AmateurRadio.com

No, not really. Here’s actual evidence…

In a previous article, I quoted current ARRL CEO David Minster NA2AA from one of his monthly Second Century columns in QST. I’ll reproduce it below:

“We know, and have known for years, that ARRL members represent the lion’s share of active hams. Moves to grow the hobby…have grown the total number of licensees, but not the number of radio-active hams.”
CEO David Minster NA2AA, QST (March 2023: 9)

Let’s take Mr. Minster at his word at first but also look his words carefully.

“Lion’s Share.” What does that mean? We hear hyperbole quite a bit in common parlance these days. Example: GOAT is the greatest of all time. (No one even mention Michael Jordan or Lebron James, ok?) In common usage, it is more glorified smack-talk than a concept that is defined, much less measured for systematic comparisons. (Apologies to the sports statistician students from my courses over the years.) The Chief Executive Officer’s monthly column in QST should be taken more seriously than water cooler talk, no? This is certainly the case since the ARRL has tax-exempt status from the IRS, whose guidelines say they should be truthful and transparent in statements to their donors (and members). As the Chief Executive Officer, his public words matter for they speak for the company.

Here’s what this phrase means:

If a person, group, or project gets the lion’s share of something, they get the largest part of it, leaving very little for other people. Collins English Dictionary

This means he is saying the vast majority of “active” hams are League members. The converse is that only a minor, irrelevant portion of “non-active” hams are members of the ARRL. It’s a rhetorical term without a numerical meaning. But for rhetoric to be successful in an argument, it must have the weight of convincibility which is largely evidence-based. So vast majority means much more than 51 percent which is a simple majority. Another argument is “The greatest part of something, to the point where alternatives are nearly irrelevant in size.” We will just proceed with Minster’s usage as reflecting some proportion which leaves the remainder as irrelevant in size.

Now, let’s wrestle with the oft-used phrase, “active ham,” to which Minster prefaces with radio, or “radio-active.” Unless we have some clear understanding of what this word means as well, then Minister’s claim can be whatever he wants it to be. Hams use it with their own, mostly unspoken, meaning. For example, a nearby club that I’ve been working with to regain their nearly dormant footing in Vicksburg has claimed the slogan of “VARC: Radio Active on the Mississippi River.” (They actually have the state’s only nuclear power plant just south of them in Port Gibson, MS but that’s not what the slogan refers to.) A dictionary definition references acting or capable of acting or currently in operation, in effect, in progress and so forth. It is one or more behaviors, for sure. But which ones and how much does it take to be considered “active” in the hobby? Every ham would have an opinion, of course, but we need systematic observation of amateurs’ behavior to better determine how to conceptualize activity in the hobby.

Active is “acting or capable of acting or currently in operation, in effect, in progress and so forth.” Collins English Dictionary

For the CEO of the ARRL, it means that a ham who engages in one or more of these unspecified amateur radio activities has a paid membership in the League (including Life Members like me). As I pointed out in a previous article, CEO Minster does not show any data, cite any studies (be careful of just citing a study as you might not actually have one), or go beyond the rhetoric of “we know in Newington” thus and so. Do we just take him at his word since HQ knows stuff without the transparency of how they know the stuff the CEO says they do?

Fortunately, we don’t have to. Our neighbor to the north, the Radio Amateurs of Canada, conducted a very detailed national survey prepared by Paul Coverdell VE3ICY with data collected in 2021. It covered all hams in Canada, including both RAC members and non-members. In fact, about 21 percent of all respondents were NOT members of RAC. When RAC President Phil McBryde VA3QR asked me if I would analyze the data for them, I did with the full report here. Paul VE3ICY included over 30 specific amateur radio activities as well as the time spent on them. Unless one thinks that ham radio operators in Canada are fundamentally and exceptionally different from those in the U.S., these are by far the best data on specific measures of ham operator behaviors. (Ok, Canadian hams do appear to be nicer, on average, but this is just my experience!)

Conceptualizing active ham operators

The scientific problem is that there is no recognized concept regarding being an active ham. Imagine trying to measure the impedance of any antenna without the concept? Couldn’t you see the morass of Youtube videos with a myriad of “meter measurements” that go all over the place except the places where the concept of impedance should take them? Sounds silly at first to those not well-read in the philosophy of science but how should we conceptualize ham activity?

