Google effect gets starker for the media
Internet major assembles audiences in a far more cost-effective and targeted way
Alphabet Inc, which is almost entirely Google, had 98,771 employees as of December. That news, contained in the annual 10-K report the company released last week, got me thinking. Google, you may remember, passed the entire US newspaper industry in advertising revenue in 2010. The numbers are now not even remotely close.
This is not a perfect comparison: Google’s ad revenues are global, while the US newspaper industry’s are not. Still, it’s directionally correct, and pretty telling. So I wondered whether Google/Alphabet was about to do the same on the employment front. Formal answer: They’re getting there.
With 139,900 payroll employees as of December, US newspapers are still ahead of Alphabet. The trend is clearly not their friend, though, and there’s a twist. Last July, Bloomberg’s Mark Bergen and Josh Eidelson reported that Alphabet employs about as many “red-badged contract workers” as white-badged full-benefits staff.
If Alphabet’s contractors in fact outnumber its employees, then it may really have 200,000-plus people working for it, far more than the US newspaper industry. Yes, newspapers have stringers and other contract workers who don’t show up in their payroll employment counts, either, and the Alphabet numbers are global while the newspaper numbers are not. It’s not that meaningful a comparison. But it’s still unsettling.
In the second half of the 20th century, newspapers in the US made money mainly by selling ads. They used some of that money to hire journalists, but they also had lots left over for the owners. In 1997, the average operating profit margin of American newspapers was 19.5 per cent. Gannett Inc’s was 26.6 per cent. These big margins came under lots of criticism as newspapers began to struggle in the 2000s. I remember talking to Craig Newmark of Craigslist in those days and his message was that newspaper owners needed to stop whining so much about lost classifiedad revenue, stop laying off journalists, and get used to smaller profit margins.
They did get used to lower profits — Gannett’s operating margin for the four quarters ending in September was 6.1 per cent — but, as is clear from the above chart, they didn’t stop laying off journalists.
Now, of course, it’s Google and Facebook that make money selling ads and boast big operating margins: Alphabet/Google’s was 22.9 per cent in 2018, Facebook’s 44.6 per cent. They assemble audiences in a far more cost-effective and targeted way than newspapers ever did. In recent years, they have also gotten to be big employers (Facebook reported a headcount of 35,587 as of December 31). What they don’t do, with occasional exceptions, is hire journalists.
At first glance, it looks like the overall job losses stopped in 2010. But that ungainly and fast-growing category of “internet publishing and broadcasting and web search portals,” while it includes a lot of digital-only journalists, covers most employees at Alphabet and Facebook, too. “All other information services,” which has also been growing, includes, well, me, alongside libraries, archives and some other things. The broadcasting sector, which has shrunk a little but held up much better than newspapers and magazines, has lots of its employees on the entertainment side.
Estimated numbers
A more direct if less reliable and timely measure here is the annual count of “reporters and correspondents” from the Bureau of Labour Statistics’ annual Occupational Employment Statistics. There were an estimated 38,790 in the US in May 2017, down from 53,060 in 2006. The same basic story, in other words — with the added twist that by now there are probably fewer reporters and correspondents than Facebook employees.
Lots of other jobs have been wiped out by technological change over the centuries, so it can seem a bit self-involved when a journalist dwells on lost journalism jobs. OK, it is self-involved. But journalists do play a useful societal role, one that has yet to be effectively taken over by artificial intelligence, crowdsourcing or other innovations. And while a mix of new business models, previously neglected older business models (subscriptions, mainly) and non-profit funding sources will probably be enough to keep national journalism going on a tolerably effective scale, at the local level, things really don’t look encouraging.
Scholars are starting to quantify the effects of this loss of local journalists. A study published in December by two communications professors and a political scientist concluded that local newspaper closures made voters more likely to vote a straight political ticket, thus increasing partisan polarisation. “When they lose local newspapers, we have found, readers turn to their political partisanship to inform their political choices,” the authors concluded. Earlier last year, a study by three finance professors found that municipal borrowing costs go up 0.05 to 0.11 of a percentage point after a local newspaper closes, with the authors arguing that “loss of government monitoring” accounted for the rise. That doesn’t sound like a lot, but as Bloomberg’s Danielle Moran explained, “For a $65 million debt issue, that amounts to about $71,500 annually — enough to cover a teacher’s salary — or about $2 million over the life of a 30-year bond.”
Google does lots of useful things, and it now employs lots and lots of people. But they generally don’t show up at city council meetings and ask difficult questions.