Home > Causes of Inflation, Rant, Theory > The Internet Has Not Killed, and Will Not Kill, Inflation

The Internet Has Not Killed, and Will Not Kill, Inflation

Every few years or so, this story goes around to great acclaim: inflation is dead, killed by the internet. Recently, we have been hearing this story again, quite loudly. The purchase of Whole Foods by Amazon helped bring commentaries like these to the fore:

Credit Suisse’s Varnholt Says Internet Killed Inflation” (Bloomberg)

Low U.S. Inflation? It’s Your Phone: BlackRock Bond Manager” (New York Times)

Amazon Deal for Whole Foods Casts Doubt on Fed’s 2% Inflation Goal” (Barron’s)

And the list goes on and on. These are some of the more-reputable outlets, and they simply misunderstand the whole phenomenon. This isn’t unusual; almost no one really understands inflation, partly because almost no one these days actually studies something that most people presume isn’t worth understanding. (But pardon my ranting digression.)

The internet has not killed, and will not kill, inflation.

In the late 1990s, the internet was having a much greater relative impact. We went from having essentially zero internet in 1995, to a vast array of businesses in 1999 – most of whom were busy transferring money from capital markets to consumers, by raising equity investments which were then use to subsidize money-losing businesses (see especially: Amazon). And inflation? Core CPI in 1999 was 1.9% (Median CPI was 2.03%).

“But there’s more internet now than there was then!” runs the natural objection. Yes, and the internet was dramatically more impactful in 2001 than it was in 1999. Indeed, as the penetration of the internet economy exploded further despite the recession of 2000-2001, core inflation rose to 2.8% (Median CPI topped out at 3.33%) by late 2001.

There is always more innovation happening, whether it’s the 1940s or the 2010s. Innovation is a relatively steady process on the economy as a whole, but very dramatic on parts of the economy – and we tend to fixate on these parts. But there is no evidence that Uber is any more transformative now than Amazon was in the late 1990s. No evidence that Amazon now is any more transformative than just-in-time manufacturing was in the 1980s (in the US). And so on.

“But the internet and mobile technology pervades more of society!” Really? More of society than the J-I-T manufacturing innovation? More of society than airlines and telephones, both of which were de-regulated/de-monopolized in the 1980s? More of society than personal computers did in the 1990s? We all like to think we are living in unique times full of wonder and groundbreaking innovation. But here’s the thing: we always are.

“But Amazon bought Whole Foods and disrupted the whole food industry! How can you be more pervasive than food?” It remains to be seen whether Amazon is able to do what Webvan and FreshDirect and other food delivery services have been unable to do, and that is to remake the entire delivery chain for food at home. But let’s suppose this is true. Food at home is only 7.9% of the consumption basket, which is arguably less than the part of society that Amazon has already reorganized. Moreover, it’s a highly competitive part of society, with margins that are already pretty thin. How much fat is there to be cut out by Amazon’s efficiency? Some, presumably. But after Amazon makes some kind of profit on this improvement, how much of a decline in food prices could we see? Five percent, over five years? 10%? If Amazon’s “internetification” of the food-at-home industry resulted in a 10% decline in prices of everything we buy at the grocery store, over five years, that 2% per year would knock a whopping 0.16% off of headline inflation. Be still, my heart.

“In any event, this signals that competition is getting ever-more-aggressive.” No doubt, though it is ever so. But here is the big confusion that goes beyond all of the objections I’ve previously enumerated: microeconomic effects cause changes in relative prices; macroeconomics is responsible for changes in the overall price level. Competitive pressures in grocery may keep food prices down 10% relative to price increases in the rest of the economy. But suppose the money supply doubles, and all prices rise 100%, but food prices only rise 90%. Then you have your 10% relative deflation but prices overall still rose by a lot. If the governments of the world flood economies with money, no amount of competition will keep prices from rising. This is why there wasn’t deflation in 2010, despite a massive economic contraction in the global financial crisis and concomitant cutthroat competition for scarce customers in many industries.

So inflation isn’t dead, and neither is this myth. It will come back again in a few years – I am sure of it.

  1. Peter MCT
    June 21, 2017 at 4:06 pm

    No, but Amazon has 🙁

    • June 21, 2017 at 4:28 pm

      Dang it Pete, read the article FIRST before you comment! I address Amazon specifically!

  2. Marshall Jung
    June 26, 2017 at 10:48 am

    Playing the long game here…but what happens when Amazon is the dominant player for entire sections of the retail industry?

    Anecdotally, I have peers that pay more for everyday items (even with Prime) just because it’s convenient to have it delivered. I don’t know how to factor in the “savings” of additional time to dedicate elsewhere or effort involved with driving to a retail outlet.

  3. Peter
    December 13, 2017 at 6:50 pm

    Historically would agree with your sentiment Michael, however when looking at relative timeline of Moore’s Law, at some point in the near future transformational technologies appear to exceed inflationary pressures. Up until recent history, whilst technology may appear to be rapidly changing to the casual observer, it has been at a relatively low point in terms of development cycle. The next few decades will likely challenge long held economic dogma as the rate of change surpasses our ability to predict near future developments and initially at least, there will be increasing economic disparity as process technology removes requirement for skilled operators in favour of unskilled labour. Working in process engineering, I frequently see robots implemented to save labour which invariably require significantly more skilled labour than they displaced in order to maintain up-time. As this quirk of development process passes into the AI era and labour requirement diminishes, the case for inflation becomes weaker.

    • December 14, 2017 at 10:22 am

      I disagree. I think those are really neat developments, but not as significant as, say, the cotton gin or the combustion engine, neither of which killed inflation. The problem is that you are focusing on the knowns while the unknowns, and the unknown unknowns, will have a larger role in the future than the knowns will (which is why sci-fi always sounds ridiculous thirty years on, even though based on solid science at the time). FWIW, many people believe Moore’s Law is reaching a limit, and therefore future computational advances will be SLOWER. While this is distinct from the accelerating APPLICATIONS of that computing power, it’s something that needs to reasonably be taken into account. And that’s the sort of thing I mean. Everything you say about AI and robotics is true. It’s just an incomplete story. And while it’s POSSIBLE that the amount of money in circulation relative to productive output will no longer, for the first time in ten thousand years, matter for inflation…”history counsels caution” to use one of Greenspan’s famous phrases!

      Thanks for writing!

  1. October 10, 2017 at 12:43 pm
  2. October 11, 2017 at 6:25 am

Leave a Reply

%d bloggers like this: