Covenants, Constitutions, and Distinct Law Types: Investigating Governments’ Restrictions on CSOs Using an Institutional Approach

Abstract

A growing number of researchers study the laws that regulate the third sector and caution the legal expansion is a global crackdown on civil society. This article asks two questions of a thoroughly researched form of legal repression: restrictions on foreign aid to CSOs. First, do institutional differences affect the adoption of these laws? Second, do laws that appear different in content also have different causes? A two-stage analysis addresses these questions using data from 138 countries from 1993 to 2012. The first analysis studies the ratification of the International Covenant on Civil and Political Rights (ICCPR) and constitution-level differences regarding international treaties’ status. The study then uses competing risk models to assess whether the factors that predict adoption vary across law types. The study finds that given ICCPR ratification, constitutions that privilege treaties above ordinary legislation create an institutional context that makes adoption less likely. Competing risk models suggest different laws have different risk factors, which implies these laws are more conceptually distinct than equivalent. Incorporating these findings in future work will strengthen the theory, methods, and concepts used to understand the legal approaches that regulate civil society.

Introduction

Research studying the laws and policies regulating voluntary association around the world has been a growth industry since the mid-2000s. Despite the growing body of work on public regulation of civil society, important questions remain. While focusing primarily on the adoption of legislation restricting civil society organizations (CSOs), previous work has not yet studied the broader legal context in which lawmaking occurs. This context is crucial because it informs the institutional constraints that inhibit or facilitate the adoption of new laws. It is necessary to consider the effect that preexisting legal commitments and constitutional rules have on CSO legislation. Moreover, laws that appear qualitatively different—such as those passed by Oman (2000) that categorically prohibit foreign funding versus those that merely require ex post notification or accounting requirements (e.g., Pakistan 2003 and Uruguay 2004)—might be promulgated for different reasons and implemented with varying degrees of success. This raises another important consideration: If the laws that restrict CSOs differ in content, could it be that different factors affect the adoption of different laws?

This article makes two key contributions. Its first builds on current theory to introduce institutional variables, specifically preexisting institutions and constitutional differences, which many prior analyses omit. The research design accomplishes this by simultaneously analyzing two key institutional variables. The International Covenant on Civil and Political Rights (ICCPR) is the first. It is a critical international human rights treaty that commits its parties to promote human rights and fundamental freedoms (Donnelly 2013; Henkin 2000). Ratification makes the covenant a preexisting institution that constrains legal attempts to undermine civil and political rights. Ratification may not be sufficient, however, to protect these rights because constitutional rules condition whether the obligations enshrined in the ICCPR affect domestic lawmaking. From an institutional analysis perspective, the consequences of ICCPR ratification depend on constitutional rules.

The second analysis disaggregates the types of laws to explore nuanced relationships. Given reports of the growing number of laws restricting civil society around the world (e.g., Amnesty International 2019; Anheier and Toepler 2019; CIVICUS 2018; Musila 2019; Rakner 2019), it is empirically expedient and sometimes necessary to classify all laws as equally restrictive and presume marginal and unimportant differences. The second analysis tests the degree to which this presumption holds. It does so by reorganizing laws restricting foreign funding into three qualitatively different groups that fall along a continuum from highly to minimally-restrictive. The quantitative analysis assesses whether the factors that predict adoption for the “pooled” conceptualization also predict adoption of distinct law types. It also discusses whether the presumption of conceptual equivalence produces either type I or type II errors.

To review, this analysis uses preexisting institutions and legal differences to build theory and further explain the adoption of laws that restrict CSOs. It accomplishes this by investigating two research questions: First, do preexisting institutions affect the adoption of restrictive laws? Second, do the political factors that predict the adoption of highly-restrictive laws also predict the adoption of moderately and minimally-restrictive ones? These are essential questions that advance theory. As an example, leading research shows that 39 countries adopted restrictive foreign financing laws between 1999 and 2013 (Dupuy et al. 2016). Figure 1 uses two indicators to highlight the range of countries that have adopted restrictive foreign funding laws. The vertical location of each marker identifies the level of democracy in each country along a continuous scale with larger values indicating higher levels of democracy. Shapes identify states’ Freedom House Status category in the year of adoption. As expected, the data show a strong relationship between authoritarianism and adoption. The data also show that democracies and hybrid regimes adopt these laws. Indeed, we see almost as many Free or Partly Free countries adopted restrictive laws as Not Free states. This figure raises the question of whether different countries are passing the same law, or if different countries pass different laws.

Fig. 1
figure1

Sources “Hands Off My Regime!” (Dupuy et al. 2016: Table 1); Values of Democracy (V-Dem); Freedom House

Adoption of more restrictive foreign funding laws. Solid shapes identify the year a country adopts its restrictive law, and hollow shapes identify countries that passed a second law during the analysis period. These include Belarus (2001, 2003); Indonesia (2004, 2008); and Uzbekistan (2003, 2004). Background shows country-year values of nearly 100 additional countries that did not adopt laws.

Two theory-driven arguments motivate this study. First, countries with preexisting institutions to promote civil and political rights are less likely to adopt laws that restrict voluntary association. Second, beginning with the premise that laws are more different than presumed, the factors that predict adoption vary across highly, moderately, and minimally-restrictive laws. I test these hypotheses using a form of event history analysis known as competing risk models (CRMs) to gain refined estimates on how different factors—such as the level of democracy, voting alignment with superpowers in the United Nations, preexisting institutions, and organized civil society activity—relate to the adoption of particular laws. The analysis uses several sources to rebuild the dataset used in “Hands Off My Regime!” (Dupuy et al. 2016). The models analyze a sample of 138 countries between 1993 and 2012, of which 37Footnote 1 adopted laws identified as highly, moderately, or minimally-restrictive.

A summary of my results follows. First, constitutions and preexisting institutions matter. Given ICCPR ratification, countries with constitutional rules that explicitly place treaties above ordinary legislation are less likely to adopt restrictive laws. This institutional arrangement is significant across law types and robustness checks. Second, findings suggest that laws are qualitatively and quantitatively different. Nuanced results from the CRMs show the risk factors that predict the adoption of one law type rarely accurately predict the adoption of two or more types. These findings suggest that “pooling” distinct laws into one broad category can produce false positives that overstate the variables’ relationship with the adoption of different law types. Likewise, presuming equivalence of distinct laws makes analyses susceptible to false negatives that wrongly reject the importance of a factor that is strongly associated with the adoption of a specific law type. I conclude by discussing the implications these results have for studying the regulation of civil society around the globe.

Theory

Institutions Affecting Voluntary Association

Vast areas of the literature discuss civil society as a crucial factor for important sociopolitical outcomes such as governance, democracy, and interpersonal trust (Aligica 2018; Brass 2016; de Tocqueville 1840; Edwards 2004; Fukuyama 1995; Putnam 1993). The relationship is not unidirectional, and politics can and do affect CSOs. One manifestation of politics affecting civil society appears in a country’s CSO regulatory regime, or the legal framework of various laws and constitutional protections that create carefully institutionalized regulatory systems that structure the activity of CSOs (DeMattee 2019). These constellations of legal protections have been part of the literature on non-governmental organizations and international development for over 35 years (Brass et al. 2018) and exist alongside self-regulation instruments (Breen et al. 2017, 2019; Crack 2018). A growing literature attempts to predict and explain the adoption of restrictive laws that seek to minimize the political influence of civil society, or what some refer to as the “Closing Space” phenomena (Carothers 2015; Carothers and Brechenmacher 2014). This research explains that state repression is not only physical but also legal and judicial. It argues that regimes use laws to reconfigure regulatory regimes to protect their hold on political power (Carothers 2006; Christensen and Weinstein 2013; Dupuy et al. 2016).

