Abstract
Religious congregations hold a unique status among public charities in that they are exempt from having to ask for recognition of their tax-exempt status from the federal government. Despite this privileged status, more than two hundred thousand congregations are registered with the Internal Revenue Service (IRS), and the rate of registration appears to be increasing. We examine regional, religious, and administrative patterns among the congregations registered with the IRS. Certain denominations, such as the Latter-day Saints, appear to have taken a centralized stance against congregational registration, whereas others, such as the Lutheran Church–Missouri Synod, have made a centralized decision for congregational registration. Other denominations appear to have left the decision up to individual congregations. Of those congregations that are registered, the method of registration, whether independent or through a denomination’s group exemption, appears to be related to the denomination’s centralization of authority.
For tax purposes, religious congregations in the United States are considered a public charity (i.e., 501(c)(3)) by the federal government. However, congregations are unique because they are considered tax-exempt without registering with the Internal Revenue Service (IRS). 1 Any other public charity, including religion-related charities that are not congregations, must apply for and receive recognition before being considered tax-exempt. A congregation, however, is given de facto exemption without even asking for it. 2 Despite this privileged position in tax law and regulations, a large number of congregations appear to have registered with the IRS.
The fact that so many congregations are registering with the IRS either independently or through a denomination’s group exemption presents a puzzle. Why would any organization subject itself to more paperwork, state entanglement, and potential regulation than is necessary? There are a number of channeling mechanisms shown to be important in the decision to register (Cress, 1997; McCarthy, Britt, & Wolfson, 1991). These include increased ease in gaining local and state exemptions, postal discounts, grants and donations, and other benefits. Technically, a congregation should be able to receive many of these benefits without a formal recognition letter from the IRS. Such a letter, however, could ease the process, especially for newer congregations without much existing history in a community or congregations that are part of a religious minority.
In this research, we explore patterns of registration rates among religious congregations, discuss the potential mechanisms underlying those patterns, and evaluate the potential strengths and weaknesses of the IRS registry for future research. Our primary source of data is the IRS’s Business Master Files (BMF). The BMF lists every public charity registered with the IRS. It is updated by the IRS as organizations register and disband. 3 Whereas the IRS only makes the most recent BMF available on its website, the National Center for Charitable Statistics (NCCS) at the Urban Institute has been archiving the BMF for over the past decade. The NCCS typically archives the BMF once or twice a year. 4 We primarily utilize the BMF version that was archived in November of 2010. Each BMF contains basic information about each registered organization, including its name, address, and its foundation type. The latter includes a category for those organizations that claim to be a “church or a convention or association of churches.” We filter cases using this variable for many of the analyses below.
How Many Congregations Are Registered?
Specifying how many active religious congregations have registered with the IRS is more difficult than it might at first appear. The most obvious way would seem to be to directly look at the IRS database. We can, in fact, generate a precise number of nonprofits claiming to be a congregation in this database. As seen in Figure 1, the number of registered nonprofits claiming to be a religious congregation has tripled in the past 20 years from just more than 70,000 in 1995 to more than 215,000 in 2013. These numbers, however, might not be entirely accurate for a number of reasons.
Number of nonprofit organizations registered with the IRS claiming to be a congregation by year.
As Scheitle (2010) notes, it is not uncommon for organizations that are not what most people would think of as a “church” to claim church status with the IRS. Some religious charities do this because they genuinely see themselves as a congregation even though an outsider might primarily see them as a noncongregational religious publishing or missionary charity. For example, the Christian organization Young Life provides camps and clubs for children, teenagers, and college students around the world and has annual revenue more than US$250 million in recent years, but it was granted church status by the IRS in 2005 (Bostwick, 2007; Scheitle, 2010).
In other cases, organizations claim church status because, as discussed above, it frees them from many of the filing and transparency requirements with the IRS. Neither the IRS nor the court system is oblivious to this fact. A court in 1980 stated that (quoted in Scheitle, 2010),
. . . our tolerance for taxpayers who establish churches solely for tax avoidance purposes is reaching a breaking point. Not only do these taxpayers use the pretext of a church to avoid paying their fair share of taxes, even when their brazen schemes are uncovered many of them resort to the courts in a shameless attempt to vindicate themselves. (p. 147)
A much more significant concern with the IRS numbers is the issue of disbanded organizations. There is little follow-up in the years after a congregation registers with the IRS because, unlike other public charities, congregations are not required to file any sort of annual financial report. They are even exempt from the e-postcard filing requirement that was instituted in 2008 that was meant to update the IRS about the status of small nonprofit organizations. 5 This means that a congregation could register with the IRS and later close but remain on the IRS registry indefinitely. Given this, we must assume that at least some of the congregations listed in the IRS registry have now disbanded but are still listed.
