Virtually all sources are wrong

Before explaining the title of this post, let me point out that  I will be interviewed on monetary policy by Charlie Deist for 1 hour tomorrow morning (8-9am Sunday, Pacific time, or 11-12am, EST) Here is the link.

Here is Wikipedia:

Virtually all sources agree on John D. Rockefeller being the richest American in history.

The second place is disputed, held by Andrew CarnegieCornelius VanderbiltJohn Jacob Astor IVBill Gates or Henry Ford depending on the source; most sources agree on Carnegie. Further places are a matter of even bigger debate.

This is just silly—John D. Rockefeller was not even in the top 10 of the richest Americans, even adjusting for inflation.  So that led me to wonder if Wikipedia’s claim was true.  Alas, I discovered it was true.  Virtually all sources agree that Rockefeller was the richest American of all time (even though he wasn’t even close.)  The New York Times agrees.  Forbes agrees.  Business Insider agrees.  CNNMoney agrees.  How could this happen?  How could virtually all sources be so wrong?

Let’s start with the facts.  Rockefeller first hit $1 billion in net worth in 1916.  That money was worth about $23 billion in 2018 dollars.  When he died in 1937 he was worth about $1.4 billion, roughly $24 billion in 2018 dollars.  That’s well down the Forbes billionaire list.  You could argue the CPI is flawed, as the cost of mansions has probably rising faster than the overall CPI.  But the CPI understates the improvement in health care, tech, etc.  So Rockefeller was clearly far less wealthy than Jeff Bezos (worth well over $100 billion), or Bill Gates at his peak in 1999 (worth close to $150 billion in 2018 dollars.)

So why are virtually all sources just completely wrong about something so easy to check?  Two reasons:

1.  Laziness

2.  Motivated reasoning

They all seem to have relied on someone’s bright idea to calculate real wealth by taking each individual’s share of the national economy at the time they were alive, then multiplying that figure by the size of the current US economy.  So if Rockefeller’s wealth was 1.5% of GDP when he died, then 1.5% of current GDP is roughly $300 billion.

To which I can only say: Huh?

First of all, why GDP and not the share of national wealth?  More importantly, even if you buy this procedure, Rockefeller clearly was not the richest American ever.  Who was?  I don’t know his name, but there is one American who had a wealth equal to roughly $100 trillion.  Yup, I said trillion, with a “T”.  Who was that lucky ducky?  He was the first man to walk across from Russia to Alaska, roughly 15,000 years ago.  The moment he arrived, he and his family controlled 100% of American wealth, for the simple reason that he was the only American.  And since Americans now own about $100 trillion in wealth, that’s how rich he was—15,000 years ago.  (Or only $20 trillion, if you prefer using GDP.)

I suppose if you want to be picky, then you could argue that America was not yet a country.  OK, we became a country around 1776, or 1783.  Even using the latter figure, our population was less than 1% as large as today.  So this “fraction of national income” technique for comparing wealth over the generations would imply that the average American back in 1783 was more than 100 times wealthier than the average American is today, for the simple reason that there are 100 times more Americans today.  Each American today tends to hold, on average, only about 1/100 as a big a share of national wealth as the average American held in 1783.  When Paul Revere rode past Lexington and Concord, they were packed with families worth at least $10,000,000, more likely $100,000,000.

I hope I don’t need to go on with this nonsense.  I presume that people in the media got lazy and took someone else’s technique, even though that technique was completely crazy, because they liked the conclusion.  They wanted to be able to say, “Yeah, you think Bill Gates is rich, he was nothing compared to Rockefeller.”  In other words, they wanted to sound like our dads.

And I think this explains much of modern progressivism and libertarianism.  People want to believe X, so they latch on to all sorts of dubious data points that support the conclusions that they have already decided they would like to reach.  I also think this helps explain the popularity of blogs like Slate Star Codex, Marginal Revolution, etc.  People go to those blogs knowing that the writers are willing to take an honest look at the data, and not try to fit the data to their preconceived ideas.



25 Responses to “Virtually all sources are wrong”

  1. Gravatar of Jeremy Bancroft Brown Jeremy Bancroft Brown
    24. February 2018 at 19:50

    This is hilarious, thanks.

  2. Gravatar of E. Harding E. Harding
    24. February 2018 at 20:50

    Great post, Sumner; never really thought of the reductio ad absurdum before (and never bothered to check the CPI calculations).

