volume adjustments were made in two ways: either directly to the 2006/07 expenditures prior to price updating, or incorporated into the indicators used to price update 2006/07 expenditures. Table 2 lists the volume adjustments that were made to audio-visual and computing equipment, the reasons for making these adjustments, and the data sources used to determine whether to adjust and by how much.

Table 2

Item and sample selection

The CPI basket of goods and services is reviewed about once every three years to ensure it continues to reflect household purchases. As the available technology for audio-visual and computing equipment has changed rapidly, the goods and services in the CPI basket must also change. For example, 23 inch black and white 'consolettes' were introduced into the CPI basket in the 1960s, but now many consumers watch free-to-air and pay channels on larger flat-panel colour television sets. Consolettes are no longer included in the CPI basket, while free-to-air digital television receivers and a variety of flat-screen TVs are now included in the CPI basket. The article Fifty years of television in New Zealand details changes to televisions over the past five decades and is available on the Statistics NZ website.

The selected sample of products tracked in the CPI represents household purchases, and was largely determined by analysing retail transaction data obtained from GfK.

The CPI basket has changed to reflect recent changes in the audio-visual and computing market. Cathode ray tube (CRT) television sets, which had been in the CPI basket for over forty years, were removed as part of the 2008 CPI basket review. Flat-panel televisions were first added as part of the 2006 CPI review and the range was widened as part of the 2008 review to explicitly include both LCD and plasma display flat-panel television sets. Free-to-air digital television receivers were also added in 2008, reflecting the introduction of Freeview.

Video cassette recorders were added to the basket in 1983 and were replaced in 2006 with combination VCR/DVD players alongside the introduction of DVD player/recorders. DVD players were added to the basket in 2002. The DVD/VCR combination player was then removed in 2008. In 2008, recordable DVD discs replaced video cassette tapes and music digital downloads were added.

Other audio-visual and computing equipment items that were added to the basket as part of the 2006 review include:

• home theatre systems, which complemented stereo systems already in the basket
• MP3 players, which replaced personal portable compact disc players
• laptop computers (desktop computers were already in the basket)
• digital recording media (recordable compact discs and digital camera memory cards).

Television repair was removed from the basket in 2006, as expenditure had fallen.

Price collection

Prices for audio-visual and computing equipment are collected in two ways: field collection and Internet collection. Most audio-visual and computing equipment prices are collected at retail outlets by Statistics NZ's staff. Prices for digital downloads and free-to-air digital television receivers are collected from the Internet. Prices for free-to-air digital television receivers from the Internet complement direct price collection by Statistics NZ staff.

Prices for audio-visual and computing equipment are collected from suitable retail outlets in the 15 CPI urban areas: Whangarei, Auckland, Hamilton, Tauranga, Rotorua, Napier-Hastings, New Plymouth, Wanganui, Palmerston North, Wellington, Nelson, Christchurch, Timaru, Dunedin, and Invercargill.

Each quarter, prices for audio-visual equipment are collected from about 45 appliance stores, 20 department stores, 30 camera or photo shops, and 30 automotive parts stores (for car audio). For computing equipment, prices are collected from about 45 appliance or computer stores and 15 department stores. Prices for recording media are collected from about 45 appliance or computer stores, 20 department stores, 30 camera or photo shops, 25 record stores, and 15 book or stationery shops.

Digital download prices and free-to-air digital television receiver prices are collected from the Internet around the middle of the quarter.

Estimation

Each quarter, regional average prices for each audio-visual and computing equipment item are calculated by outlet-weighting the prices collected at different outlets within each region. All regions are aggregated to obtain the New Zealand quarterly item index by weighting together regional price movements from the base (June 2008) quarter to the current quarter, using regional population weights.

Prices for digital downloads are national prices and used to directly calculate the New Zealand item-level index. This is because these prices are collected from the Internet and apply nationally with no regional variation.

Prices for free-to-air television receivers collected from the Internet are combined with collected prices for each region prior to the calculation of the New Zealand index.