Analysts discuss the degree to which differences in constitutions, ratification of international treaties, and domestic laws affect de facto civil liberties and the organizational ecology of CSOs (Elkins et al. 2009; Hathaway 2002; Salamon and Toepler 1997; World Bank 1997). From an institutional analysis perspective, constitutions are the ultimate preexisting institution because they structure the terms and conditions of governance and establish the boundaries of governmental activity (Buchanan and Tullock 1961; Ostrom 1997) and inform our understanding of how predatory states develop a credible commitment and the state capacity necessary to undertake protective and productive roles for society (Boettke and Candela 2019; Buchanan 1975). Constitutions sit atop a multi-level rulemaking process that determines the creation of laws and working rules that affect decision-making (Brennan and Buchanan 1985; Cole 2017; Ostrom 2005). Relevant examples are systems of checks and balances, whether ratification of international treaties constrains domestic lawmaking, and power bestowed to the executive. At the same time, constitutions affect individuals by creating and maintaining a sense of shared identity (Breslin 2009; Murphy 1993; Mutunga 1999; Pitkin 1987), protecting and enforcing property rights conducive to development (North and Weingast 1989), and constitutionalism that guarantees a state’s commitment to “a set of inviolable principles” such as freedom of association and religion (Elkins et al. 2014: 38). Studying constitutional differences provides insight into the constraints on rulemaking and the importance of preexisting rules. Although constitution-level rules change slower than collective-choice or operational rules (Ostrom and Ostrom 2004), scholarship on the topic finds the average life expectancy of constitutions is only 19 years with a “decline in constitutional life spans after World War II” (Elkins et al. 2014: 131).

At the international level, the 1966 International Covenant on Civil and Political Rights (ICCPR) is considered a principal treaty in the area of international human rights (Donnelly 2013; Henkin 2000; ICNL 2009, 2015; Kiai 2012).Footnote 2 The ICCPR guards the civil and political rights of citizens in the majority of countries in the world with new signatories ratifying the agreement each year (United Nations Office of Legal Affairs 2018).Footnote 3 Article 22(2) outlines the limited conditions under which restrictions on association are permissible with similar language appearing in several subsequent regional treaties such as Article 16(2) of the American Convention on Human Rights (1969), Article 11 of the African Charter on Human and Peoples’ Rights (1981), and Article 11 of the European Convention on Human Rights (2010) .

International law uses Article 22(2) to establish legal criteria to evaluate the legitimacy of laws affecting voluntary association (UN Human Rights Committee 2006, 2007, 2015). A three-part test sets a threshold that all rules regulating voluntary association must be:

  1. 1.

    Prescribed by law that uses sufficiently precise and accessible language;

  2. 2.

    Established to meet legitimate aims specified by Article 22(2) to include “national security or public safety, public order, the protection of public health or morals or the protection of the rights and freedoms of others”; and

  3. 3.

    “Necessary for democracy” in that they meet a pressing social need in a proportional manner.

In general, the three-part test suggests that states may regulate (prescribe by law) CSOs to perform specific actions in the interests of transparency and accountability (legitimate aims) if such requirements are properly scoped to prevent real dangers to democracy (necessary for democracy).

Ratification of the ICCPR provides information concerning when it becomes a preexisting institution. But the strength of the ratification as a preexisting institution depends on the status constitutional rules give international treaties. Thus, the preexisting institution is most influential when a constitution elevates the ICCPR’s international commitments above ordinary legislation. Table 1 shows the year each country adopted its restrictive law (Law adopted). It also displays the institutional context present in the country the year it adopted its law: the ratification status of the ICCPR (ICCPR ratified) and whether constitution constitutional rules make international treaties superior to ordinary legislation (Treaty Superior). Of these 39 countries, 25 ratified the ICCPR before the start of the observation period and nine ratified during the observation period. But two of those countries, Indonesia and Pakistan, ratified after adopting restrictive laws. In total, 33 countries adopted laws in an institutional context that either had no preexisting commitments to promote civil and political rights or had no constitutional rules enforcing ICCPR ratification.

Table 1 Varying institutional contexts—adoption, ICCPR ratification status, and constitutional rules.

Countries that abstained from adoption also changed their institutional contexts during the observations period. Of the 101 countries that did not adopt restrictive laws, 23 changed their constitutional rules regarding international treaties, and 44 ratified the ICCPR during the analysis period. Figure 2 summarizes the institutional context for all countries in the sample. Abstainers designate countries that do not adopt a law, while adopters identify those that eventually do. The integer values express the number of countries that sort into the given institutional context for at least 1 year. Percentage values denote the proportion of country-year observations that match the institutional context. The top branch of Fig. 2 is the institutional arrangement that maximizes the effect of the preexisting institutions. The next highest branch is less constraining because ratification exists alongside constitutional rules that give lawmakers the option to follow or ignore the ICCPR’s commitments. When the ICCPR is not ratified (Fig. 2, bottom branches), the constitutional rules concerning international treaties are inactive because ICCPR ratification is absent.

Fig. 2
figure2

Sources Dupuy et al. (2016, Table 1); United Nations Office of Legal Affairs; Comparative Constitutions Projects; United Nations Treaty Collection

Frequencies of Institutional Contexts. Integer values express the number of countries that sort into the given institutional context for at least 1 year during the analysis period (138 countries, 1993–2012). It is possible for a single country to represent multiple institutional contexts because ratification status and constitutional rules vary within countries over time. Circle sizes signify the proportion of country-year observations that match the given institutional context. The area of the four circles on the right is the same as the two in the center and the one on the left.

Constitutions and international treaties are preexisting institutions that shape the institutional context that affects lawmaking in several ways. Principally, the adoption of laws or policies is commonly an exercise of incremental change rather than significant reordering (Lindblom 1959; Pierson 2000). Next, preexisting policies provide society with the opportunity to form opinions about policy and give lawmakers and bureaucrats experience in implementing and adapting objectives to local conditions and preferences (Lowi 1964, 1972; Pierson 1993, 1994). Finally, understanding policy adoption as part of a link in a historical process of institutional change means the adopted law or policy has one of the four relationships with preexisting institutions: independent, complementary, contingent, or substitute (Mahajan and Peterson 1985). Analytical frameworks of institutional analysis underscore the importance of history and show that one period’s policy outcome shapes the rules of future political action arena (Cole et al. 2014; McGinnis and Ostrom 2014; Ostrom 1990, 2011; Ostrom and Cox 2010).

For this analysis, constitutions both define the processes of lawmaking and determine the degree to prior commitments constrain future legislation. The former are the rules of rulemaking, and the latter are preexisting institutions. The ICCPR is an international treaty whose parties accept additional commitments to promote civil and political rights. Once ratified, the treaty both modifies current institutions and constrains future attempts to alter the regulatory regime of voluntary association. But the strength of international treaties as preexisting institutions depends on constitutional rules. Some constitutions fail to discuss treaties altogether, some give treaties a status that is less than or equal to domestic legislation, and still others explicitly grant treaties a status that is superior to ordinary laws. Two institutional hypotheses follow:

H1A

ICCPR ratification decreases the probability of adoption.

H1B

The institutional arrangement that minimizes the probability of adoption is one that (A) ratifies the ICCPR and (B) possesses constitutional rules that make international treaties superior to ordinary legislation.