    I also think this helps explain the popularity of blogs like Slate Star Codex, Marginal Revolution, etc. People go to those blogs knowing that the writers are willing to take an honest look at the data, and not try to fit the data to their preconceived ideas.

    MR is kind of a mess, but, yes, psychiatrist Schlomo Schlomovich does try to take an honest look at the evidence to the best of his ability every once in a while.

  3. Gravatar of Doug M Doug M
    24. February 2018 at 21:06

    Using $20 billion current dollars as the estimate for Rockefeller’s wealth, for most of the 20th century he he was the richest man in American history. It wasn’t until the mid 90’s that Gates passed this mark. And since then about about 20 others.

    It sounds like the record keepers just don’t want to update their data.

  4. Gravatar of dtoh dtoh
    24. February 2018 at 21:27

    Who was the greatest basketball player of all time?

    Who was the greatest accumulator or wealth?

    Does “richest” equal “highest net worth?”

  5. Gravatar of Lorenzo from Oz Lorenzo from Oz
    24. February 2018 at 21:29

    Also, economising on information. What is the minimum I can get away with to seem smart/cool/morally sound? In a world of info-tsunamis, the demand for markers of righteousness allowing one to economise on information goes up and up.

    So, folk can get into ever more intense pools of mutual self-righteousness while being increasingly comparatively ignorant.

    I am currently writing a book on marriage — I have a publisher’s contract and all — and reading about sex, gender, social psychology, comparative anthropology, etc has made me increasingly aware of the economising-on-infomration effect. It is particularly extreme in matters Muslim/Islamic but is a much wider pattern. Worth, I have decided discussing in an Afterword.

    That, and how toxic social constructionist views are. Plus the similarities between the 1920s war against the Demon Drink and the current. much more widely prohibitionist, war against the Demon Domination. (Both largely male demons, and both targets of flushed-with-success women’s movements.)

  6. Gravatar of dtoh dtoh
    24. February 2018 at 21:57


    So one more question. Growing up in the suburban Midwest, someone in my neighborhood who lived in a 1000 square foot house would have been extremely poor. Someone living in the same neighborhood 500 years ago in a house half the size would have been extremely rich.

    So how could a poor person be richer than a rich person.

    It seems to me that “richness” is a measure of wealth relative to one’s peers not a measure of absolute net worth.

  7. Gravatar of ssumner ssumner
    24. February 2018 at 22:05

    Doug, You said:

    “It sounds like the record keepers just don’t want to update their data.”

    No, that’s not the problem, the problem is the technique they used, which is beyond idiotic.

    dtoh, You said:

    “It seems to me that “richness” is a measure of wealth relative to one’s peers not a measure of absolute net worth.”

    That’s a reasonable argument, but it doesn’t in any way provide a defense for these claims about Rockefeller. If that was their goal they might have wanted to deflate his wealth by a wage index. Or compare his wealth to average wealth at the time. But what they did is just idiotic–indeed idiotic squared.

  8. Gravatar of MFFA MFFA
    25. February 2018 at 05:24

    The reductio ad absurdum is a master act. My general rule looking at what comes out of the mainstream medias is too assume that journalists have absolutely no clue of what they are writing about, and be on the watch to identify the few exceptions. Of course a lot of people just take everything at face value. We need more Sumners!!

    Ps: did you consider correcting the Wikipedia page? The list of inflation corrected net worth (the normal definition, not the crazy stupid one user here) should be much more close to the standard of Wikipedia in terms of accuracy than this “most sources agree…” argument

  9. Gravatar of Jeff Jeff
    25. February 2018 at 06:03

    There’s an easy correction. Use share of contemporaneous GDP per capita, rather than share of GDP itself, as the measure. So your ancient Siberian hiker would, on arrival, have just 100 percent of per capita GDP, which numerically is 1.

    I found a source that says GDP per capita in 1916 was $5459 in 1990 dollars. In 1916 dollars that’s about $446, so Rockefeller’s number in 1916 is $1 billion / $446 is about 2.24 million.

    If Bill Gates in 1999 was worth $150 billion 2018 dollars, then since the CPI is about 50 percent higher now than it was in 1999, Gates had $100 billion in 1999 dollars. GDP per capita in 1999 was $34,602 so Gates number then would be roughly 3 million.

    So Gates was about 34 percent richer than Rockefeller by this measure, and Jeff Bezos at $100 billion 2018 dollars is about equal to Rockefeller.