Minster writes and speaks as if it is a light switch: either on (active) or off (inactive). Does it make sense to think of active or inactive hams only that way? If we asked two hams if they are “active,” each might say yes. Delving further, the first one might just be an ARRL member and really read QST every month. The other might be an ardent EmComm operator, participating in several weekly Nets and activating for emergencies with a local ARES team. And be an ARRL member. Are both “active”? Well, they say they are because each refers to a different standard of reference in doing so. It seems much more theoretically beneficial to conceptualize ham operator activity as a dimmer switch, or on a continuum rather than an on/off dichotomy, especially since the “off” of absolutely no activity has no internal variation while the “on” of some activity can range dramatically. I explore this distinction with these national survey data. The reader will see that there is quite a bit of variation within the “on” part of Minster’s dichotomy!

In my report to RAC on their 2021 national survey of Canadian hams, I analyzed the 32 specific operating behaviors. (Each is listed there for your reading so I won’t reproduce them here.) It’s clear that activity as directly measured by asking hams themselves if they do specific things varies quite a bit. I’ve excerpted Figure 2 from the report below. The average number of activities in a given month is about nine (9). These range from one (1) to twenty-nine (29). Quite a bit of variation. A dichotomy conceptualization of activity would not characterize these observations very well. Not a lightbulb turned off or on but a dimmer that varies greatly and shines rather bright if you’re doing a couple of dozen activities during a month!

It’s important to recognize the potential “sample selection bias” that truncates the distribution on the left (i.e., the truly inactive hams, those that would score zero here, may not be adequately represented in the survey). I’m not sure that this is different for non-members and members but we simply do not know. We will see if there are any differences in those in the survey who are not very active below. Given that these are the very first national sample data on specific operator behaviors, the results make this question pale by comparison to what we do learn from them.

In the right panel, box-and-whisker plots visualize how each age varies in total number of activities. The median number of activities (illustrated with the black line in the box itself) does not diminish that much with age. The small set of teens in the sample to make that group pop up in number but the overall distribution shows that middle age hams have small segments who really, really get involved in a diverse set of activities.

To emphasize the key here, the notion of being an “active” ham is best thought of as a continuum from an external, objective viewpoint. This allows for us to see how amateurs themselves say they behave instead of us individually recounting just what we see on a daily basis. This externally objective conceptualization is a canon of science. Water cooler hyperbole per se is not very useful to understand things from a scientific perspective.

Is Measured Ham Activity Related to National Association Membership?

If being a member of RAC, and by inference, ARRL, contains the lion’s share or “vast majority” of active hams, then this distribution of the total number of activities, and the time they spend on them, would show marked differences for members and non-members. It might show that measured years of self-defined “being active” would also be different, too. Wouldn’t “radio active” hams know they are active since we now have specific measurements on that activity? Let’s see. Here is the total number of over thirty specific activities reported by hams in Canada, separated by RAC membership. The box-and-whisker plots illustrate the distribution of each group. I’ve annotated the display for readers unfamiliar with this data visualization technique.

There is less than one average activity difference in favor of RAC members in comparison to non-members (means of 9.55 vs 8.81). This difference is likely to be non-zero (p = .006) in the population but substantively it is very small. The medians, shown in the boxes as black lines, are exactly one activity different (9.0 vs 8.0). The percentage of highly active and very lowly active hams are very similar for both RAC members and non-members. Finally, the general shape of the two distributions of amateur radio activity is almost identical. The capstone comment is that imagine all of the differences among Canadian hams in this survey in measured operator activities (see previous histogram). Of all these differences in activity, only 0.6 percent is due to membership in their national society! Less than one percent.

The total number of activities is one aspect of ham operator behavior. Another his how much time do they spend on these activities? Surely, an operator spending much more time on an activity is more “active” than one who spends less, right? Combine these for all activities since some hams are more pluralistic in their enjoyment of the hobby than others.

I’ll compare those data for members and non-members now in the boxplots below. Notice in the top panel that the raw numbers of total hours are bunched around the medians for each group. But the median hours spent in ham radio activities are actually higher for non-members (41.0) than members (35.5)! This equates to about 1.4 hours and 1.2 hours, respectively, in a 30-day month. Recall that the median (50th percentile) has one-half of the respondents below and above that point. The mean score will be influenced by very high or low scores. Each group has them (shown as the symbol * in the boxplot). Perhaps League members are more highly active hams as Mr. Minster alluded to in his QST editorial column. Not really, as the means for members (93.26) and non-members (120.19) still show that Canadian hams not in RAC report higher levels of activity than do members.