Given these issues, one might consider other data sources in estimating the number of congregations registered with the IRS. There is, in fact, one possibility. The 1998 National Congregations Study is the only nationally representative survey that we are aware of that has asked congregations whether they have “filed with the Internal Revenue Service for its own official 501(c)(3) status?” 6 Unfortunately, this question was not asked in the 2006-2007 edition of the National Congregations Study (NCS), so we are limited to these somewhat outdated data. Nonetheless, it might provide a way to estimate how far off the IRS numbers were in 1998 and a way to correct for future years’ estimates.
Twenty-seven percent of congregations in the 1998 NCS said they were listed as a subordinate of their denomination, just below 40% said that they were listed individually, 26% said that they were not listed, and 8% said that they did not know their status. If we combine the first two numbers, this would equate to about 67% of congregations being listed with the IRS. This becomes problematic, however, when we compare these numbers with the IRS database. Estimates for the total number of congregations in the United States range from 325,000 to 400,000 (Cnaan & Curtis, 2012; Hadaway & Marler, 2005), so based on that 67% we would expect between 217,000 and 268,000 congregations listed with the IRS in 1998. But in fact there were only 95,101 organizations claiming to be a congregation in the September 1998 IRS database. This is less than half of what we would expect based on the NCS.
One factor in producing these contradictory findings is the congregation’s interpretation of the “through our denomination” response. It is possible that the congregation assumes that because the denomination has received recognition from the IRS, all its congregations have, by default, received recognition. This is not actually the case unless the denomination has applied for the group exemption and has specifically listed the congregation as a subordinate.
But even if we discount all the “through our denomination” responses (which is likely too harsh given that certainly some of those congregations are listed as subordinates of a denomination), the almost 40% of congregations that stated that they have received recognition independently would still produce an estimated 130,000 to 160,000 listings with the IRS, which is still many more than were actually listed with the IRS at the time of the NCS. This suggests a couple of other factors in producing these conflicting estimates. First, even though the question asked explicitly about whether the congregation had “filed” for an exempt status, it is possible that many of the congregations interpreted the question as asking whether the congregation is tax-exempt. As noted earlier, congregations are unique in nonprofit law in that they are by default granted exempt status even without asking for it explicitly. Another issue might be that many congregations are simply uncertain or unclear about their filing status. Indeed, almost 8% stated that they “don’t know,” so it is possible that a significant proportion of those that did give a more definitive response were not actually certain either.
In short, neither survey-produced estimates nor the actual IRS counts of congregations can be trusted, as both are undoubtedly overestimates. It might be possible to take a sample from the IRS registry and investigate the status of each congregation through mailings, phone calls, and Internet searches, but this could prove a surprisingly difficult and flawed approach as well. Because of the lack of reporting requirements for congregations, the address and phone information for many congregations listed with the IRS could be out of date even if the congregation is still active. This could lead one to misinterpret disconnected phone numbers and invalid addresses as a closed congregation. Most congregations are very small and resource poor (Chaves, 2004), so they may not have much of an electronic footprint, either. For instance, the 2006 National Congregations Study found that 55% of congregations did not have a website (Chaves & Anderson, 2008). In addition, some congregations may change denominational affiliation and possibly congregational name (such as may have occurred with recent schisms in various Protestant denominations over issues of homosexuality, for instance, resulting in Episcopal congregations joining the Anglican Church in North American and Evangelical Lutheran Church in America [ELCA] Lutheran congregations joining the North American Lutheran Church), which may also produce the appearance of a congregational closure when it could instead be unregistered with the new name or registered with its new denomination.
Are the IRS Congregations Representative?
It might not be a problem whether the IRS registry is an incomplete subset of the population of congregations or whether it contains disbanded congregations if these weaknesses are randomly distributed. Unfortunately, there are significant reasons to think that this not the case, as seen in Table 1. If we look at the regional distribution of congregations in the IRS database to the known regional distribution of the congregational population in the National Congregations Study we see significant discrepancies. In particular, the IRS database seems to under-represent congregations in the south and over-represent congregations in the west and, to a lesser extent, the northeast. Certain denominations and denominational traditions, such as Baptists, are concentrated in the south. So, this discrepancy could be explained if congregations of these denominations and traditions are less likely to register with the IRS. In fact, there is some evidence of this. The 2006 NCS found that 30% of congregations are Baptist, but only 10.7% of congregations in the November 2010 edition of the IRS database have the word “baptist” in their name. We must be cautious, of course, as there are limitations of relying on a congregation’s name to identify its denominational or religious tradition. First, even if a congregation has the word “baptist” in its name, we cannot determine the specific Baptist denomination to which it belongs. Second, some traditions and denominations do not readily lend themselves to particular keyword searches. Finally, some congregational names may simply be misleading as a proxy for tradition or denomination. However, this method is at least suggestive of potential patterns.