Restrictions on Foreign Aid to CSOs

Most research on the adoption of restrictive laws disagrees on the extent to which international or domestic factors influence adoption. Research by scholars testing the influence of policy diffusion across borders has mixed findings with some finding no significant relationships (Dupuy et al. 2016) and others finding strong support of neighborhood effects and international linkages (Reddy 2018). Others focus on the intervention of influential global leaders who either protect states who attempt to pass restrictive laws (Christensen and Weinstein 2013) or serve as the object of emulation in a leader–laggard model of policy adoption. This research is part of a larger body of work that discusses the laws that affect civil society around the world with a recent review identifying over 50 distinct types of provisions (Bloodgood et al. 2014; DeMattee 2019; Kameri-Mbote 2002; Maru 2017; Mayhew 2005; Ndegwa 1996; Salamon and Toepler 1997, 2000; Sidel 2017).

Despite the broad approach used by some analyses, many recent studies focus on provisions that restrict access to financial resources even though such provisions belong to a more substantial subgroup of provisions regulating organizations’ financial and non-financial resources (DeMattee 2019; Kiai et al. 2017). Legal experts identify various tactics of “philanthropic protectionism” that restrict the flow of resources to CSOs (Rutzen 2015) such as requiring prior governmental approval, capping the total amount of international funding, requiring burdensome reporting, and mandating the routing of foreign funds through state-controlled financial institutions.

With so many provisions restricting foreign aid to CSOs, it is empirically expedient and sometimes necessary to classify all laws under the same monolithic group—“restrictive laws”—which suggests there exist only slight differences among them. This practice may be a misstep for theoretical and conceptual reasons. Law and policy differences are central to theories that explain the adoption and effects of regulation.Footnote 4 Further conceptualization of these instruments can advance our understanding of the laws that restrict CSOs. Three ordering logics demonstrate that the consistent categorization of distinct law types is possible. The analysis conducted here investigates governments’ restrictions on CSOs ability to access foreign funding (Fig. 3, top row) and organizes these legal provisions from prohibitive (highly-restrictive, left) to notification (minimally-restrictive, right). The figure provides examples of restrictive provisions discussed by scholars (bottom row).

Fig. 3
figure3

Sources: Appe and Marchesini da Costa (2017), Carothers and Brechenmacher (2014), Chikoto-Schultz and Uzochukwu (2016); Christensen and Weinstein (2013); Cunningham (2018); Dupuy and Prakash (2017); Dupuy et al. (2016); Gershman and Allen (2006); Gugerty (2017); Hodenfield and Pegus (2013); Kameri-Mbote (2002); Maru (2017); Mayhew (2005); Rutzen (2015); Salamon and Toepler (1997); Sidel (2017); Tiwana and Belay (2010); Wolff and Poppe (2015); World Bank (1997)

Continuum of Restrictive Laws, Regulations, and Policies

An ordering logic grounded in access to foreign funding is not useful for conceptualizing other legal provisions that affect CSOs. For instance, a recent review found that analysts regularly discuss four broad subtypes of legal provisions: governance, formation, operations, and resources (DeMattee 2019). These broad subtypes include over 50 distinct provisions including whether organizations must meet a defined membership or financial threshold before they can register (e.g., Maru 2017: 56), or whether the government has the discretion to unilaterally intervene and dissolve an organization (e.g., Mayhew 2005: 745–746). The breadth of these legal provisions requires alternative logics to accommodate consistency in our conceptualizations. International law provides one alternative. Here, the three-part test organizes law types from highly-restrictive and illegitimate to minimally-restrictive and contestable (Fig. 3, second row). Transaction costs are another conceptualization logic (Salamon and Toepler 2000, 2012). Laws that impose debilitating transaction costs are predatory (highly-restrictive), while less restrictive laws may be either inefficient or proscriptive. The three logics use different criteria but produce a consistent categorization of restrictions on CSOs’ ability to access foreign funding.

Readers may disagree with the ordering of these provisions along the continuums of restrictiveness, illegitimacy, or transaction costs. And others may still propose more useful categorizations as prior attempts to classify laws, policies, and regulations have shown (e.g., Gormley 1986; Lowi 1964, 1972; Salamon 2002). The significant point on which I hope many agree is that the broad category of restrictive laws is not monolithic, and theory predicts these differences have consequences for research and theory.

Two sets of conceptual hypotheses follow. One tests the conceptual equivalence of factors, and the other evaluates their distinctness. If laws are equivalent, then the factors that predict adoption for one type of law should also predict the adoption of other laws. This means the factors believed to be positively associated with the adoption of restrictive laws—such as foreign aid flows, competitive elections, and the degree to which CSOs seek to topple the existing political system—should predict the adoption of different laws similarly as indicated by the variables’ sign, effect size, and significance. Likewise, factors believed to be negatively associated with the adoption of restrictive laws—such as higher levels of democracy—should also be consistent in their sign, size, and significance. Four conceptual equivalence hypotheses follow:

H2A

Higher levels of overseas development aid increase the probability of adopting restrictive laws, and that relationship is generally similar for all restrictive law types.

H2B

The context of electoral competition increases the probability of adopting restrictive laws, and that relationship is generally similar for all restrictive law types.

H2C

Higher levels of organized opposition by CSOs to the current political system increases the probability of adopting restrictive laws, and that relationship is generally similar for all restrictive law types.

H2D

Higher levels of democracy decrease the probability of adopting restrictive laws, and that relationship is generally similar for all restrictive law types.

The second set of conceptual hypotheses proposes the predictive power of individual factors varies across law types. At least four types of variations may exist that signal conceptual distinctness. First, a factor can increase the probability of adopting highly-restrictive laws, but the effect wanes for less restrictive types. Russia is believed to possess highly-restrictive CSO laws (Benevolenski and Toepler 2017; Toepler et al. 2019); therefore, a shared ideology with Russia—as measured by voting alignment in the UN General Assembly—should increase the probability of adopting highly-restrictive laws.

Second, a factor may decrease the probability of adopting highly-restrictive laws but also increase the probability of adopting minimally-restrictive ones. Although CSOs may lobby to dissuade lawmakers from passing laws that harm the sector’s organizational ecology, this relationship is likely too weak to identify in the data because lawmakers likely excluded CSOs when such laws are under consideration. CSOs may support minimally-restrictive laws, however, to establish “reasonable regulation” that promotes trust, accountability, and transparency. Therefore, greater participation by CSOs in the lawmaking process increases the probability of adopting minimally-restrictive.

Third, a factor may decrease the probability of adopting highly-restrictive laws, but the effect wanes for less restrictive types. The institutional arrangement of treaties and constitutions discussed in H1B is expected to exhibit this behavior because, according to international law’s three-part test, countries can prescribe by law actions in the interests of legitimate aims if such requirements are properly scoped and necessary for democracy. The presence of this particular institutional arrangement decreases the probability of adopting restrictive laws that do not meet the three-part test.

Finally, a factor may increase the probability of adopting minimally-restrictive laws, but be unrelated to the adoption of more restrictive types. The USA is believed to possess a legal approach that supports CSOs (Barber and Farwell 2017; Salamon and Toepler 1997), but its tax code contains provisions that may be characterized as minimally-restrictive. For example, its laws require CSOs to file tax forms if they wish to be formal tax-exempt organizations and 501(c)(3)s must pay taxes revenue earned through activities that are unrelated to their charitable mission. Therefore, shared ideology with the USA—as measured by voting alignment in the UN General Assembly—increases the probability of adopting minimally-restrictive laws. Four conceptual distinctness hypotheses follow:

H3A

Greater voting alignment with Russia in the UN General Assembly increases the probability of adopting highly-restrictive laws.

H3B

Greater participation by CSOs in the lawmaking process increases the probability of adopting minimally-restrictive laws.

H3C

An institutional context where ICCPR ratification exists and constitutional rules make treaties superior to ordinary legislation decreases the probability of adopting highly-restrictive laws, but this relationship wanes or vanishes for moderately and minimally-restrictive types.

