Theoretical Motivation and Literature Review

Reviews of the history of regulations regarding charitable solicitations (e. g., Barber and Farwell 2016) discuss the steady increase in the number of states that regulate these activities since the early twentieth century. One significant motivation has been to instill confidence in donors, and the general public, that charitable contributions will fund donor-directed activities. However, to date, very little published research has focused specifically on the impacts of state regulation of charitable solicitations, in part because of the challenges of characterizing state regulatory environments for nonprofits. One of the earliest attempts to map the state-level landscape of nonprofit regulation was a study done in the 1970s by the Ohio Attorney General’s Office, as one among many studies conducted for the Commission on Private Philanthropy and Public Needs (the “Filer Commission”) (Office of the Ohio Attorney General 1977). This study used state statutes and surveys of state regulators to identify and assess various forms of authority that could be present in each state.

In a later study, Irvin (2005) used a simpler method to divide states into two groups: those with registration and reporting requirements for nonprofits, and those (six at the time) with no such requirements (164). Irvin’s study attempted to calculate the net benefits of nonprofit regulation generally by comparing the costs of compliance with state regulations against the benefits to society of such regulation. Her measures include total donations to charity, which, according to Irvin, signal public trust in the sector; the costs of compliance; and qualitative data on enforcement activity in the states with no regulation (168–172). Irvin argues for partial deregulation of the sector because costs of compliance are high and because regulatory requirements do not seem to be related to support for the state’s charities (173).

More recently, several scholars have investigated the link between state regulations and outcomes other than the amount contributed to charitable causes. Desai and Yetman (2005) examine state laws regarding governance and reporting requirements, as well as fully adjudicated cases reported by the media to enforce the state laws, to map the landscape of regulations addressing governance in nonprofit organizations within each state. The state laws examined in their study include factors ranging from notice requirements to legal standing, cy pres provisions to audit requirements, among others (7–8). The authors use data on these regulations to construct two indices, a “detection index” and a “prosecution index,” by counting the number of requirements present in state statutes (9). They find that, in states with higher values for the detection index, harmful outcomes such as excessive managerial compensation, unreasonable fundraising expenditures, and reduced “charitable spending” are less prevalent (19–20). A 2016 study regarding private foundation compliance with new state-level regulations (Galle 2016) also involved mapping part of the regulatory landscape. Galle used data about staff and resources available to state attorney general offices as well as state statutes granting legal standing to private individuals (18–20) to measure the opportunity for enforcement. However, the study found no correlation between the resources for enforcement and private foundations’ compliance with state law (25).

Although some of these studies had limited geographical scope – Irvin’s study, for instance, compared the six nonregulated states, which are primarily western states with low- density populations, with the others – other attempts to fully characterize state regulatory environments have additional conceptual limitations. As argued in a recent study published by the Center on Nonprofits and Philanthropy at the Urban Institute and Columbia Law School’s Charities Regulation and Oversight Project (Lott et al. 2016a), public enforcement actions, i. e., civil or criminal trials, are only a small fraction of the enforcement activity taken by state regulators (Lott et al. 2016a, 5–6). Thus, studies that focus on litigation activity reported in the media often fail to account for other types of enforcement activity, often more subtle or one-on-one between the regulating office and the organization. In addition, state attorneys general typically also have common law authority to bring certain types of enforcement actions, such as prosecuting fraud (National Association of Attorneys General 2013). As a result, studies that focus solely on the statutory authority of the regulators often fail to account for the complete range of enforcement activity, and will likely underestimate the level of enforcement capacity in a state where few laws regarding charitable solicitations are on the books.

In general, the impact of “enforcement” on the effectiveness of state regulation upon fundraising needs further investigation and data collection by researchers. In recent decades, many states have changed their statutes to emphasize the importance of public disclosure of solicitation-related activities, rather than the need to report these requirements to regulatory staff (Barber and Farwell 2016). Although some have speculated that reporting requirements may induce organizations to change their behavior in response to a perceived increase in awareness by donors, most agree that regulators need to play an active role in enforcing regulations (ibid.). For this reason, enforcement must be measured directly to reach conclusions about impact, especially if the goal is to establish connections between state fundraising regulations, the ability of charities to raise significant amounts in contributions, and the level of “public trust” in the nonprofit sector. However, to date, the existing literature has not yet identified the most effective measures and data sources for enforcement; by counting fraud cases reported in the media, fully adjudicated cases taken to trial, and other observable court-based actions, most studies are not considering the entire spectrum of tools of enforcement available to state charities oversight officers.

