I write this for journalists, first year economics students, and the general (non-economist) public. Yes I'm being pedantic. But I hope it helps. I'm trying to clear up a very common confusion in how we talk about inflation.
Your car has an odometer, which measures kilometers driven. And your car has a speedometer, which measures kilometers per hour.
The CPI price level is like the odometer. It tells you how high the price level is (compared to some arbitrary base year, which is set at 100 for convenience).
The CPI inflation rate is like the speedometer. It tells you how quickly the price level is rising (usually per year).
[OK, assume your car is a vertical rocket ship, or at least pointing uphill, if that helps. And the CPI is just the price of a basket of goods, so it averages out the prices of the individual goods.]
The only difference is that Statistics Canada only checks the odometer once a month. And the speedometer (usually) measures how many kilometers traveled per year, compared to the same month last year (so they don't have to worry about seasonal adjustment). And the speedometer can record a negative speed, which means the car is going in reverse and the odometer is going backwards (remember Ferris Bueller?)
If you say "the Consumer Price Index (CPI) rose by 1.2% in November, compared to 0.9% in October and 1% in September" that actually means something quite different from what you probably think it means or meant to say. [That's an actual quote from a UK newspaper, except I have substituted the actual 1.2% for the 1.1% economists had predicted.]
If you take that quote literally, it means that the price level was [a little more than] 3.1% higher in November 2016 than in August 2016. And if the price level rises by 3.1% every 3 months or quarter year, that means the inflation rate is [above] 12.4% per year. Which is a very high inflation rate for any non-basket case country. [Ignore the stuff in square brackets; it's just to keep the math nerds off my back, because a 10% increase followed by another 10% increase adds up to a 21% increase not 20%.]
What actually happened, and what the journalist presumably meant to say, was this:
This is the long way of writing it: "The CPI price level rose by 1.2% from November 2015 to November 2016; it rose by 0.9% from October 2015 to October 2016; it rose by 1% from September 2015 to September 2016"
This is the short way of writing it: "The 12 month CPI inflation rate was 1.2% in November, compared to 0.9% in October and 1% in September"
Inflation was low, and below the Bank of England's target.
The CPI price level rose in all three of those months, compared to the same month one year before. The 12 month CPI inflation rate did rise from October to November, but it fell from September to October.
A rising price level means the inflation rate is positive. Your car is going forward and the odometer is rising.
A rising inflation rate means the price level is rising more and more quickly. Your car is accelerating and the speedometer is rising.
A falling price level means that the inflation rate is negative. Your car is going backward and the odometer is falling. Another word for this is deflation.
A falling inflation rate means the price level is rising more and more slowly. Your car is decelerating and the speedometer is falling. Another word for this is disinflation.
unless by "A falling inflation rate" you mean a negative rate shouldn't the last sentence read "It means you car is accelerating more slowly and the speedometer is rising more slowly"?
Posted by: Bryan K | December 13, 2016 at 11:14 AM
Bryan: No, it's right as I wrote it. The odometer is rising more slowly.
Posted by: Nick Rowe | December 13, 2016 at 11:21 AM
We certainly think differently.
To me, the odometer is like GDP. Both measure the distance traveled. Both are a snapshot taken at a point in time.
The speedometer is measuring how much distance is traveled during a predefined time period. (A rate, distance per period). This would be equivalent to the quantity (difference) between two GDP measurements.
So what is inflation? Inflation is a change in the scale itself. With inflation, we begin measuring distance in miles, then we move to a new scale where each mile is longer than 5280 feet.
Am I wrong? :-(
Posted by: Roger Sparks | December 13, 2016 at 11:54 AM
Roger: You are talking about the level of output (GDP) and the growth rate of output. I am talking about the level of prices (CPI) and the growth rate of prices (inflation rate).
Since we measure prices in terms of money, there is no difference between saying "The price of the CPI basket of goods is rising in terms of money" and "The price of money is falling in terms of the CPI basket of goods."
The trouble is, as soon as you say "the price of money" a lot of terribly wrong people assume you are talking about the rate of interest, and start talking about "the money market".
Posted by: Nick Rowe | December 13, 2016 at 12:51 PM
This seems like a good analogy to me, though I suppose I am not in a position to judge.
There is one other thing I think worth mentioning to a general audience: if one takes "literally" the quote that "the Consumer Price Index (CPI) rose by 0.1% in November", it likely means that the measured CPI rose by something like 0.18%, because the quoted single-month rate (or CPI level) is seasonally adjusted, and a ballpark adjustment for November in the US or UK is around -0.8%. Many people do not appreciate how large monthly variations can be in comparison to absolute monthly rates. The reason for quoting year-on-year inflation (November 2015 to November 2016) is that washes out possible errors in the estimates for seasonal adjustment.
Posted by: Phil Koop | December 13, 2016 at 03:36 PM
Phil: good point. I've added a short bit saying they do year over year so they don't have to worry about seasonal adjustment.
Yep, you are probably not the best person to judge whether the odometer/speedometer analogy works! You will get it of course, but will my intended audience?
Posted by: Nick Rowe | December 13, 2016 at 03:47 PM
Cars that go fast are generally considered better than cars that don't go as fast. I'm not sure that translates into a desirable normative statement about an economy :-). Other than that, I think your post is very clear.
Posted by: Oliver | December 13, 2016 at 04:18 PM
I remember Ferris Bueller!
Finally a post I feel comfortable with (Bueller and CPI). Thanks, Nick! ;-)
The odometer/speedometer analogy works quite well, although I'm not sure if it adds much. Most people probably understand that a price can rise or fall, and that it can rise or fall fast or slowly. What might be harded to understand is the basket and the year-on-year comparison (instead of month-on-month), and there the odometer/speedometer analogy offers no help?
But it's always useful to try to put things a bit differently, and just trying to think in terms of odometer/speedometer is most likely helpful -- due to the ensuing discussion, both in this thread and in the reader's mind.
Posted by: Antti Jokinen | December 13, 2016 at 04:19 PM