H3D

Greater voting alignment with the USA in the UN General Assembly increases the probability of adopting minimally-restrictive laws.

Methods and Data

This study uses competing risk models (CRMs), which are a type of event history analysis and reveals considerably more information about social phenomena with multiple types of outcomes (for review see Allison 2014: 53–66; Box-Steffensmeier and Jones 2004: 155–182; Jones 1994). I use a logit model as my primary modeling strategy and use two additional methods for robustness checks. The first is a Cox proportional hazards model (Cox 1972, 1975). Preliminary testing showed the proportional hazards assumption holds for some laws but not others, which requires the Cox model to incorporate time-varying coefficients. The second robustness check concerns rare events. The data show that 37 of 138 countries adopted a law between 1993 and 2012. For some, these laws represent only 37 country-year “events” among nearly 2400 “non-events” (approximately 1.5%), making them rare events because there are “dozens to thousands of times fewer ones [than zeros]” (King and Zeng 2001: 138). When statistical software corrections are unavailable,Footnote 5 methodologists prescribe specific data collection and sampling strategies to minimize bias (King and Zeng 2001: 141–143).Footnote 6

Dependent Variables

The dependent variable is the adoption of a law that restricts foreign aid to CSOs. In the first analysis, the variable is coded as 0 and changes to 1 if countryi adopts the law at time t. This pooled approach is typical for most of the literature on this topic and is used here to test the institutional hypotheses (H1A and H1B). Three countries passed multiple laws during the analysis period,Footnote 7 but specifications here analyze only initial adoptions. While losing these country-year observations that occur before these subsequent adoptions is statistically inefficient, doing so allows for consistent estimation of factors that predict the adoption of initial laws without having to model the effects of initial adoptions on later adoptions.

In the CRM analyses, the coding of the dependent variable depends on the law type. Testing the conceptual hypotheses (H2A-D and H3A-D) requires recoding the dependent variable for each CRM. In these models, the content of adopted laws as described by Dupuy et al. (2016: Appendix Table 3) determine their values. Prohibitive laws are highly-restrictive and contain strong language regarding what organizations cannot do.Footnote 8 The coding protocol identified laws as prohibitive/highly-restrictive if the description of the legal restriction contains forms of the qualifier “prohibited.” The dependent variable equals 1 if a country adopts a law in that year and that law is coded as prohibitive, 0 if a country adopts a law in that year but the law is not prohibitive, and 0 if the country does not adopt a law in that year.

Coding for red-tape and notification laws follows the same protocol. Red-tape laws are moderately-restrictive and communicate ex ante conditions that organizations must meet before receiving funds.Footnote 9 The protocol coded laws as red-tape/moderately-restrictive if the description of the law included variations on terms “restrictions on,” “required to,” and “approval for.” This category also includes laws that require “government monitoring of contracts” (i.e., Ecuador 2011) because foreign funding is allowed given the ex ante condition. The variable equals 1 if a country adopts a law in that year and that law is a red-tape type and 0 otherwise. Notification laws are minimally-restrictive and contain instructions for what organizations must do after receiving foreign funding.Footnote 10 The protocol uses terms such as “notification,” “reporting,” and “taxation” to code notification/minimally-restrictive laws. The variable equals 1 if a country adopts a law in that year and that law is a notification type and 0 otherwise.

Independent Variables

Access to data for such a large number of countries is difficult. Thus, this analysis uses the Varieties of Democracy Project (V-Dem) that provides data for 201 countries from 1789 to 2018 (Coppedge et al. 2018), to include as many countries as possible in the analysis. Data are added from other sources as necessary. The UN Office of Legal Affairs and the Comparative Constitutions Project (CCP) provide data to test hypotheses H1A and H1B. The former provides information on whether and when a country ratifies the ICCPR. For each country-year observation, ICCPR ratified equals 1 if the country ratified the human rights treaty and 0 if it did not. The CCP provides constitutional texts for 214 independent countries from 1789 through 2013 (Elkins et al. 2014). Treaties Superior equals 1 for all constitutional systems that explicitly states international treaties are superior to ordinary legislation. The variable equals 0 if the constitution does not mention international treaties or gives them a status equal or inferior to ordinary legislation. Executive power measures the authority given to the country’s chief executive and follows the paper on the constitutional boundaries of executive lawmaking (Elkins et al. 2012, 2014). The variable ranges from 0 to 7 with higher values indicating more constitutional powers entrusted to the chief executive.Footnote 11 Analyses do not lag institutional variables because they frame the institutional context of lawmaking in the current year.

The models lag all control variables discussed below. Electoral competition measures whether elections to fill chief executive offices and the legislative body are characterized by uncertainty, meaning that the elections are, in principle, sufficiently free to enable the opposition to gain power (Coppedge et al. 2018: 299). The variable equals 1 when electoral competition exists. This operationalization differs from others that use the National Elections Across Democracy and Autocracy (NELDA) dataset (Hyde and Marinov 2012), which provides information on all national elections from 1945 to 2012. The NELDA dataset does not contain information for years that did not experience national elections. Thus, I use the V-Dem variable to achieve a consistent measure of electoral competition for election and non-election years.

The World Development Indicators (World Bank 2018) provide country-year data for population, GDP (constant 2010 US$), and the net official development assistance received (constant 2014 US$). Following the practice of prior scholarship, this analysis removes countries with GDP per capita exceeding $12,615 throughout the observation period (Dupuy et al. 2016: 9). Net ODA is divided by the total population to normalize ODA on a per capita basis. Zero and negative ODA per capita values are set to $0.01 before transforming the variable with a natural log function to achieve a normal distribution. Multiplying ln(Net ODA per capita) and electoral competition produces an interaction term that recent research finds relevant for explaining the adoption of restrictive laws (Dupuy et al. 2016).

The Yearbook of International Organizations (Union of International Associations) is often used to measure the vitality of civil society with a country. Those data only include information for intergovernmental and international non-governmental organizations. Scholarship finds both local and international non-state actors can affect politics in developing counties by merely providing the necessary infrastructure to develop civil society (Brown et al. 2008; Stremlau 1987), providing charitable service without political motivations (Brass 2012; Frantz 1987), and serving as a catalyst for policy change in an international system (Kajese 1987; Keck and Sikkink 1999). This suggests the role and character of CSOs, from both local and international origins, matter more than merely the number of non-state actors in attendance. Two variables control for the degree to which CSOs influence lawmaking. CSO routinely consulted measures the degree to which policymakers consult major CSOs with higher values representing more significant consultation (Coppedge et al. 2018: 176). CSOs are anti-system measures the level of organized opposition to the current political system with higher values representing stronger anti-system activity (Ibid, p. 178). Both variables were initially collected using ordinal intervals and then converted to a continuous interval using a Bayesian item response theory measurement model (Ibid.).

The Polity2 indicator represents the balance of autocratic and democratic authority in any particular regime context (Marshall et al. 2017: 17). The indicator is commonly used to control for regime type in cross-national analyses, to limit analyses to cases with specific profiles of democratic quality, or proxy variables of theoretical interest such as institutional effectiveness (Baldwin et al. 2019; Bauerle Danzman et al. 2017; Hellwig and Samuels 2008). Unfortunately, the data are only available for countries whose populations exceeded 5000,000 in 2006 (Marshall et al. 2017). This population cutoff omits small countries such as Belize which adopted its law in 2003 but only had a population of approximately 300,000 in 2006. In this analysis, I use Imputed FH/Polity2 to control for regime type for as many countries as possible. The variable uses Freedom House Political Rights and Civil Liberties values, and the original Polity2 variable to impute values for countries where Polity data are missing (Coppedge et al. 2018: 290). The scale ranges from the least democratic (0) to most democratic (10) and is shown to perform better in terms of validity and reliability than its component indicators (Hadenius and Teorell 2005).

