Selling 100 Hyundai Accents is clearly not the same as selling 100 Bugatti Veyrons - one of the flaws of just counting unit sales. Tracking only unit sales can obscure what’s really happening in the marketplace.Tory Zimmerman
Everyone knows you can get statistics to say just about anything you want.
That's because not all stats are created equal.
In a new study, Statistics Canada takes a look at the three main types of measures it uses to track economic activity: physical quantities (the number of cars sold), dollars (total car sales) and volume (cars sold minus price changes).
Under some conditions, the three ways of measuring demand will move in tandem. But it's hardly a given.
Consider the car sales: Between 1993 and 2008, auto sales volumes doubled from $35-billion to $70-billion. But the actual units sold increased only 40 per cent, from 1.2 million to 1.7 million.
Only part of the sales volume increase is due to automakers selling more cars. The rest of the increase was caused by the change in the types of vehicles sold and a proliferation of pricey new features, such as ABS brakes and air bags.
Quantity data is great, Statscan says, because the numbers are quick and easy to collect, and they give a quick snapshot of how demand is changing, particularly for things that are pretty much the same, such as bulk commodities.
The flaw is that one car is just like another when you're tracking unit sales. Selling 100 Hyundai Accents is clearly not the same as selling 100 Bugatti Veyrons. Tracking only unit sales can obscure what's really happening in the marketplace. A shift to smaller cars, for example, could produce lower profits for auto makers, but higher sales.
"The number of vehicles sold doesn't capture the full diversity of the auto industry's numerous models," the report pointed out.
Another problem with quantity is that you can't tally up different units of measure. Energy exports, for example, are made up of very different sources, from cubic metres of gas, litres of gasoline to megawatts of electricity.
That's where current dollar stats come in handy. You get a single dollar value that reflects the diversity of products and prices.
But current dollar stats also have limitations, Statscan says. Sticking to the energy example. Energy exports rose sharply between 2002 and 2007. But to figure out whether exports rose due to higher prices or increased energy demand requires a third type of measure -- volume.
Volume, or so-called "real data," measure the entire spectrum, excluding price changes over time.
That's really important when you look at sectors, such as energy, where prices can fluctuate wildly. Volume captures what's really going on in the market, with the relative contribution of prices and units. That way you figure out whether energy exporters are generating higher sales based on higher prices, higher volumes or a combination of both.
The lesson is obvious. A single set of data can be very misleading. That's why it's always best to come at an issue from multiple angles.
Follow Economy Lab on twitter @Economy_Lab