  10. Gravatar of Jeff Jeff
    25. February 2018 at 06:06

    Actually, Bezos would be some 10 percent or so less wealthy than Rockefeller.

  11. Gravatar of Jeff Jeff
    25. February 2018 at 06:12

    And even that calculation is wrong. It would be correct if Bezos had 100 billion 2018 dollars in 1999, but he didn’t. He had them in 2018, when GDP per capita is about $60000, so his number is about 1.67 million, about two-thirds of Rockefeller’s 2.24 million.

  12. Gravatar of Philo Philo
    25. February 2018 at 08:01

    Yes, people “latch on to all sorts of dubious data points that support the conclusions that they have already decided they would like to reach”; but why focus on modern progressives and libertarians? This is a universal human tendency.

  13. Gravatar of John John
    25. February 2018 at 08:26

    If we want to discuss the strong version of this argument, you could argue that captured flow of economic income is a proxy for relative economic power, which isn’t nearly so absurd (even if the exact math wasn’t done quite right).

  14. Gravatar of ssumner ssumner
    25. February 2018 at 10:33

    MFFA, No, I don’t know how to do that, but commenters should feel free to do so, and perhaps could cite this post if it would help.

    Jeff, Yes, that would clearly be far better. But it still wouldn’t really tell us who was the richest American of all time (in terms of real purchasing power), rather it would tell us who was richest relative to the average wealth of the day.

    It would still be a very unconventional definition, and as you say, Rockefeller still would not be number 1.

    John, Yes, but even then the winner would be the first American, not Rockefeller. It would also remove him from lists at the global level, instead some billionaire in Liechtenstein would probably come first.

    I think it’s important that people not abuse the language, and “richest” has a pretty clear meaning, which is not share of national income. If America’s population doubles, that doesn’t suddenly make me less rich than my grandfather, who might otherwise have had the same real wealth.

  15. Gravatar of Charlie Deist Charlie Deist
    26. February 2018 at 10:41

    Hi Scott,

    Interesting post – thanks for the link, and again, for taking the time to chat.

    Reading between the lines, I think the subtle dig at libertarian ideology is well-deserved.

    In asking “What would Milton say?” there’s probably more than a hint of motivated reasoning on my end. Partly, I wrote that title to appeal to a certain kind of libertarian audience, but regret it to the extent that it a) obscures the originality of your work and b) hinders new knowledge by cementing the old paradigms.

    My hope was actually to convert the three interviews – with you, Brad DeLong, and Robert Wenzel – into a short ebook, and write my own synthesis connecting the chapters. I lean Austrian because I think that monetarism (and even market monetarism) depends too much on ideal institutions to maintain credibly. You need a Benjamin Strong at the helm – an enlightened statesman who can calm fears of either extreme that we talk about in the interview.

    I almost wrote my thesis on the fact that Milton Friedman’s death functioned similarly to Strong’s death (both happening on the eve of the downturns). It would have been more of a study of human nature than the transmission mechanism of monetary policy – especially with respect to people’s tendency to copy a simple answer, coming from an authoritative voice. Maybe that explains the phenomenon you describe in this post. I think if Friedman had been around, he would have offered intellectual “cover” for conservatives and inflation hawks to go along with a program of temporarily higher inflation. Without that clear voice, Bernanke didn’t have the confidence to pursue something that would have upset the balance.

    Given that we don’t live in a world of philosopher kings, I think we may just be better off with a free banking system, where failures are at least localized rather than system-wide. But I’m getting off-point.

    As a final note, I think the challenge for us in or close to the market monetarist camp is to be able to explain these ideas to someone who hasn’t spent a few semesters trying to figure all of this stuff out. I’ll have the transcript shortly, and will try to spotlight some of the areas where I think the average person would lose the thread.


    P.S. The link now has the audio, as well.

  16. Gravatar of lxm lxm
    26. February 2018 at 12:17

    “And I think this explains much of modern progressivism and libertarianism. People want to believe X, so they latch on to all sorts of dubious data points that support the conclusions that they have already decided they would like to reach.”

    Why limit motivated reasoning to progressivism and libertarianism. I speak with many conservative who are pros at motivated reasoning.

    Why you might even try watching Fox News. I can’t watch it for long because I always break out laughing and my brother-in-law hates me for it.

    The problem is how to get around motivated reasoning and get real dialogue going. But the ruling elites don’t seem to want this.