Researchers use UN voting behavior in numerous ways, including constructing measures of policy preference similarity between two states (Chapman 2009) and testing whether foreign aid is a reward for UN votes (Morgan 2018). I use votes in the UN’s multi-dimensional issue space to triangulate a measure of governmental ideology that is comparable across time. UN votes with the USA (%) and UN votes with Russia (%) are continuous variables of all votes during a particular session. The variables range from 0 to 100 with higher values indicating greater alignment with the superpower (Bailey et al. 2017; Voeten 2013). The correlation between these variables is approximately − 0.46. A third variable, Regional Diffusion, controls for policy diffusion through a normative pressure process. The variable represents the percentage of states within a country’s World Development Indicators regional group (World Bank 2018) that adopted the law type studied as the outcome.

The Political Terror Scale (PTS) provides data for PTS average, which measures local political terror and unrest (Gibney et al. 2017). PTS provides three separate indicators coded from annual human rights reports published by Amnesty International, Human Rights Watch, and the US Department of State. Each variable is measured on a 5-point scale with higher values indicating higher levels of abuse and physical integrity rights violations. In this analysis, the control variable averages all PTS scores available for each country in the given year.

Research shows states’ decisions to violate human rights are negatively related to their judicial effectiveness, which is the primary enforcement mechanism of legal obligations at home and abroad (Powell and Staton 2009). Rule of Law Index measures the degree to which laws are fairly enforced and to what extent the actions of governmental officials comply with the law. The index is a latent variable that uses a Bayesian factor analysis of 15 indicators (Coppedge et al. 2018: 235–236). Methodologists show the latent variable index is superior to using a single indicator or averaging several measures (Linzer and Staton 2015). The variable Time represents the number of years a country has gone without adopting a law since entering the dataset. To maintain a consistent sample, countries always leave the dataset the year they adopt any law. But the value of the dependent variable for those countries varies according to the competing risk model used. The Appendix contains a table with descriptive statistics for all variables (Table 5). The top panel summarizes dependent variables for the 138 cases, while the bottom panel summarizes all variables for the 2398 country-year observations.

Results

This section presents the results in the same order as the hypotheses. The primary modeling strategy and its robustness checks show that the institutional arrangement that minimizes the adoption of restrictive laws is one where the constitutional rules explicitly place ICCPR ratification above ordinary legislation. Competing risk models find that the factors that predict the adoption of one law type do not necessarily predict the adoption of a different type, which suggests laws are more distinct than equivalent.

The discussion of results follows recommended practice and uses marginal effects at the means (AMEs) to summarize practical implications of independent and linked interaction variables (Amrhein et al. 2019a, b; Cameron and Trivedi 2005; Greenland 2017; Hanmer and Ozan Kalkan 2013; Long 1997; Wasserstein et al. 2019). I do this because logit models produce coefficients that are difficult to interpret or depend on specific values that are inconsistent across cases.Footnote 12 AMEs are direct, interpretable measures that compute the marginal change of the factor across all cases in the sample and then calculates an average size of the effect in the sample (Long and Freese 2014). The implication for the reader is that numeric values discussed may not appear in the accompanying regression tables; therefore, I report p values of two-tailed hypothesis tests for all AMEs to allow readers to judge significance for themselves.

Results of Institutional Hypotheses Using Event History Analysis

Table 2 shows the results of an event history analysis using logistical regression to test the institutional hypotheses: first, that the ratification of the ICCPR decreases the probability of adopting restrictive laws (H1A). Second, that the institutional arrangement that minimizes the probability of adoption is one where ICCPR ratification exists alongside constitutional rules that make international treaties superior to ordinary legislation (H1B). The baseline specification (model 1) suggests contexts with higher values of electoral competition and increased voting alignment with Russia increase the probability of adopting restrictive laws, whereas higher levels of democracy decrease the probability of adoption. Model 2 introduces ICCPR ratification as an independent variable and executive power as a control variable. The data suggest that ratification has no apparent relationship with adopting restrictive laws. This finding is contrary to the first institutional hypothesis that predicts ratification and decreases the probability of adoption (H1A). However, model 2 omits constitutional rules that condition the status of international treaties. Subsequent models (3–5) analyze constitutional differences in various ways. These latter models show that constitutional rules that treat international treaties as superior to ordinary legislation decrease the predicted probability of adopting a restrictive law.

Table 2 Pooled event history analysis (EHA) with logistic regression

The interaction term specified in model 5 tests the second institutional hypothesis (H2A) that argues the institutional arrangement that minimizes the probability a country adopts a restrictive law is one where constitutional rules elevate ICCPR ratification commitments above ordinary legislation. The interaction term and the main effect have a strong relationship with the predicted outcome. Given ICCPR ratification by the average country in the sample, a discrete change in its constitutional rules to privilege international commitments decreases the probability of adoption by one-half a percentage point (− 0.005, p = 0.01). Though the magnitude might seem small, the effect is larger than both a standard deviation increase in the level of democracy within a country (− 0.002, p < 0.01) and entering a context of electoral competition (0.003, p = 0.08).

The data show the effect of this institutional arrangement varies by context. For an otherwise average country with ICCPR ratification, the average effect of constitutional rules that privilege international commitments is 2.5 percentage points larger when states are a standard deviation below the mean level of democracy (− 0.026, p = 0.01) than when states are one standard deviation above the mean (− 0.001, p = 0.08); the difference is significant at the 0.03 level. For the average country, the average effect of this institutional arrangement is 1 percentage point larger when political competition is present (− 0.013, p =0.01) than when absent (− 0.003, p =0.03); the difference is significant at the 0.07 level.

Figure 4 shows the effect of preexisting institutions in different contexts. For the average country, possessing constitutional rules that make international commitments superior to ordinary legislation generally decreases the probability of adopting restrictive laws for all levels of executive power (Fig. 4, left panel) and net ODA per capita (Fig. 4, right panel). Comparing the effect when executive power is low (2) versus high (6) in an otherwise average country, the data suggest the decrease in predicted probability is smaller for lower levels of executive power (− 0.004, p =0.01) than higher ones (− 0.012, p =0.03), and the difference is significant at the 0.07 level. However, the average effect of preexisting institutions is only slightly smaller when the natural log of net ODA per capita is 0 ($1 per capita) compared to 6 ($403 per capita), but the difference is not significant (p = 0.78).

Fig. 4
figure4

Effect of constitutional rules on adopting restrictive laws. Figure shows the average marginal effect (with 95% confidence intervals) of the discrete change of constitutional rules that elevate international treaties above ordinary legislation. Thin gray lines show the distribution of observations across the data space using kernel density (executive power, left panel) and a rug plot (ODA per capita, right panel)

These results support the second institutional hypothesis (H1B) that the institutional arrangement that minimizes the probability a country adopts a restrictive law is one where ICCPR ratification exists alongside constitutional rules that elevate international treaties above ordinary legislation. Robustness checks support these findings. The average effect size for this institutional arrangement varies across context, but its impact appears strongest in settings described as undemocratic, politically competitive, or possessing a strong constitutional executive.