Because of the skewed distribution in the raw number of hours per month (small share of highly active hams in each group), the lower panel uses a log form of the number of hours reported. This gives the reader a clearer visualization of just how similar the two groups are in hours of activity spent each month. They are almost identical! To make the visual point in numerical terms, of all of the differences in hours of activity, only 0.2 percent is associated with RAC membership. Almost zero.

In the full Report, I combined some of these 32 specific activities into meaningfully similar indexes, such as EmComm, Competition (contesting), and so forth. Perhaps it is only among some activities that national association members gain this “lion’s share” of active hams. Let us see briefly.

In the set of six panels above, I’ve assembled comparative boxplots by RAC membership. Only one shows any differences for members and non-members: competition, such as in contests and DXing. None of the other activity indexes are any different by membership status.

Individual activity comparisons by RAC membership are detailed in the final report but do are not inconsistent with the stark lack of differences by membership status shown here (i.e., not favoring either one). Perhaps it is the contesting and DXing segment of ARRL membership that is most on CEO Minster’s mind when he asserts that the Lion’s share of “radio-active” hams are League members. That would be the only finding here that is in line with that viewpoint. I try to explain why this might have some influence on him in the final section.

Is the Scope of the Amateur Radio Career Where Membership Makes a Difference?

To give CEO Minster’s assertions every chance to be validated, perhaps his statement may be accurate over a longer duration of time. Like nearly all such surveys of behavior, the scope of time being referenced is a recent month. Would national association members be licensed longer or have a greater share of the length of their license tenure being “active” as they defined it? Would the temporal duration during the ham radio operator’s career better fit the “Lion’s share” description by CEO Minster? Fortunately, we can examine the RAC Survey 2021 national dataset directly.

As described in the Report, I have taken the measurements for years licensed (or tenure) and self-defined years “active” to make the following comparisons for national association membership. I’ve taken the difference (i.e., License tenure – years active) to compute a percent of the length holding an amateur radio license has been spent being active. I’ll call this the percent of active years since obtaining a license. It can range from zero to 100 with two-thirds reporting one hundred percent. This metric gets at the career aspect of being “radio active” rather than just a recent snapshot of ham activity. This variable cannot be more than 100 percent so it is truncated on the higher end. The two boxplots below demonstrate that there are virtually no differences in career active status for national association members and non-members.

Another nuance to length of license tenure and self-defined active status career is illustrated in the scatterplot below. This is simply the years active plotted by length that a license has been held. RAC membership status is denoted with different-colored circles as shown in the legend. Not surprising is the positive relationship: the longer the license tenure, the longer the ham reports being active. But note also that this is not nearly as high as many hams like to think it is. It’s the same regardless of RAC membership status (compare the two R-squared coefficients for similarity: .146 vs .144). The clustering of small circles at the zero point illustrate hams that have continuously been active. Note that the left axis means more years of inactivity (tenure – active). The more one is from the vertical axis reflects less activity since licensed. How different are RAC members from non-members in whether they’ve been continuously active? I’ll examine that next.

Using the data on “100 percent active” or not and membership status, the following crosstabulation answers this question. There is at best a three percent difference in “always active” status. For a survey sample, this is not a statistically significant difference (p = .303, ns) and thus the only observed differences are simply random fluctuations.

The time scope of the ham career also does not reflect any statistical differences, much less a Lion’s share.

What Can We Conclude About the Lion’s Share Thesis?

The recent statement by ARRL CEO David Minster NA2AA in his monthly QST column said that the Lion’s share of “radio-active” hams are indeed members of the League. This means that those hams who are not League members are a small, irrelevant segment of those hams who are “active” on the bands. He offered no data or other reference to this statement, simply that it is well-known, for some time, in Newington. As the CEO of a tax-exempt corporation, his public statements have greater weight than mere idle chatter, like that the reader may encounter at a hamfest. Indeed, IRS official guidance requires tax-exempt charities who solicit donations to be both truthful and transparent in their public and official statements to their constituents.

