Dec 17, 2009

Can Newspaper Ad Revenue Recover in 2010?

This is a short excerpt of a longer post I am preparing on newspaper print advertising trends, but a post today by Alan Mutter encouraged me to publish this small excerpt.

Mutter expresses skepticism about the rosy advertising scenario presented by U.S. newspaper executives in a survey conducted by Kubas Consultants. As this table from the Kubas report shows, newspaper executives are hoping for flat advertising revenue in 2010, with a 15% increase in online (approximately 10% of 2009 revenue) offsetting small declines in every print category.

I am particularly curious about the poll's findings regarding retail advertising, which after the steep declines in classified advertising over the past few years, will account for just over 50% of total print ad revenues in 2009, making it the biggest driver of the overall results.

The following graph (click to enlarge) plots retail ad revenue per subscriber against per capita personal consumption expenditures (PCE) for 1952 - 2009. Both the ad revenue and PCE data have been adjusted to constant 2008 dollars using the consumer price index (CPI-U).

Presumably there is a relationship between the amount of goods and services companies sell to consumers each year (represented by PCE) and the amount those companies are willing to spend to reach each consumer or household by advertising in newspapers.

And indeed there is a relationship, or more accurately, several relationships. From 1952 through 1990, newspapers sold retail advertising equal to about 1.76% of personal consumption expenditures on a predictable basis. Coming out of the 1990 recession, however, there was a clear downward shift in the relationship and for the next decade, retail ad sales ran at about 1.52% of personal consumption. (see note 1).

After the 2001 recession, there was another shift down in the relationship between ad spending and personal consumption; and for the next five years, retail ad spending in newspapers was basically flat in real terms. It's worth noting that in real terms, retail ad spend actually peaked in 2000.

Finally, since 2006, real retail ad revenue has plummeted as the recession took hold (in December 2007) and consumers abandoned their free-spending ways.

Recessions often act as antidotes to inertia by making households and companies (link added 04/15/10) reassess how and where they spend their money. Advertisers almost certainly used the past two recessions to reassess their spending in newspapers, and it's a fair bet they'll do the same as the current recovery unfolds.

A slavish reading of the most recent trend suggests that retail ad spending will simply continue to plummet like a rock falling off a cliff. This could happen, but I doubt it. Ad spending, even in newspapers, should respond to an eventual recovery. But making economic predictions around recessions and recoveries is notoriously difficult since prior recessions and recoveries provide limited relevant precedent. To paraphrase Tolstoy, if periods of prosperity are all alike, every recession unfolds in its own way.

But I'll take a stab at it. In a related analysis, I have compared changes in retail ad spending to changes in PCE on a quarterly basis. This analysis suggests that it will take real growth of 3.2% in per capita PCE just to keep retail ad spending per subscriber flat. It would take a sharp "V-shaped" recovery to reach that level of growth, but the WSJ has published a consensus estimate of 3.8% for GDP growth for 2010.

If GDP does grow at 3.8% and per capita PCE matches this growth, retail ad spend per subscriber could come in at +1.5% in real terms. Add 2.0% for inflation if you believe that newspapers will have any pricing power, then subtract 3% for subscriber attrition and you get 0.5% growth.

More pessimistically, if you assume GDP growth (and PCE) come in at 1.7% (the average from the 1992 and 2002 recoveries), retail ad spend per subscriber could come in at minus 4.0%. Assume no pricing power and subscriber attrition of 5% and overall retail ad spending would shrink by 9%

The precision of the numbers should not mislead about the confidence of the forecaster, but for my money, I think a range of flat to down 10% for newspaper print advertising feels about right.

note 1. A more precise, but likely less clear statement about the relationship would say, "From 1952 to 1990, for every $1,000 increase in real consumption expenditures per capita, spending on retail ads (per subscriber) in newspapers increased by 1.76% of this amount, or $17.60." For each of the regression lines, the intercept terms are small enough to make me comfortable with statement in the text.

note on data sources. Annual retail ad spend and paid circulation data comes from the Newspaper Association of America website. Personal Consumption Expenditure data, Consumer Price Index - All Urban and Population data are from the St Louis Federal Reserve Bank's FRED database.


  1. Robert, I like your approach to the numbers. I have also looked at correlation of newspaper ad rev to PCE (in my analysis I added residential real estate spending), and agree that the current pattern is a break from that of prior recessions. See my graph here:

    If you'll allow a bit scatology here, during my publishing days when my ad managers would predict positive performance after a whole series of negative months, I would ask them if they knew the reason turds are tapered. The answer is that if they were not, your ass would slam shut. The point for the present situation is that you don't come off four consecutive quarters of 25-30% losses and go straight to breakeven. The curves never look like that. A 10% loss for 2010, which we more or less agree on, would actually be a pretty good performance after the disaster of 2009.

    But looking further out, there is simply no real recovery possible. What publishers need is growth, not more palatable losses. And they don't have a business model that will produce growth.

  2. Martin, thanks for the comment and the reference to your earlier post.

    I wish I'd come across it before; I spent a long time staring at these numbers thinking there must be some mistake. I'll be adding some more to these thoughts and I'll be sure to point readers to your work.

    As for the longer-term outlook, I agree. At a 1.4% steady-state growth in real PCE per capita (the average for the past 20 years) current trends suggest that retail ad revenue per subscriber will decline at 4.8% in real terms.

    Subtract your own guess for subscriber attrition and add back two percentage points for inflation if you're feeling generous and I'm pretty sure the conclusion is 5% to 10% annual declines into perpetuity.

  3. Standard and Poors has issued a research note, summarized here,that forecasts slow or negative revenue growth for print-related media in 2010, even with an economic recovery.