Results of Conceptual Hypotheses Using Competing Risk Models (CRMs)

CRMs indicate whether the risk factors that predict the adoption of laws vary across different law types. Table 3 shows the results of the logit primary modeling strategy, and the appendix contains the full regression tables (Tables 7, 8, 9). Models recode dependent variables, but the specification and sample remain the same. “Pooled laws” (model 1) presumes laws are the same and analyzes them as a monolithic group. Models 2–4 disaggregate this monolithic group to assess equivalence (hypotheses H2A-D) and distinctness (hypotheses H3A-D). The “only prohibitive laws” model shows the factors that predict the adoption of highly-restrictive laws containing strong language regarding what organizations cannot do. In column three, the “only red-tape laws” model identifies the factors that predict the adoption of moderately-restrictive laws that erect ex ante conditions for what organizations must do before receiving foreign funding. Finally, “only notification laws” reports factors associated with the adoption of minimally-restrictive laws that prescribe actions that organizations must take after receiving foreign funding.

Table 3 Competing risk model (CRM) with logistic Regression

In comparing the pooled logit results to the competing risks estimates, it is clear that the two approaches produce different results regarding the relationship a factor has on the adoption of law types. In general, the coefficients’ signs in the pooled model are the same as those in the CRMs. The pooled results cannot differentiate among law types and so represent an “average” or “dampened” effect. The CRMs show more refined estimates on how a type-specific covariate relates to a particular law type. The pooled model produces false positives (type I errors) or false negatives (type II errors) when coefficients are type-specific hazards that correlate with the adoption of only one law type. The CRMs’ nuanced results identify these spurious results. For example, the pooled model’s emphasis on electoral competition is a false positive because the type-specific hazard positively correlates with the adoption of red-tape laws only (p < 0.05). The data show no relationship for the adoption of other law types. The pooled model’s rejection of CSO consultation’s significance is a false negative because the coefficient positively correlates with the adoption of notification laws (p < 0.01). I continue this exercise to assess the conceptual equivalence and distinctness of law types.

Assessing Conceptual Equivalence

Table 4 organizes data to evaluate the equivalence and distinctness of law types. The top panel contains the conceptual equivalence hypotheses (H2A-D), and the bottom includes the conceptual distinctness hypotheses (H3A-D). Each row includes the hypothesis followed by the average marginal effect of a discrete change in the factor for each law type. The Appendix contains an expanded table that combines this information with data from the robustness checks (Table 10).

Table 4 Assessing conceptual equivalence and distinctness

If factors predict adoption of different law types in generally consistent ways, then such patterns signal laws are conceptually equivalent and are perhaps best analyzed as a monolithic group. If this is the case, factors associated with the adoption of restrictive laws should predict the adoption of different law types in similar ways as indicated by the variables’ sign, effect size, and significance. The data do not support the hypothesis that ODA per capita is a significant, positive, and consistent predictor of restrictive laws (H2A). The analyses found no evidence suggesting a statistical relationship between a standard deviation increase in ODA per capita and the adoption of any law type. Robustness checks confirm this finding.

The data partially support the hypothesis that the context of electoral competition increases the probability of adopting restrictive laws. Although the factor has a statistically significant coefficient in all pooled models, CRMs show the presence of electoral competition robustly predicts the adoption of moderately-restrictive, red-tape laws only. For the average country in the sample, a context of electoral competition increases the probability of adopting red-tape laws by less than 1 percentage points in the primary modeling strategy (0.002, p =0.07) and almost 1 percentage point in the rare events robustness check (0.009, p = 0.14). Electoral competition weakly predicts the adoption of highly and minimally-restrictive laws. Hypothesis H2B is unsupported because electoral competition is not a significant, positive, and consistent predictor of all restrictive laws.

The data do not support the claim that organized opposition by CSOs to the current political system is a significant, positive, and consistent predictor of restrictive laws. The CRMs find no relationship between a standard deviation increase in CSOs’ organized opposition to the current political system and the adoption of any restrictive law. Therefore, H2C is not supported.

The data find evidence that higher levels of democracy decrease the probability of adopting restrictive laws. The factor is generally significant and always negative. However, the CRMs show that for the average country, the change in the predicted probability caused by a standard deviation increase in democracy is consistent for only two types of laws in only one modeling strategy: in the primary modeling strategy, a discrete change decreases the predicted probabilities of adopting red-tape (− 0.001, p < 0.01) and notification (− 0.001, p < 0.10) laws by the same amount. The Cox model robustness check shows the average marginal effect of a positive discrete change is insignificant across the types of laws. This statistical insignificance may be because the democracy–adoption relationship varies over time. According to the Cox model with time-varying coefficients (Table 8), the level of democracy has a significant correlation with highly-restrictive, prohibitive laws only. The factor’s main effect is negative (p < 0.05) and its time-varying component is positive (p < 0.05), which suggests the impact of level of democracy as an adoption deterrent declines over time.Footnote 13 This time-varying relationship is inconsistent across law types, however. The hypothesis that higher levels of democracy decrease the probability of adopting restrictive laws in a manner that is consistent for all types of restrictive laws (H2D) is unsupported.

Assessing Conceptual Distinctness

If the predictive power of individual factors varies across law types, then the laws are conceptually distinct and best analyzed in a disaggregated manner. The pooled models suggest greater voting alignment with Russia in the UN General Assembly increases the probability of adopting restrictive laws. This relationship appears unique to only one law type, but not the type predicted. For the average country, a positive discrete change in voting alignment with Russia produces an average marginal effect that is significant for predicting minimally-restrictive, notification laws only. The Cox model robustness check shows the average marginal effect of a positive discrete change is insignificant across law types. Like other factors, this statistical insignificance may be because the factor’s effect varies over time. According to the Cox model with time-varying coefficients (Table 8), voting alignment with Russia correlates with highly-restrictive, prohibitive laws only. The factor’s main effect is positive (p < 0.01) and its time-varying component is negative (p < 0.01), which suggests the impact of voting alignment with Russia as an adoption propellant declines over time.Footnote 14 Together these findings partially support the hypothesis that increased voting alignment with Russia increases the probability of adopting only one law type (H3A), but the hypothesis incorrectly predicted the law type.

Data show that for the average country, a positive discrete change in CSO participation produces an average marginal effect that is significant for predicting minimally-restrictive, notification laws. Depending on the modeling strategy, a discrete change increases the predicted probabilities of adopting notification laws by less than one percentage point in two of the three modeling strategies (p < 0.05). This relationship suggests increased CSO participation in lawmaking may coincide with a push for reasonable regulation from within the sector. Notably, the same discrete change produces an average marginal effect that decreases the predicted probability of adopting prohibitive laws. That relationship conforms to theory but lacks statistical evidence (p > 0.20). In support of hypothesis H3B, the data show increased participation by CSOs in the lawmaking process increases the probability of adopting notification laws only.

The data find evidence that ICCPR ratification combined with constitutional rules that make international treaties superior to ordinary legislation decreases the probability of adopting restrictive laws. Table 4 shows this institution–adoption relationship is not unique to any one law type. Contrary to hypothesis H3C, the effect is not strongest for highly-restrictive laws and then wanes for less restrictive ones. Instead, the data show that in the average country, the preexisting institution decreases the predicted probability of adopting moderately-restrictive, red-tape laws by less than one percentage point in two of the three modeling strategies (p < 0.05, Table 10). This institutional context has smaller effects for prohibitive and notification laws, but these relationships have weaker statistical evidence in those models.

For the final hypothesis, the data do not support the claim that greater voting alignment with the USA in the UN General Assembly decreases the probability of adopting highly-restrictive, prohibitive laws. Instead, the analysis finds no relationship, and H3D is not supported.

The above eight hypotheses carefully examined factors’ type-specific relationships to assess whether qualitatively different laws are conceptually equivalent or distinct. A signal of conceptual equivalence occurs when factors predict adoption in ways generally consistent across law types. Such patterns would imply—but not prove—conceptual equivalence and support pooling various law types into a monolithic group. Of the four factors examined, only electoral competition and level of democracy showed robust signs of a relationship with adoption. The data show the relationship between these two factors and adoption varies in effect size—as measured by both regression coefficients and average marginal effects—and significance across law types. Thus, the analysis is unable to produce strong evidence supporting the argument that restrictive laws are conceptually equivalent.