Quality assurance

The aim of the CPI is to measure the price change of the same product at each sampled outlet or business over time. In practice, the sample of products changes as products become unavailable, change, or no longer reflect purchases made by New Zealand households. When this occurs, there may be a change in the quality of the product and an adjustment is made so that only the estimated 'pure' price change is shown in the CPI. For example, a price increase (or decrease) that is deemed to be purely the result of better (or poorer) quality materials or service should not affect the CPI.

The quality adjustment of prices collected for audio-visual and computing equipment is a difficult task for statistical agencies worldwide. Price change shown in indexes in the audio-visual and computing subgroup can result from either changes in shelf prices or from changes in the quality of monitored products. For example, if an LCD television monitored in the CPI in the March 2011 quarter was superseded by a later model of better quality in the June 2011 quarter, and the shelf price was $1,999 in both quarters, this would be shown as a price fall in the LCD television price index.

Following price collection each quarter, analysis is undertaken to identify and appropriately adjust for any quality changes in the sample of products to ensure that only 'pure' price change is shown in the CPI. This is done in a number of ways depending on the product that is being analysed. For computers, if the sampled computer priced in the previous quarter is no longer on offer, a replacement product of the same brand is collected. Provided the products are not too dissimilar, a quality adjustment would be made based on the estimated value of the different characteristics (eg hard drive space, processor speed, etc) when comparing the specifications of the old and new products. The estimated value of the difference in characteristics is based on the price of different options for these features obtained from available market information. If the model of a particular computer monitored in the March 2011 quarter, with 2Gb of RAM, was upgraded when prices were next collected in the June 2011 quarter to 4Gb of RAM, an adjustment would be made by using the difference between the market value of 2Gb of RAM versus 4Gb of RAM.

For flat-panel television sets, a quality assessment is required whenever a model is superseded and replaced. Several options are considered depending on the comparability of the replacement product, and the relationship between the price and the quality of the replacement product when compared with the previous product. The 'comparable replacement' method is used when the replacement product is judged to be very similar in quality to the old product, such as a newer model with only small superficial changes. In this case, any change in shelf price between the old and new models is shown in the price index. When the replacement is judged to be of different quality to the old product, the method used is to infer the pure-price movement from products that are directly comparable from within the same region as the product being replaced. This is called the 'class mean imputation' method. Given the downward trend in television prices, class mean imputation often results in a price fall being recorded in the CPI for flat-panel televisions when products are replaced.

External influences on prices

The technology used in audio-visual and computing equipment is evolving rapidly. The range and quality of audio-visual and computing equipment are influenced by technology upgrades. Many sampled products, such as computers, digital cameras, multi-function (printer/scanner/copier) devices, and stereos, may no longer be available at the time of price collection (three months apart) or may quickly become unrepresentative. For goods with short model lives, newer products are introduced into the CPI basket as product replacements are made. However, new goods that are significantly different to existing goods in the CPI basket are often not introduced into the CPI until a three-yearly review of the basket is undertaken.

Care is required as to when these new technology goods are introduced into the CPI basket. This is due largely to the product 'life cycle'. The product life cycle is often characterised by new goods being first introduced at higher prices, then prices falling as their popularity grows. Prices may then remain largely unchanged for some time as the product 'matures', before falling again as the product is superseded by newer, higher quality products. Finally the product exits the market.

If an item is introduced into the CPI basket early in its life cycle, the weight assigned to the item in the CPI might quickly become too low. This is because volumes of purchases are often lower in the early stages of the product life cycle and then grow quite quickly. This may result in the CPI not fully reflecting the initial price falls associated with new products. However, if the weight is based on expected future sales, this may over-represent these price falls. This emphasises the need to frequently reweight items in the CPI basket to reflect the growth in sales.

If the introduction of a new technology good or product into the CPI basket is left too late, the CPI may miss the price falls that characterise the beginning of a new good or product's life cycle. This could cause the CPI to overstate price change for audio-visual and computing equipment.

Back to Price Index News: July 2011