A Commodity Perspective On Recent Trends in Retailing

by R. Lussier, D. McDowell and E. Cryderman*

Introduction

The 1990s witnessed a number of important changes in the retail industry. The “big box” phenomenon crossed the border from the US to Canada. “Power centers”, or groupings of big box stores, sprouted across the country, dramatically changing the face of retailing. Numerous store types subscribed to the one-stop shopping philosophy, expanding both their physical size and selection of merchandise. For instance, drugstores and general merchandisers actively sought to tap new markets: drugstores moved into food and sundries, while general merchandisers moved into drugs and expanded their food sections. Diversification, however, was not universal. Department stores spun-off specialized stores to handle furniture and household goods while reducing these lines at their regular stores.

The Monthly Retail Trade Survey (MRTS), although extremely important, could not always answer probing questions on retail trade. For example, for years the trend of food shipments by manufacturers and retail sales of supermarkets and other food stores was very close (Figure 1). But they diverged significantly in the mid-1990s, with shipments of food continuing to grow while retail sales of food slowed and actually dropped in 1995-96. This trend baffled analysts, with no apparent explanation in the available data (although it was clear that people had not stopped buying food and manufacturers had to be shipping it somewhere). Analysts believed that a shift in food sales by store type was a possible reason, one that could not be detected by trade group sales estimates but would be by a commodity-based survey.

Figure 1

All of the above emphasized the need for an ongoing commodity-based survey. Previous commodity surveys were irregular, with gaps of up to 15 years between surveys. Data from the new Retail Commodity Program provide timely answers to a myriad of questions that were hitherto unavailable. What were Canadians buying from retail stores? From what store types were they buying it? Have personal spending patterns changed? When do Canadians buy their goods? Where do Canadians buy given commodities? What do certain store types sell? What is happening to market share?

The Retail Commodity Program

In the mid-1990s, a goal was set to obtain commodity-based estimates for both total retail trade and for each of the existing trade groups. First, a workable and useful commodity classification was developed. Second, the survey was designed so that the aggregate industry and commodity estimates were consistent.

The resulting program consists of two components: the Quarterly Retail Commodity Survey and the Monthly Commodity Survey of Large Retailers. (The former includes the latter). The same commodity detail is collected in both surveys.

The Quarterly survey covers about 8,000 of the 14,000 business in the MRTS sample. The objective of the Monthly Survey of Large Retailers is to provide the most timely commodity estimates possible. To do so, the survey is based on a select group of 80 large retail enterprises that effectively form a longitudinal panel. These enterprises represent Canada’s largest retailers of food, clothing, home furnishings, electronics, sporting goods, and general merchandise. Together they represent about 38% of total retail sales (excluding recreational and motor vehicle dealers).

What do consumers buy – the supply perspective

It is critical for retailers to know what consumers are spending their money on, in order to respond to shifting consumer demands. The Retail Commodity Program is able to track the sales share of over one hundred commodities. In 2002, sales of new and used motor vehicles, parts and labour represented 29% of total retail sales. To this, one can add sales of automotive fuels, oils and additives (7% of total retail sales) to get a picture of transportation-related spending made in retail stores. Second in size is food and beverage sales, some 22% of all sales. Next are clothing and accessories with 9% and home furnishings and electronics with 8%.

Have spending patterns changed?

The Retail Commodity Program sheds light on the changes in the importance of each commodity in the consumer’s retail basket. In 1997, food and beverage sales represented 24.3% of retail sales. Although food and beverages has increased between 1.4% and 4.0% annually from 1997 to 2002, their share dropped significantly to 21.7% of retail sales in 2002.

Evidently something else has taken a bigger share of the consumer dollar. Table 1 indicates that health and personal care products account for a larger portion of what consumers spend. In 1997 they represented 6.4% of retail sales, but this surged to 7.5% in 2002, largely due to rising drug sales. Drugs include prescription drugs, as well as over-the-counter drugs, vitamins, herbal remedies, and health supplements.

Home furnishings and electronics were also a big winner, increasing from 7.0% to 7.9% of retail sales over the 5-year period.

Finally, automotive fuel, oil and additives is another commodity that took a larger share of the retail dollar, partly due to large price increases for gasoline.

Table 1: Shift in Commodity Sales Share

When do Canadians buy goods?