Regional Distribution of All Congregations Compared With IRS-Registered Congregations.
Source. 2006-2007 National Congregations Study (weighted to reflect congregational population) and November 2010 IRS Business Master File.
Note. IRS = Internal Revenue Service.
How Do Congregations Get Registered?
As mentioned earlier, a congregation can apply individually for tax-exempt recognition, or it can receive tax-exempt recognition through its denomination or national organization, assuming the latter has a group exemption and claims the congregation as a subordinate. Either way, the congregation will be listed in the IRS database, as the holder of a group exemption (i.e., the denomination) provides a list of individual chapters (i.e., congregations) to the IRS for registration. Each subordinate has its own unique Employer Identification Number, though, so it is not immediately apparent that they are connected in the IRS database.
It is worth emphasizing that not all denominations have a group exemption, and even among those that do have one, the exemption does not always encompass all of a denomination’s member congregations. A denomination could have some of its congregations recognized under a group exemption whereas other congregations are individually recognized as exempt or not recognized at all. Factors determining this include the congregation’s preference, the denomination’s policies about being included in the group exemption, and likely the denomination’s extent of control over individual congregations.
As an example, consider the variation in policies between the two largest Lutheran denominations in the United States. The Lutheran Church–Missouri Synod’s (LCMS) Congregational Treasurer’s Handbook states,
Congregations listed in “The Lutheran Annual” are included in a blanket tax-determination letter from the IRS . . . The listing of congregations is updated annually. Based on this letter, all listed congregations including those in the formative stages are exempt from income taxes as organizations described in Section 501(c)(3) of the Internal Revenue Code, (2014: 8-2).
Note that this statement does not ask individual congregations to apply for inclusion. However, the ELCA’s website for “Congregation Administration” states that member congregations can receive recognition under the denomination’s group exemption, but it requires congregations to apply for this inclusion.
We see evidence of these two group exemptions in the IRS database. As shown in Table 2, if we search for congregations with the word “Lutheran” in their name, we find that the ruling dates of these congregations, or the date of their tax-exempt recognition, are heavily clustered around the ruling dates of the two denominations’ group exemption dates. 7 The LCMS’s group exemption date is 1941, and 45% of congregations in the IRS database with “Lutheran” in their name have a ruling date of 1941. Another 33% of IRS-registered congregations with “Lutheran” in their name have a ruling date of 1987, which corresponds to the ELCA’s group exemption date. Interestingly, the LCMS reported just more than 6,040 congregations in a 2010 enumeration (Grammich et al., 2012), which corresponds almost exactly with the number of congregations in the IRS database that appear to be part of that denomination. However, ELCA claimed more than 9,800 congregations in 2010, but less than half of those appear to be included in its group exemption. This likely reflects the more unilateral and centralized policy of the LCMS relative to the ELCA regarding including its congregations in its group exemption.
IRS Exemption Dates of Congregations With “Lutheran” in Their Name.
Source. November 2010 IRS Business Master File.
Note. IRS = Internal Revenue Service.
We should note that just because all these congregations have the same ruling date does not mean they were actually established in that year. In the case of a group exemption, the subordinates receive the ruling year of the parent organization. Researchers and organizations such as the NCCS already caution against using a nonprofit’s ruling date as a precise measure of its founding date (NCCS, 2010). This is partially due to the IRS’s recordkeeping limitations for pre-1960s organizations, but also because official recognition by the IRS often comes a few years after an organization is founded even for new nonprofits. For organizations such as congregations that are often part of a group exemption, though, the ruling date is even more problematic as a proxy for founding date.
As seen in Table 3, there are significant variations across denominational traditions in whether a congregation is registered independently or as part of a group exemption. It is well-known that denominations vary in the amount of governance and control they have over their member congregations. Sullins (2004) classified Protestant denominations into three levels of centralization, and these appear to overlap with the route in which congregations become listed with the IRS. Sullins’s “most centralized” denominations include the Episcopal Church. In the IRS database, congregations with the word “episcopal” in their name are overwhelmingly (85%) registered through a group exemption. Sullins classifies many Baptist denominations (e.g., Southern Baptist Convention, National Baptist Convention, USA, American Baptist Church in the USA, and so on) as “moderately centralized.” In the IRS database, congregations with the word “baptist” in their name are split between independent and group-affiliated exemptions. Finally, Sullins classifies many Pentecostal and holiness denominations (e.g., Assemblies of God, Church of God in Christ) as “decentralized.” Accordingly, if we look at the IRS database, congregations with the word “Pentecostal” in their name tend to be more independent exemptions instead of group exemptions. However, congregations that are most likely to be registered independently appear to be those of non-Christian traditions. For example, of congregations with the keywords “Islamic” or “masjid” (mosque) in their name, 93% are listed independently. Similarly, 100% of those with the keyword “Hindu” or “mandir” (temple) are listed independently.