  17. Gravatar of Frank Taussig Frank Taussig
    26. February 2018 at 14:54

    Hi folks,

    ¨When he died in 1937 he was worth about $1.4 billion, roughly $24 billion in 2018 dollars¨

    True…But! Veracity here hinges on the adverb ¨ roughly¨ in the above phrase. Start to unpack the roughness of this kind of estimate and soon realize that translation of past purchasing power into present values is MUCH harder than it looks.

    It’s a guessing game at best. A really fun game for econ nerds IMHO.

    UIC economists Samuel Williamson and Lawrence Officer tackle this problem at their great website Measuring Worth. It’s a labor of love, Highly recommended.

  18. Gravatar of Scott Sumner Scott Sumner
    26. February 2018 at 20:41

    Charlie, I had the same thought about Friedman’s death. Excellent comment.

    lxm, Fox News? I didn’t include it because I assume that none of my readers take it seriously. It’s become a joke. I take libertarianism and progressivism seriously.

    Frank, I agree.

  19. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    27. February 2018 at 08:10

    Why waste time and energy decrying the misinformation about something as trivial as the comparative wealth of John D Rockefeller? There are far more serious examples of ‘Everything you know about X is wrong.’ Consider;

    The homicide rate doubled from the early 1960s to the late
    1970s, increasing from 4.6 per 100,000 U.S. residents in 1962
    to 9.7 per 100,000 by 1979….

     In 1980 the rate peaked at 10.2 per 100,000 and subsequently fell
    to 7.9 per 100,000 in 1984.

     The rate rose again in the late 1980s and early 1990s to another
    peak in 1991 of 9.8 per 100,000.

     The homicide rate declined sharply from 9.3 homicides per
    100,000 in 1992 to 4.8 homicides per 100,000 in 2010.

    The number of homicides reached an all-time high of 24,703
    homicides in 1991 then fell rapidly to 15,522 homicides by

     The number of homicides increased steadily from the early 1950s
    until the mid-1970s….

     Between 1999 and 2008, the number of homicides remained
    relatively constant, ranging from a low of 15,552 homicides in
    1999 to a high of 17,030 homicides in 2006.

    These homicide numbers were still below those reported in the 1970s, when the number of reported homicides first rose above 20,000 (reaching 20,710 in 1974)

    How many news stories (or Wikipedia articles) mention any of the above government statistics when there’s a mass shooting like the one last week in Florida?

    I.e., the decline in the homicide rate is coincident with a massive INCREASE (about 50% in the past quarter century) in the number of firearms in private hands in the USA. I don’t know, since I don’t watch TV news shows, but if any news outlet has mentioned that Americans are less likely to be murdered by a firearm today, than 25 years ago, it probably was on Fox News.

  20. Gravatar of Carl Carl
    27. February 2018 at 09:09

    The Washington Post used the metric of the number of average incomes in the year and country where a person lived. They assume a 6% interest rate on assets to express everything as income. I’m not sure they take into account inflation.

    Rockefeller still beats Gates by this measure, but loses out to Khodorovsky and Slim.

  21. Gravatar of Carl Carl
    27. February 2018 at 09:09

    Forgot the link:

  22. Gravatar of Fred Fred
    27. February 2018 at 10:56

    Professor Sumner,

    Have you seen this recent paper from the 2018 US Monetary Policy Forum?

    I haven’t read much more than the abstract and some of the coverage in the financial press on it. The “popular” view seems to be that this paper doubts the impact of LSAP’s. From reading the abstract, it seems that the bulk of the demonstration in the paper was questioning the impact of LSAPs on long-term Treasuries, as if that’s the epicenter of transmission. If that’s the case, it seems to toss the entire paper into doubt: after all, I was taught that expansionary monetary policies can reduce, increase, or hold constant the level of interest rates. Absent some careful decomposition of the effects of QE on yields, this doesn’t seem very reliable.

    I’d love to see you write a post on this, if you’re interested.

  23. Gravatar of Nate Nate
    27. February 2018 at 13:59

    Fixed! You guys know anyone can edit wikipedia right?

  24. Gravatar of ssumner ssumner
    1. March 2018 at 09:48

    Carl, Wow, that WaPo story is nuts. They mix up purchasing power and relative income.

    Fred, I have a post at Econlog.

  25. Gravatar of ssumner ssumner
    1. March 2018 at 09:50

    Thanks Nate!

Leave a Reply