The Truth About Witness Retention Rates

While browsing around the internet for information on Jehovah's Witnesses, a person is very likely to stumble across a statement like the following:

"Jehovah's Witnesses have the lowest retention rate of any religious group! Two out of every three Witnesses jump ship! They are leaving in droves!"

But what is the real story behind those figures? Do they accurately represent the true retention situation among Witnesses? This post will attempt to provide a brief answer.
Do note though, that this author doesn't claim to be a statistician, and he doesn't claim to have all of the overall growth variables included. This post will merely provide a very simple look at the numbers and at the major issue that causes confusion, i.e., the ignoring of the convert retention rate.

The primary sources of information for Witness retention rate data have historically been Pew surveys, specifically their U.S. Religious Landscape Surveys.[1] Notice the important part of that title though: "U.S." Most of the data we have on Witness retention rates (including the data used in this post) is for the United States only. Retention rates in other parts of the world (especially Latin American countries) could be significantly higher than the U.S. rates.

The Pew study linked to in the references gives retention rate data primarily on those who have been raised around their current religion, but does not give retention rate information on converts. This is an important distinction to make, because some religions (like the Witnesses) are more convert heavy than others. Even a very low retention rate among life long adherents could be statistically insignificant if life long adherents are not well represented in the overall group.

Another important factor to keep in mind is that many Pew surveys tend to have very low relative numbers of actual Witness participants. In the 2008 U.S. Religious Survey referenced below, there were only 200 or so Witness participants out of an overall participation of around 35,000. Thus, while it can accurately tell us the ratio of Witnesses to the general population, it might not be such a reliable source of information for more specific things, such as retention rates.

While keeping the possible sources of error in mind, we can still determine the makeup of that small, 200 or so person sample and draw some conclusions from it. By knowing the general makeup of the group, we can essentially 'weight' the retention numbers based on their overall significance and then find the compound retention rate. The religious history of current Witnesses was as follows in the Pew survey:

Total Group Size Estimate: 35,000
Childhood Witness: 0.6%; 35,000 x 0.006 = 210
Entering Percent: 0.5%; 35,000 x 0.005 = 175
Leaving Percent: 0.4%; 35,000 x 0.004 = 140
Current Witness: 0.7%; 35,000 x 0.007 = 245

210 Childhood JW

-

140 Leavers from Childhood JW

---------------

70 Raised JW and Stayed JW

+

175 Joiners From Other Groups

---------------

245 Current JW

70 / 210 = 33% retention among those raised JW.

 The Pew people themselves ended up with 37% retention for born-ins, so this author differs from them by about 4%.

Now those "Joiners" from other groups are assumed to be the converts who have already gone beyond the 'decision' stage, and who decided to stay with the JWs. There would also naturally be a certain number of converts from other religions who chose to not stay with the JWs, and who would technically not be represented at all in the Pew data since it cannot measure more than one religious change, e.g., "from Protestant to JW back to Protestant" would not register the JW stage in the Pew survey. This unrepresented group would increase the total number of converts in general to a number higher than 175, but exactly how much higher depends on the retention rate of converts.

What is the retention rate of converts? The Pew survey does not allow us to determine that number; however, many NRM researchers and those who study the Witness Year Book statistics (such as Cumorah farther below), would approximate the convert retention at around 55-75%. Using the high and low ends, we can guesstimate the total number of JW converts in the representative group of 35,000.

175 / 0.55 = 318 on the low convert retention side.
175 / 0.75 = 233 on the high convert retention side.

70 + 175 = 245 total retained JWs
210 candidate born-in JWs + x candidate convert JWs = y total candidate JWs
Compound Retention Rate  = 245 / y

210 + 318 = 528 total candidates
210 + 233 = 443 total candidates

245 / 528 = 46%
245 / 443 = 55%

Compound Retention  = 46%-55% or around 50% averaged out.

This convert heavy structure of the group also explains why the Witnesses have not just completely fallen apart and disappeared from the face of the earth (as you would expect any group to do that only retains 33%-37% of members). They are not losing 63-66% of their TOTAL membership, they are losing 63-66% of a smaller subgroup within the total number of Witnesses (~43% of candidates, i.e., those who still have the choice of either staying or leaving, are born-ins while ~57% of candidates are converts), which is spread out over several years (because it is not technically an annual loss). Factor in a much larger retention rate for those converts (around 60% or higher), and it becomes apparent why the Witnesses are still growing at a faster pace than most of the other religious denominations even with low born-in retention rates: the high convert numbers compared to born-ins and the high convert retention more than makes up for the losses coming from those born-in drop-outs.

Also factor in death rates, birth rates, new baptism rates, the rate of convert loss over time vs. the rate of born-in losses over time (which are probably different), the reinstatement rate of DF'd individuals over time, ect., and eventually you should end up at the 2-3% overall annual growth figure. But this author isn't going to explore all of that in any depth.

The research site Cumorah.com[1] also had some interesting comments on the Pew studies in relation to LDS retention rates:

The Jehovah's Witnesses, on the other hand, reported a net gain, with converts outpacing the children of members lost to disaffiliation. To put it another way, 71% of individuals presently identifying themselves as Jehovah's Witnesses were converts, compared to just 23.5% of Latter-day Saints...The Pew survey documents that the LDS Church is significantly more successful than Jehovah's Witnesses at retaining (at least in the sense of retaining self-identified religious preference) individuals born in the faith, but is much less successful at retaining baptized converts... In contrast, the Jehovah's Witness proselyting program has been substantially more effective when assessed not from the angle of raw baptismal statistics, but from the perspective of making and retaining adult converts who continue to affiliate with the faith and constitute the bulk of its committed membership. The Witnesses' own annual statistics consistently demonstrate retention of 55-65% of baptized converts worldwide, more than double LDS convert retention rates.

Conclusions

• The retention rate among those raised as Witnesses was estimated by this author to be around 33% and by Pew to be around 37% (4% difference).
• Those retention rate numbers derived from the Pew survey are not very reliable considering the small sample size of Witnesses.
• The retention rate for Witness converts is generally agreed to be much higher (anywhere between 55% and 75% depending on the source).
• The compound retention rate for both converts and born-ins was estimated by this author to be around 50%.

Possible Sources of Error

1. As was already mentioned, the sample size of Witnesses used for the survey was very small compared to the size of the overall representative group, so small in fact that the percentages given for the "Leaving," "Entering," "Childhood Religion," and "Current Religion" in relation to the overall sample only contained a single significant figure. If instead of 0.6% "Childhood Religion," they had "0.64%," then there would have been a sizable different in the final calculated retention rates (and indeed, those significant figure issues may have been what caused the 4% difference between the author's calculated childhood retention rate and Pew's calculated rate). And that few hundreths of a percent error compounded over four different categories could make it even worse.
2. This author also rounded the total group size to 35,000 for the sake of convenience. But because the calculations were all founded on percents instead of whole numbers, the rounding shouldn't have produced significant error.
3. The assumed convert retention rate may have caused some error, but it was intentionally left as a range in order to average out that error.
4. As was mentioned near the start, the data was taken from U.S. residents only, even though Witnesses reside in nearly every country on earth. The U.S. Witnesses represent maybe 1/7th of the global Witness population, and likely have one of the lower retention rates compared with other countries (especially Latin American ones).

--------------------------------------------------------------------------

References

[1] http://religions.pewforum.org/pdf/report-religious-landscape-study-full.pdf
[2] http://www.cumorah.com/index.php?target=church_growth_articles&story_id=14