Bitcoin and other digital currencies are emerging technologies with limited transparency and as such there are little hard data on who uses Bitcoin and why. What limited surveys typically reveal is that Bitcoin users are generally male, relatively young and reside in the US. However, they may be just one subset of the digital currency community.  

Recent dramatic Bitcoin price increases suggest that the user base for digital currency is much broader than these nonrandom, online surveys would suggest. Another approach to better defining the Bitcoin user is to look at the digital currency ATM market.

The number of Bitcoin ATMs globally, but particularly in the US, have skyrocketed in 2017 alone. Analyzing the locations of the oldest Bitcoin ATMs in select US cities should give clues about the demographics ATM operators are trying to reach.

Even with all of the limitations, the ATM data still reveal key information about a Bitcoin ATM user targeted by the owners of the ATMs. Targeted users appear to be from all income levels and neighborhoods.   

Recent surveys

Existing user surveys are briefly summarized and then compared to an analysis of the demographics of a sample of US Bitcoin ATM locations.  

A 2015 survey of almost four thousand responses found that almost 60 percent of respondents were under the age of 35, 90 percent of the responses identified themselves as male, 49.85 percent of survey respondents said that they were located in North America and 65.8 percent stated that they were Caucasian.  

Users were evenly distributed across income categories. There are problems with this analysis. This is not a random survey, but a survey of people that voluntarily chose to participate and predominantly accessed the survey’s main site and social media platforms. As such, this survey is linked to North American and European users and is not an unbiased survey.

Another online survey of Bitcoin users collected 1,193 responses in 2013.  This survey reveals results that appear to be very similar to that of the previous survey with the exception that this poll finds that 96 percent of Bitcoin users are male.  Like the previous survey, these results are based on self-selection making it biased as well.

To get away from the North American and US bias of these surveys, bitcoinsinireland has conducted annual online surveys in a variety of languages.  The 2016 survey used 13 languages, had 771 respondents and again found these users to be younger and male.  Over 1/6 of respondents had lost Bitcoin due to a hack or shutdown.

It is likely that these online, volunteer surveys may be finding only one sector of the Bitcoin user demographic, a computer savvy, young, male adopter of Bitcoin. The Bitcoin community may be broader than these online surveys suggest given the rapid growth in market value and adoption of the technology.  

Another rapidly growing sector of the Bitcoin technology is the digital currency ATM. We do not know much about these users, but we do know the locations of the ATMs and can make some general inferences about who the ATM operators may be trying to reach via their location decisions.

ATM Data

Digital currency ATM locations were collected from coinatmradar.com which has mapped ATM locations and status since 2014. There are currently 1,485 Bitcoin ATMs on their website, with 964 ATMs located in the United States, with most of them installed in 2017. In addition to mapping locations, coinatmradar provides a list of major cities in the US with ATM locations.

Assuming that ATM operators are trying to be profitable and will relocate their machine investment if a specific location is not generating usage, a subset of the ATMs in major cities was selected with installation dates of 2014 and 2015 that appear to still be in operation.

In a few cities without installations in those years, the first installation in 2016 was selected. Using this method, 104 ATMs were identified in major cities around the US that have been in operation for up to three years.  

The next step was to link the ATM geographical location with Census tract information to get a better idea of the demographics of the locations chosen by the ATM operators to reach their targeted customers  ATM location addresses were matched with Census tract neighborhood data (2,500-8,000 people) for income data categories of  upper, middle, moderate and low. Census creates these categories based on comparing the Estimated Median Family Income for the Census tract with the broader metropolitan area data.

The results are displayed in the chart below. Of the 104 ATMs in operation as early as 2014, 25 percent were in upper-income neighborhoods, 33.7 percent were in middle-income neighborhoods, 26.9 percent in moderate income areas, and 14.4 percent in low-income tracts.  

The installations were located in high traffic venues, including gas stations minimarts, delis, pizza parlors, laundromats, shopping malls, and tourist locations. This corresponds generally to a 2015 analysis by coinatmradar which listed the top five Bitcoin ATM location types as:

  • Restaurant/pizzeria 12%

  • Cafe/coffee shop/gelateria – 9%

  • BTC embassy / company office – 7%

  • Shopping mall – 7%

  • Supermarket / food store / Deli – 6%

What does this mean?

Currently, US  Bitcoin ATMs are distributed across the income spectrum with roughly 40 percent of all the Bitcoin ATMs used in this survey located in tract areas that according to the US Census are below the median family income level.  

Over 14 percent of the ATMs are located in areas where the median family income averages $26,000. It can be assumed that the reason that these ATMs were placed in those areas was because the operators who put them there thought they would make a profit. This seems to contradict the notion that Bitcoin users tend to be wealthier than most.  

There are several possible explanations as to why the ATMs would be located in lower income areas. One reason could be that residents living below the median family income level are less likely to open bank accounts and more likely just to use cash, making the ATM a convenient choice. Residents may also find that using Bitcoin preferable to an online exchange as it offers the appearance of anonymity, if not the reality.  

Another reason could be that a percentage of the population living in the below median family income level area are, for various reasons, disproportionately interested in transferring remittances to others in other geographic locations.  In this case, the sender buys Bitcoins for cash at an ATM location, sends the Bitcoin to the recipient, who then sells the Bitcoin for cash at a receiving location or buys goods directly. This could be a simple and in some cases less expensive remittance option.  

A broader question must also be asked and answered. Why set up Bitcoin ATMs in any income area when it is possible to use online exchanges?  

One reason could be that transactions at a Bitcoin ATM are quick and require less documentation compared to online exchanges. Another reason could be due to lack of trust in the exchanges.  

The 2016 bitcoinsinireland survey found a substantial number of users had experienced Bitcoin loss through hacking. A Bitcoin ATM could allow a person to buy Bitcoins and then just as quickly store it in a safer offline wallet. Again, a feeling of anonymity when using a Bitcoin ATM may exist, even though federal financial regulations applying to ATM operators may not justify this feeling.  

A beginning Bitcoin user may also feel less intimidated by an ATM than establishing an online account with an exchange.

Limitations of the ATM Data

Using the location of Bitcoin ATMs to extrapolate data on the average Bitcoin user has its own problems.  

The assumption here is that Bitcoin ATMs are placed in neighborhoods where the residents are poised to take advantage of the ATM.

The flaw in that line of reasoning is that in some cases, such as tourist areas, many of the Bitcoin ATM users are not going to be from that neighborhood, but somewhere else.  

Another potential problem is that if a Bitcoin ATM fails to make a profit, then the operators will either move the ATM to a better neighborhood or shut it down entirely.  As a result, the current analysis is accurate now, but that could change quickly.  

This analysis used the oldest Bitcoin ATMs found in the US in order to mitigate this problem, but it hasn’t been eliminated entirely. The final issue is that the coinatmradar data is voluntarily submitted and might not track every Bitcoin ATM in the US.

The coinatmradar data is currently the best data available about Bitcoin ATMs, but limitations are inevitable.

- By Jennifer L.  Moffitt

Jennifer L. Moffitt (jennifermoffittlaw.com) is a privately practicing attorney located in Cheyenne, Wyoming specializing in regulation and compliance with respect to digital currency. She has a Bachelor’s Degree in Economics from the University of California, Berkeley, a Master in International Studies from the University of Otago in New Zealand, and a J. D. from the University of San Diego.