Not only do these null results fail to demonstrate restrictive laws are a monolithic group, but the findings provide additional support to the argument that restrictive laws are conceptually distinct. The case for conceptual distinctness relies on the relationship between individual factors and adoption to vary across law types. Several factors portray this behavior and deserve mention: Electoral competition, voting alignment with Russia, and CSO participation in lawmaking have robust relationships with only one type of law. These relationships are consistent according to regression coefficients, robustness checks, and average marginal effects.

The remaining variables—level of democracy and preexisting institutions—predict the adoption of multiple law types with varying degrees of success. Looking at the factors’ regression coefficients is the minimal approach. Here, while holding all else constant, the factors predict adoption for red-tape and notification laws reasonably well, but only preexisting institutions robustly predict adoption of prohibitive laws. A stricter approach uses average marginal effects. This shows that for the average country, the average effect of a discrete change in democracy or preexisting institutions robustly predicts one and perhaps two law types depending on the significance threshold used. In total, the evidence from five of the eight factors examined implies the laws that restrict foreign aid to CSOs are conceptually distinct and should be theorized, conceptualized, and analyzed accordingly.

Discussion

This analysis makes two original contributions. First, the study incorporates institutions in two ways: first, it looks to the past and countries’ ratification of the ICCPR as a preexisting institution. Second, it looks up to constitution-level rules to understand the degree to which international commitments constrain lawmaking. The data show the institutional arrangement that minimizes the probability a country adopts a restrictive law is one where ICCPR ratification exists alongside constitutional rules that elevates international treaties above ordinary legislation. This critical finding calls for a turn to history when analyzing the institutional development of the laws and policies that regulate civil society around the world. Boring into the conceptualization of these laws is the second original contribution. Previous work tends to analyze the laws that restrict CSOs as a monolithic group and has not yet conceptualized them as distinct law types. My second analysis demonstrated that disaggregating law types reveals counterintuitive and overlooked relationships. For example, the positive relationships that voting alignment with Russia and CSO consultation have with increasing the probability of adopting minimally-restrictive, notification laws. The mechanisms behind these associations beyond the scope of this research but deserve discussion here and investigation in the future.

To begin, as a government’s decisions on international matters align with Russia’s, analysts might expect a higher probability that the country passes harsh, Putin-style laws as a result of diffusion through learning or emulation. The analysis shows that this is not the case. Instead, the data show a positive correlation between voting alignment with Russia and the adoption of minimally-restrictive, notification laws. This may seem like a peculiar finding, but a focus on institutions offers two possible explanations. One the one hand, CSO regulatory regimes are political institutions believed to change incrementally over time. Thus, the adoption of minimally-restrictive laws is not necessarily an endpoint but may instead be a waypoint or entry point toward a more “bureaucratically illiberal” regulatory regime that uses laws and policies to raise the transaction costs for CSOs to emerge and operate (DeMattee 2019: 10). This institutional explanation emphasizes that laws accumulate and unless their language is precise, governments may enforce them widely and differentially.

On the other hand, describing the Russian case as entirely closed to civil society may be a mischaracterization of its institutionalized legal approach. Such a conclusion fails to account for the complicated relationship between formal legal rules and social norms to produce what Cole (2017: 6) and others call “working rules.” While laws may closely resemble working rules in some contexts, in other settings informal rules and social norms amend legal rules to produce working rules. A regulatory regime may enforce one set of friendlier working rules for apolitical CSOs that provide services complementary or supplementary to government, while using a harsher set of working rules to crackdown on CSOs that are politically inconvenient (for a recent example of the Burundian case see Popplewell 2018). This explanation matches work discussing “the Russian government’s divergent positions towards civil society” (Benevolenski and Toepler 2017: 64; Salamon et al. 2015; Toepler et al. 2019). This is also congruent with research on dictators’ decision-making studied as both a trade-off between suppressing rivals and delivering public services (Woldense 2018), and the rational use of their tools of office to gather information that prolongs their stay in power and maximizes their gains from officeFootnote 15 (Malesky and Schuler 2011; Sartori 1976; Smyth and Turovsky 2018; Wintrobe 1998). Forthcoming work on the Russian case suggests this is a possibility and identifies CSO laws are one of several strategies regimes use to gather information about political opponents’ public support and resources (Smyth 2019: 17–21).

An entirely different set of explanations seem to drive the positive relationship between the adoption of minimally-restrictive laws and consultation with CSOs. The magnitude of this relationship is substantial: a standard deviation increase in CSO consultation is roughly twice the size of the same discrete change in voting alignment with Russia. Not only does the effect size deserve attention, but the sign on the coefficient itself challenges received wisdom. Why would a political opportunity structure open to CSOs also be one that, ceteris paribus, propels the adoption of restrictive laws? One explanation is that CSOs successfully pull lawmakers away from highly or moderately-restrictive bills and negotiate agreement around minimally-restrictive laws. Although supported by theory, this explanation lacks supporting evidence in these data because neither the prohibitive nor red-tape models show CSO consultation to be negative and significant.

An alternative account asserts these minimally-restrictive, notification laws might not be as undesirable as analysts interpret them to be. Notification laws may be attempts to minimize patterns of wrongdoing within the sector (Gibelman and Gelman 2004) or correct the dualism foreign funding causes among domestic CSOs (Chahim and Prakash 2013; Pallas et al. 2018: 259). Perhaps these notification laws are a type of “reasonable regulation” that CSOs support to enjoy the privileges that accompany registration as a nonprofit or religious legal entity. And given the ballooning of development NGOs beginning in the early 1990s (Anheier and Salamon 1998; Cammett and MacLean 2014; Reimann 2006; Schnable 2015), CSOs’ support for minimally-restrictive laws may be an attempt to cull ‘briefcase NGOs’ from the sector while at the same time earn regulatory legitimacy (Popplewell 2018) and show themselves as transparent and accountable to citizens, donors, and governments (Arhin et al. 2018).

Conclusion

Scholars offer many reasons why countries adopt laws that restrict foreign funding to CSOs. Despite the growing body of work on this topic, scholars have not studied the institutional context of lawmaking, nor has research investigated how factors’ relationships with adoption vary across law types. This paper has explored these understudied areas to explain the conditions that propel or deter the adoption of laws that restrict CSOs. It has made both theoretical and conceptual contributions in the process.

The first analysis introduced institutional variables to further our understanding of the factors that predict the adoption of restrictive laws. The argument focused on preexisting institutions to explain the changing constraints lawmakers face. Ratification of the International Covenant on Civil and Political Rights (ICCPR), which legal scholars consider the principal treaty guarding civil and political rights (Donnelly 2013; Hathaway 2002; Henkin 2000; ICNL 2009, 2015; Kiai 2012), should have a strong and negative association with the adoption of laws that impinge on voluntary association. Yet, the analyses here show that ICCPR ratification is not a sufficient condition. Incorporating constitutional rules into the analysis explains why ratification of the ICCPR does not itself stifle efforts to adopt restrictive laws. Constitutions condition the degree to which international commitments constrain lawmaking. Thus, the effect of ICCPR ratification as a preexisting institution is strongest when constitutional rules give international treaties a status that is superior to ordinary legislation. Although the average effect size for this institutional arrangement varies across context, its impact appears strongest in settings described as undemocratic, politically competitive, or possessing a strong constitutional executive.