Creating a State Index of Regulatory Breadth

Our index of regulatory breadth is based in part on a recently published report (Lott et al. 2016a) that features a complete and systematic analysis of state-level oversight and regulation of charities. The analysis featured in this report has three components: a legal analysis of laws pertaining to charities in 56 US jurisdictions; a survey of all state and territory offices with oversight, regulatory, and enforcement authority over charities (with at least one office within 47 jurisdictions completing the survey); and interviews with officials in over two-thirds of the responding offices. The report’s release was accompanied by on-line publication of a state-level compendium of state charitable regulations. We enhanced the data from this publicly available compendium (Lott et al. 2016b) by conducting further legal research into elements of regulatory requirements for fundraising.

Our analysis begins with a review of state-level regulations that require charities and/or their fundraisers to notify the state, donors, or the public about certain aspects of their organization’s status, activities or professional relationships. These requirements offer both the possibility of detection of problems within an organization and also invite public scrutiny. We utilize these regulations due in part to the availability of other relevant data; as Lott et al. (2016a) observe, “little systematic information exists on the parameters, frequency, consistency, challenges, and benefits of the state regulatory and enforcement structure.” Our index focuses on registration, notice and filing requirements, and not on common law, which often supports oversight in practice.

Following previous studies, we measure state-level regulatory breadth by counting the number of important indicators found in each state’s regulatory code specific to charitable solicitation. We also consulted current or former state regulators to find out which indicators they would consider hallmarks of a robust regulatory regime. Generally, these indicators capture one or more general characteristics of the way states oversee charities and those who fundraise on their behalf:

1. To whom do the regulations apply/what kinds of individuals/organizations are regulated?

2. What aspects of the relationship between an organization and its fundraiser(s) does the state regulate?

3. What does the state’s regulatory scheme prohibit? and

4. What information is available to the public and donors?

Ultimately, after consulting with our group of state regulators, we created an index of “regulatory breadth” using the following ten elements:

1. Does the state require registration by charitable organizations?

2. Does the state require registration by commercial fundraisers?

3. Does the state require bonding of professional fundraisers?

4. Does the state require registration by fundraising counsel?

5. Does the state oversee commercial co-venturing (by requiring that the co-venture be registered or by requiring that the charitable organization file the co-venture contract)?

6. Does the state require annual financial reporting by charitable organizations in addition to filing a copy of the 990 with the regulator (if filing 990 is required)?

7. Does the state require annual financial reporting by commercial fundraisers?

8. Does the state require a copy of any contract between a charitable organization and a commercial fundraiser or fundraising counsel be filed with the regulator?

9. Does the state require the fundraisers to provide notice to the regulator before any solicitation campaign (in addition to annual registration and/or filing the contract)?

10. Does the state require specified disclosures to donors?

These ten items form a relatively unidimensional and cohesive (Cronbach’s α=0.850) index. In addition, almost all of the final index items load onto the first principal component identified in exploratory analysis of the available data.

The compendium published with Lott et al. (2016b) contains state-level data on many of these characteristics for all fifty states plus DC. After collecting data on the remaining indicators, we then coded each state as a “yes” or “no” based on our reading of state statutes. If the statute indicated that the state required the index item in any degree, no matter how narrow the requirement, we coded the item a “yes” for that state. If the state’s statutes were silent about a particular indicator, we coded the item a “no.” We constructed the index value for each state by simply counting the number of “yes” entries across the ten items. Appendix A contains the yes/no values for each state for each of the ten indicators we selected for our index of regulatory breadth.

The index has several desirable qualities, but also definite limitations – in particular, it does not attempt to measure enforcement of the regulations. As Lott et al. (2016a) observe, state charity offices vary substantially in their enforcement capacity, due to differences in budget constraints, staff turnover and training, and the office or offices where oversight authority is located. The extreme right column of Appendix A also contains a yes/no measure that indicates whether the state charity office involves bifurcated regulatory and enforcement jurisdiction. Bifurcated jurisdiction does not by itself indicate the strength of a state charity office’s capacity to enforce fundraising regulations; the 23 states with bifurcated jurisdictions vary in many important ways. However, collaboration and coordination between the offices is especially important in these states, especially when budget and staff constraints may affect enforcement capacity.

The Growth in Giving Initiative

After creating our state-level index of regulatory breadth, we use data collected by the Growth in Giving Initiative (GiG) to develop measures of fundraising performance. The GiG, launched in 2012 by the Association of Fundraising Professionals (AFP) and the Center on Nonprofits and Philanthropy at the Urban Institute, uses a unique data collection process that allows vendors of donor software to contribute anonymized data on four gift-transaction data fields: Organization ID, Donor ID (both anonymized, and time-consistent so that a donor’s history of giving to an organization can be traced through time), Date of Receipt and Amount of Gift. Although the anonymized dataset has certain limitations – it cannot be used to trace all the organizations a single donor supports, for instance – these fields can produce dozens of key metrics of fundraising performance, including many used by professional fundraisers, such as donor retention and acquisition rates.

