9

NYDIG published this statement:

Our findings suggest that 46MM+ Americans own Bitcoin today. That’s more than 22% of adults over the age of 18.

That figure sounds extraordinarily high to me. And NYDIG is not a disinterested party: NYDIG describes itself as "a leading technology and financial services firm dedicated to Bitcoin."

NYDIG claims that its findings came from the following "methodology:"

This poll was conducted by SurveyMonkey on Jan. 6-7, 2021 among a national sample of 2,184 US consumers. Respondents for this survey were selected from the nearly 3 million people who take surveys on the SurveyMonkey platform each day. Data have been weighted for age, race, sex, education, and geography using the Census Bureau's American Community Survey to reflect the demographic composition of the United States. This poll has a margin of error of +/- 2.1 percentage points

There are two ways of looking at this question:

  1. Is the NYDIG claim supported or contradicted by other data?
  2. If we accept NYDIG's methodology statement, are SurveyMonkey "US consumer" weighted survey results valid and accurate?
Lysander
  • 583
  • 1
  • 6
  • 16
  • 5
    How are the 3 million people chosen, that take surveys on SurveyMonkey? If these are self-selected (which I would suspect) than this massively biases their sample in all kinds of very hard to predict ways that can't easily be corrected by adjusting for age, race, etc. – quarague Jun 08 '21 at 08:48
  • 11
    People who own cryptocurrencies are much more likely to fill out surveys about cryptocurrencies. – gnasher729 Jun 08 '21 at 09:19
  • 3
    This is going to depend heavily on the validity of methodology. [This](https://www.buybitcoinworldwide.com/how-many-bitcoin-users/) article discusses several methods, each with their own pros and cons. To quote a snippet: *'the total number of people who own Bitcoin depends on how we want to define "own"'.* – Jordy Jun 08 '21 at 14:02
  • 4
    @gnasher729 in fact, owners of cryptocurrency are very likely to own a computing device, and non-owers of computing devices are very unlikely to answer an online survey – Hagen von Eitzen Jun 08 '21 at 14:17
  • There's perhaps some technical questions on what it means to own cryptocurrency is well. There's several philosophies on what it means to actually own cryptocurrency rather than just hold a stake in it in some way. –  Jun 08 '21 at 18:47
  • Back in 2017, there were an estimated 2,9M to 5,8M bitcoin users *worldwide* according to this Cambridge study : https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2965436 . That figure must have grown since, but I agree with your intuition that 46MM+ Americans btc owners today seems exaggerated (by one order of magnitude, I would guess). – Evargalo Jun 09 '21 at 07:40
  • 1
    This (not neutral) website tries to estimate how many ppl own bitcoin: https://www.buybitcoinworldwide.com/how-many-bitcoin-users/ TL;DR : it's complicated and noone really knows. – Evargalo Jun 09 '21 at 07:47
  • How would anyone survey how many users of a pseudonymous system who have strong reasons to hide their identity and ownership? You might be able to put a lower bound on "how many US persons have coin exchange accounts", since those are subject to ID registration. – pjc50 Jun 11 '21 at 10:04

1 Answers1

6

SurveyMonkey describes how they select participants:

SurveyMonkey Contribute panelists take surveys for charity and a chance to win a sweepstakes prize.

Rewards panelists earn credits for completing surveys which they can redeem for gift cards or donate to charity.

All panelists share demographic info about themselves like gender, age, and region, and other targeting attributes you might be interested in, like cell phone usage or job type.

We balance Contribute and Rewards panels according to census data of age and gender.

It seems that the people who fill out SurveyMonkey surveys to earn gift cards are a demographic that's interested in "earning money online".

Basically, the audience is people who do things online to get rewards. For most question you can ask, you can likely get a representative sample by controlling for demographic factors.

Investing in Bitcoin however might be in the same class of "doing things online to get rewards" and thus the sample has a systematic bias.

A survey that would call people on the phone wouldn't have that bias and therefore likely provide more accurate answers.

Nate Eldredge
  • 9,275
  • 5
  • 46
  • 48
Christian
  • 33,271
  • 15
  • 112
  • 266
  • 2
    This is neither a useful comment, nor a useful answer. It does not answer the question "Do more than 22% of American adults own Bitcoin?" – Jamiec Jun 10 '21 at 07:59
  • 2
    @Jamiec It's quite unclear why my comment was deleted. Deleting my comment and writing something related to it is really strange moderation behavior. – Christian Jun 10 '21 at 08:45
  • I did not delete your comment, and wasnt even aware you had written one – Jamiec Jun 10 '21 at 08:46
  • I have however just looked back in the history and it said "I do think it's useful in the context of the question so it's better posted then not posted" - which is just wrong. this is a Q/A site. You either answer the question, or you don't. This is not an answer to the question, it would have been better to not post at all. – Jamiec Jun 10 '21 at 08:47
  • 2
    @Jamiec : The question species "two ways of looking at it" my answer answers one of them. – Christian Jun 10 '21 at 13:46
  • @Jamiec Read the entire question, not just the title. – barbecue Jun 10 '21 at 15:52
  • I think the passage you quoted is in reference to "SurveyMonkey Contribute" panelists. This appears to be only a particular subset of SurveyMonkey users, and I don't see any indication that the study in question was targeted to that particular subset. – Nate Eldredge Jun 10 '21 at 19:37