This led to my analysis of RAC 2021 Survey data on over 30 specific ham activity behavior to determine if there are any noteworthy differences between RAC members and non-members in this national survey. I demonstrated that the notion of being an “active” ham is best thought of as a continuum from an external, objective viewpoint. It is not useful to think of amateur radio opeartor activity as simply being a dichotomy like a light switch. Rather, the data show that there is a very variable continuum of activity.

I found almost no evidence supporting CEO Minster’s statement in his QST Second Century column. In this analysis, I tried to give every opportunity to find any area that fit Mr. Minster’s stated “Lion’s share” thesis. To cut to the chase: The single result that reaches statistical significance in favor of members was very small in practical value, less than one percent difference owing to RAC membership. It was one more activity per month that RAC members say they participated in versus non-members. The time spent in those activities was actually higher for non-members. The shapes of each respective distribution, even for the total number of activities per month, were nearly identical, further evidence of the similarity of members and non-members. All other metrics showed virtually no distinction between national association membership and non-members in Canada.

Specific areas of activity were also identical except one: Contesting and DXing where RAC members were demonstrably more active than non-members. Remember, RAC was born out of being a Section in the ARRL and modeled their governance organization in a parallel manner. If we assume that the US amateurs are generally similar to Canada, could this be what is shaping the CEO’s view from Newington?

It’s often heard that the “Contest Mafia” rules Newington by donations, serving on Committees, and becoming Directors. I’m not sure how empirically true this is but several prominent contesters have said this to me, sometimes in jest, but the idea is a real thing at minimum. Is Mr. Minster’s worldview of amateur radio in the U.S. heavily shaped by “radio-active” contesters and DXers who have higher influence in Newington? He has identified himself as avid contest participant (see QST Second Century column, November 2021: 9) and, rumor has it from some involved, enjoys organizing contesting trips outside the country with some Board members, Officers, and staff members. Thus, identifying closely with a smaller group might explain his statement and it presumes the same membership effect would hold in U.S. data. Clearly, from his column in September 2021’s QST, he has been influenced directly by a few power brokers:

Nearly a year ago now, I received strong feedback from a few important members; these were members who play an important role in our community or who were large donors. The messaging was consistent and clear: they were not happy with where the governance of ARRL had strayed. They weren’t asking me, they were telling me: I had to make it one of my top priorities. And I have.

A casual perusal of the call signs that place in the top echelons of various contests and highest on the DXCC Awards lists will also show many who are donors to the League, attaining specific mention in the Maxim Society: “ARRL is fortunate to recognize a group of individuals whose extraordinary generosity continues to support the organization at leadership level.”  It may be a stretch to make this connection but it is the only membership effect I found in all of these comparisons of activity by membership status.

Is another explanation of Mr. Minster’s “Lion’s share” thesis just a reflection of how he manages an organization? Remember, he says he comes from a lengthy career in the jewelry industry. Things are highly inflated (average mark-up for retail is 250-300%) and made to look bright-and-shiny in jewelry stores. Ever wonder why lights are bright and spotted toward the display cases? This is so as to make diamonds, gems, and other jewelry sparkle. Perhaps the CEO mis-spoke because of his career history in jewelry where bright-and-shiny promotions are the norm?

I have no idea if the Contest Mafia connection or the jewelry management background has anything to do with how far off the CEO’s statement is from recent (2021) national survey data from Canada. I’m stretching here to make sense of how wrong his column could be on this, unless they have convincing national U.S. data that they will not release to the public. And that may be at odds to the IRS guidance to tax-exempt non-profits since it intentionally lacks transparency to members.

Whether it is based on social influence from being a contester, politically influenced by powerful donors, or just a Machiavellian strategy to ignore the half-million licensed amateurs in the U.S. as not relevant to the existing League’s interests, it is just incorrect with respect to the only data on the subject that is in the public sphere.

A reader would have to argue that hams in Canada are demonstrably and fundamentally distinct from hams in the U.S. to wholly reject these results. Is there any evidence to that effect? Well, when the National Association for Amateur Radio, the ARRL, does not release data they say they have, and upon which they seem to make factual claims, we have no counter-factuals to this analysis based upon representative data from Canada. So, no there isn’t.

There is no other way to put this, based upon CEO Minster’s original statement and the results of this thorough analysis of national data on Canadian hams. If this is a “Lion’s share,” then it is a very small kitten indeed.