To date, software vendors have contributed data on over 110 million individual contributions recorded by over 10,000 client organizations from 2006 to the present. Certain types of organizations in the GiG database tend to be underrepresented, such as very small and very large organizations (both of which are unlikely to be users of commercial software). However, the database tends to be substantively and geographically representative of small and medium-sized nonprofit organizations (Giving USA Philanthropy Spotlight 2016), while the large number of organizations boosts the statistical power of our analysis.

Empirical Analysis of State-Level Generosity

To begin, we develop a series of multivariate models, where the dependent variables are measures of the organization’s fundraising performance in a given year. Using the GiG database, we create five organization-level metrics that serve as dependent variables. These metrics include total amount contributed to the organization from all donors (within a given year); total number of gifts received; total number of unique people making donations; number of existing donors who gave the following year (“donors retained”); and number of new donors who had not given to the organization. All variables are logged to temper the impact of extremely large organizations.

The goal of our analysis is to model the charitable environment of the place where the organization is located, and use the results to isolate the impact of organization-specific factors that influence fundraising performance. Although some earlier studies (Seeley 1957; Havens and Schervish 2005, 2007) developed multivariate models of the charitable environment, we borrow the model specification used by Brown and Rooney (2005). The Brown-Rooney model contains measures of a number of state-level socioeconomic and demographic characteristics as well as measures of state economic conditions. We add a proxy for the organization’s size: the number of donors contributing to the organization in a given year. Although the Brown-Rooney dependent variable – average amount contributed to charity by individuals who itemize their tax returns – is different from ours, their model specification appears to perform well overall for our organization-specific dependent variables. Appendix B contains descriptive statistics for the variables in our multivariate models.

Because many organizations in the GiG database have contributions data for all ten years (2006–2015), we can use panel-data methods to isolate unmeasured organization-specific effects that influence fundraising. Appendix C contains the results of our five panel-data multivariate regression models. For each model, we estimate a fixed-effects model and retrieve the organization-specific fixed effects from the results. Assuming that the multivariate model is correctly specified, the fixed effects are unbiased, consistent measures of these organization-specific effects.

The fixed effects can also be interpreted as measures of the extent to which the organization’s fundraising performance exceeds – or falls short of – expected levels, controlling for the measurable variables that capture the state’s baseline charitable environment. For each dependent variable, we use the median values of these organization-specific fixed effects to form state-level fundraising metrics. The cardinal values of these state-level measures are difficult to interpret, and their normative implications even more so: organizations may raise prodigious amounts of funds by engaging in unethical or illegal practices, or may attract support from donors for personal reasons that are unrelated to the activities of state regulators or any other measurable factors. However, higher/positive values for our fundraising metrics indicate that the organizations in that state tended to raise more money, collect contributions from more donors, and/or acquire more new donors than would be expected given the state’s fundraising environment. Similarly, lower/negative values indicate that the organizations have lower-than-predicted performance in collecting contributions, controlling for other factors in the state.

Comparison of State Regulatory Robustness and Fundraising Performance

Finally, we investigate the relationship between our five metrics of fundraising performance and the scores on our index of regulatory breadth. Simple correlations of these state-level measures show that states with higher scores on the regulatory index tend to have slightly lower values for all five metrics, controlling for all other factors (but, notably, not for levels of enforcement which, as noted, cannot be measured from currently available data). Data from the District of Columbia are excluded from all results, because the observations are outliers in each model. The correlations also suggest that greater regulatory breadth is more strongly associated with lower amounts for total dollars raised, total number of gifts received, and total number of people donating, while having less effect on retention of existing donors or acquisition of new donors, controlling for the fundraising environment.

However, the state-level relationships appear to be curvilinear: for each metric, we see a larger difference in median fundraising performance between states with index values between 7 and 10 (inclusive) than between states with index values between 0 and 6 (inclusive). To control for other observable differences in the structure of state charity offices, we create a four-category variable that separate states into categories based on whether they have broad or narrow regulatory breadth (scores of 7–10 on the index vs. 0–6), or bifurcated or non-bifurcated jurisdiction for nonprofit oversight. We use the resulting four-category variable to examine whether robust regulation and bifurcated jurisdiction, taken together, are associated with higher or lower state-level fundraising metrics, controlling for environmental factors. The results (seen in Appendix D) suggest, in general, that these metrics are likely to be lower in states with both high regulatory breadth (high index values) and bifurcated jurisdiction over nonprofit oversight.