Independent or Group Exemption Status of IRS-Registered Nonprofits Claiming to Be a Congregation by Keywords in Congregation’s Name.
Source. November 2010 IRS Business Master File.
Note. IRS = Internal Revenue Service.
It is important to recognize, though, that denominational centralization does not always lead to member congregations being registered with the IRS. As an example, Sullins classified the Church of Jesus Christ of Latter-day Saints (LDS), commonly called the Mormon Church, as “most centralized.” But we can find no listings of individual congregations (called wards in the Mormon Church) in the IRS registry. Although speculative, the almost total lack of registration of Mormon congregations, either as independent registrations or as part of a group registration, could reflect the simultaneous high level of centralized power of the Mormon Church and its desire to keep its congregations independent and unattached to the federal government. This fits well with historical and theological interpretations of church–state relations in Mormon history. In summarizing the LDS view on the church–state relationship, Durham and Oman (2006) state that,
the quest for state endorsement or support may distract religious organizations from their central missions, or from structuring their missions as they think best. Freedom to maintain doctrinal and organizational purity has always been vital to LDS leaders. (p. 29)
If we look at the trend in group versus individual registrations of congregations, we find some evidence that the increasing use of group exemptions has played a part in the growth of congregations registering with the IRS. As shown in Table 4, in 1998, 58% of the exemptions of nonprofits claiming to be a congregation were individual exemptions, while 41% were subordinates in a group exemption. Over a decade later, an additional 100,000 congregations were added to the IRS database and the share of group exemptions had increased to 47% of all congregational exemptions. As the last row in the table shows, an additional 150 “parent” organizations, in this case denominations, had been added to the IRS registry.
Independent or Group Exemption Status of Registered Nonprofits Claiming to Be a Congregation by Date of IRS Database.
Source. IRS Business Master Files (September 1998 and November 2010).
Note. IRS = Internal Revenue Service.
Conclusion
A few points are worth emphasizing from the above exploration. First, the number of congregations listed with the IRS is very likely an overestimate of the number of active congregations listed with the IRS due to the weak or nonexistent reporting requirements for congregations. Second, congregations can become listed with the IRS either independently or as a subordinate in a group exemption if (a) the congregation belongs to a denomination, (b) the denomination has a group exemption, and (c) the denomination lists the congregation as a subordinate. However, as we saw in our comparison of the Lutheran Church–Missouri Synod and the ELCA, denominations with a group exemption vary significantly in their policies about listing their congregations under that exemption. Interestingly, the method of being listed appears to correlate with the centralization level of the congregation’s religious tradition. Centralization of denominational authority, however, does not guarantee registration, as is seen in the apparent systematic nonregistration of Church of Jesus Christ of LDS (Mormon) congregations.
Based on all these, we suggest that social scientists be very thoughtful about using the IRS registry of congregations for sampling purposes, as it is clearly not representative in any way. However, further historical and qualitative research on the process by which congregations become registered (or not registered) could contribute to our understanding of the American religious market and our understanding of organizational processes. For instance, what is the history of decision making for different denominations in their thinking about how to approach the exemption of their affiliated congregations? Do congregations that are not registered with the IRS face any costs by their lack of registration? In addition, further research could shed light on the institutional forces that may guide a congregation’s registration. For example, some congregations may choose to register with the IRS as a means of establishing legitimacy as an organization, particularly if a congregation is not part of an established denomination from which to derive legitimacy. New Institution Theory suggests that organizations tend to imitate other organizations, particularly those perceived as successful or at least normative, to gain legitimacy, which is essential for garnering resources in a given institutional field (see DiMaggio & Powell, 1983). Independent congregations may feel that IRS registration could make members more comfortable about donating to the congregation or may reinforce the congregation’s status as a legitimate congregation, particularly in environments where many congregations are affiliated with a major denomination, such as in the northeast. However, congregations in environments where centralized denominations are less prominent and independent churches are more common, such as in the South, the legitimacy concerns that could lead to IRS registration may be less salient.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by a grant from the National Science Foundation’s Directorate for Social, Behavioral, and Economic Sciences (Award Number 0960617).