Seasonality is an important factor when analyzing retail trade data. However, trade group data often mask more distinct seasonality at the commodity level. The influence of holidays, long weekends, back to school events, and other events (Mother’s Day, Valentine’s, Halloween) can have a much greater impact at a specific commodity level than at the trade group level. Take for example candy sales around Easter or Halloween. Supermarket sales may be up slightly in October due to Thanksgiving, or in March or April due to Easter. However the spike in candy sales around this time is lost when lumped together with all other products sold at a food retailer. Commodity level seasonality provides much better information.

Although almost every commodity has a seasonal pattern, commodities that seem similar on the surface can have different seasonal patterns. Clothing is a very seasonal commodity. However, within clothing, seasonality varies from commodity to commodity.

Taking a closer look at the seasonal pattern for men’s and women’s clothing (Figure 2), the primary peaks in December, although in the same direction, differ greatly in strength. The seasonal factor for men’s clothing is significantly stronger in December than for women’s, presumably because men receive a greater proportion of their clothing as Christmas gifts.

Figure 2

Women’s clothing has a secondary peak in May, while men’s clothing has a secondary peak in June. One explanation is clothing purchases around Mother’s Day and Father’s Day. On the other hand, there does not seem to be much of an Easter effect.

Boy’s and girl’s clothing sales follow similar seasonal patterns (Figure 3). However, infant’s clothing does not have the same back to school peak as boy’s and girl’s.

Figure 3

Where do Canadians buy commodities?

It is often assumed that consumers have relatively stable buying patterns and that given commodities are bought from certain store types. With the fierce competition between store types, the same commodity can now be found in a variety of different stores. Where do people shop for certain everyday products? Where are they most likely to buy certain commodities?

Table 2 shows that in 2002 the majority of food and non-alcoholic beverage sales were made by supermarkets. But general merchandise stores surpassed speciality food stores with 9.7% of all food and non-alcoholic beverage sales, while drug stores sold 1.4%. This may not seem large, but every percentage point represents $666 million of food sales.

Table 2: Share of Food and Non-alcoholic Beverage Sales in 2002

What do stores sell?

It is revealing to see the various commodities sold by a trade group. To illustrate the point, Table 3 presents the different commodities sold by motor and recreational vehicle dealers. Around 55% of the sales are accounted for by new cars and trucks (including mini-vans and SUVs). Used automobiles represent 26%, while the rest is spread across parts, oil, labour receipts, new and used recreational vehicles.

Table 3: Share of Motor and Recreational Vehicle Dealers Sales
in 2002

Are some trade groups gaining market share?

Market share is a very important measure of success for most retailers. With over $307 billion in retail sales in 2002, every retailer and industry analyst wants to know who has the biggest piece of that pie or who is growing faster.

In addition, understanding market share is an important step towards understanding the often fickle consumer. Clothing store sales illustrate how market share can change. Table 4 shows that between 1997 and 2002, general merchandise stores (including department stores) lost clothing share. Gains were made by specialised clothing stores and sporting goods stores, with the latter increasing their market share by over half.

Table 4: Share of Clothing Sales by Trade Group

Conclusion

Statistics Canada has had a long history of collecting commodity detail for the manufacturing sector. However, until fairly recently, its attempts for the retail sector had been sporadic. The establishment of the Retail Commodity Program in 1997 has proven to be a useful analytical tool for the retail industry. The information generated can be used to track commodity sales within and across retail industries, to calculate commodity market share of various types of retail stores, and to gain a better understanding of the rapidly changing retail industry. The data show the type of outlets where consumers prefer to buy certain commodities, and shifts in the different types of commodities retailers decide to sell. Analysis of this data assists in establishing trends in commodity sales over time and in detecting and analysing seasonal factors in the retail industry.

In addition, various levels of government use the data to monitor the economy and consumer confidence. The authors are not aware of any other country in the world with such an extensive commodity program. The Retail Commodity Program appears to be a distinguished feature of the Canadian Retail Trade Statistics Program.

Recent commodity data, for about 30 series, are available free of charge through the Canadian Statistics website (www.statcan.ca/english/Pgdb/econ85.htm and www.statcan.ca/english/Pgdb/trade39.htm). Approximately 120 retail commodity series are available from both the Quarterly Retail Commodity Survey and the Monthly Commodity Survey of Large Retailers in CANSIM table numbers 080-0009 and 080-0010.

Recent feature articles

Note

* Distributive Trade Division. For more information contact Client Services (613) 951-3549 or retailinfo@statcan.ca.