Next, while researchers and practitioners discuss many types of laws that regulate CSOs, most studies analyze them as a monolithic group. The second analysis showed several benefits to disaggregating law types. First, analyses that pool laws may unknowingly overstate factors’ relevance in adopting restrictive laws. The pooled models identified several factors thought to propel or deter the adoption of restrictive laws. However, after disaggregating laws, the analyses showed these factors robustly predicted only one law type: electoral competition positively correlates with the adoption of moderately-restrictive laws, higher levels of democracy negatively correlates with moderately-restrictive laws, and stronger voting alignment with Russia in the UN positively correlates with minimally-restrictive laws. Pooling laws also risk understating or overlooking the relevance of a particular factor. The pooled models rejected the importance of CSO consultation in all instances. Only after conceptualizing laws as distinct did we identify nuanced relationships, such as the positive correlation of greater participation by CSOs in the lawmaking process with the adoption of minimally-restrictive laws.

These findings call on researchers to give increased attention to institutions and conceptualizations when studying the adoption of laws that restrict CSOs. Constitutions and history matter and analysts should consider these and other important factors when studying the adoption of laws that restrict CSOs. Findings also urge researchers not to oversimplify or ignore the differences among laws or else risk overstating or overlooking key relationships. Analysts should instead embrace legal differences and use them as analytical leverage in empirical research and theory building. Going forward, research studying the laws, policies, and regulations of civil society should pay increased attention to political institutions, preexisting laws and policies, and fundamental policy differences.

Notes

  1. 1.

    Belize and Vietnam are not analyzed because they are not coded in several datasets. These countries adopted laws in 2003 and 2009, respectively.

  2. 2.

    The ICCPR protects freedoms of expression and belief (Articles 18, 19, 27), rights to associate and organize (Articles 1, 18, 21, 22), rule of law and human rights (Articles 6, 7, 9, 14, 15, 16, 17, 25, 26), and personal autonomy and economic rights (Articles 1, 3, 8, 12, 22, 23, 25).

  3. 3.

    Costa Rica was the first to ratify (November 29, 1968), and Fiji was the 172nd and most recent (August 16, 2018).

  4. 4.

    Policy differences have theoretical implications for Kingdon’s politics and policies streams of the Multiple Stream Approach (Kingdon 1984) where various policies compete in the policy stream, and only proposals that successfully match the national mood, and the politics of policymaking get considered during the policy window (Zahariadis 2014). Through the theoretical lens of Punctuated Equilibrium Theory (Baumgartner and Jones 1991, 1993), strong, established interests in the policy subsystem make adoption of substantial policy changes less likely. Only in disequilibrium are entrenched players unable to railroad massive policy changes. In the Advocacy Coalitions Framework (Sabatier 1988; Sabatier and Jenkins-Smith 1993; Sabatier and Weible 2007), policies that are an affront to a coalition’s belief systems are met with stiff resistance, whereas minor changes may be the result of a cross-coalitional learning or the negotiated outcome of the dialogue with policymakers. While the above theories predict differences in policies affect policymaking, policy differences are shown to have different effects on public and elite opinion and the social construction of target groups (Ingram and Schneider 1990, 1991; Schneider et al. 2014) and affect the extent to which targeted groups participate in politics (MacLean 2011; Mettler and Soss 2004; Pierson 1993).

  5. 5.

    Statistical software corrections—such as relogit or firthlogit in Stata 15—for analyzing rare events with logistic regressions are not yet available for panel data or analyze requiring clustered standard errors.

  6. 6.

    This involves first collecting all the “events” and an equal number of randomly selected “non-events,” and continuing to add randomly sampled non-events and stop when the confidence intervals are sufficiently small for the substantive purposes at hand. In the rare events analysis, countries that adopt laws appear in all analyses (482 country-year observations).

  7. 7.

    Belarus (2001, 2003); Indonesia (2004, 2008); and Uzbekistan (2003, 2004).

  8. 8.

    The variable takes the value of 1 if at least one of the following nine provisions exists: “certain organizations are prohibited from receiving foreign funding”; “certain types of organizations are prohibited from receiving foreign funding”; “foreign-funded organizations prohibited from carrying out particular activities”; “foreign funding can be used only for certain purposes”; “foreign funding prohibited”; “foreign funding prohibited for certain activities”; “foreign-funded NGOs prohibited from working on certain issue areas”; “foreign-funded organizations prohibited from carrying out particular activities”; and “use of foreign funding prohibited for particular activities.”

  9. 9.

    The variable equals 1 if at least one of the following twelve provisions exists: “government approval for foreign funding”; “government approval required for particular uses of foreign”; “government may cap the amount”; “government monitoring of NGO contracts financed with foreign funding”; “government restrictions on use and source”; “government restrictions on whether foreign funding can be received”; “other restrictions on use of foreign funding”; “requirements for how organizations can receive foreign funding”; “restrictions on certain types of organizations receiving foreign funding”; “restrictions on receipt and use of foreign funding”; “restrictions on sources from which foreign funding can be acquired”; and “restrictions on use of foreign funding.”

  10. 10.

    The variable equals 1 if at least one of the following six provisions exists: “foreign funds are taxed”; “government notification of foreign funding required”; “organizations must report source of revenues”; “reporting and accounting requirements”; “reporting and accounting requirements for foreign funding”; and “reporting requirements.”

  11. 11.

    Operationally, the additive index increases by 1 for each of the following binary variables present in the constitutional system as identified by CCP: (1) power to initiate legislation (coded 1 if head of state, head of government, or government can initiate legislation); (2) power to issue decrees (coded 1 if head of state or head of government can issue decrees); (3) power to declare emergencies (coded 1 if head of state, head of government, or government can declare emergencies); (4) power to propose amendments (coded 1 if head of state, head of government, or government can propose amendments to the constitution); (5) power veto legislation (coded 0 if no vetoes are possible or can be overridden by a plurality or majority in the legislature; coded 1 if vetoes are possible but require at least 3/5 supermajority of the legislature to override veto); (6) power to challenge the constitutionality of legislation (coded 1 if head of state, head of government, or government can challenge the constitutionality of legislation); (vii) power to dissolve the legislature (coded 1 if head of state, head of government, or government can dissolve the legislature).

  12. 12.

    The logit models used here produce coefficients that represent the direction of the variable’s effect on the probability of adoption but are difficult to interpret or odds ratios whose substantive meanings depends on the specific value of the odds before they change (Long and Freese 2014: 228–235).

  13. 13.

    According to the Cox model with time-varying coefficients (Table 8), the level of democracy has a negative sign for all types of laws but is statistically significant for highly-restrictive laws only (− 3.05, p < 0.05). A standard deviation increase in the level of democracy (approximately 2.85 points on a 10-point scale) at the beginning of the observation period, while all other variables are held constant, yields a hazard ratio equal to exp(− 3.05*2.85) = 0.0001. Thus, the rate of adoption decreases by (100% − 0.01%) = 99.99% with a standard deviation increase in democracy at the beginning of the observation period. The time-varying component of the level of democracy has a positive sign for all types of laws but is statistically significant for highly-restrictive laws only (0.18, p < 0.05). This suggests the level of democracy as a deterrent for adopting highly-restrictive declines with every unit of time. A standard deviation increase in the level of democracy at year 10 of the observation period, while all other variables are held constant, yields a hazard ratio equal to exp(− 3.05*2.85 + 0.18*2.85*10) = 0.0283. Thus, the rate of adoption decreases by (100%-2.89%) = 97.11%. While holding all other variables constant, this discrete change at year 20 yields a hazard ratio equal to exp(− 3.05*2.85 + 0.18*2.85*20) = 4.79, which increases the rate of adoption by (479% − 100%) = 379%. Holding all else equal, a positive and discrete change in democracy causes a decrease in the rate of adoption by 99.99% at the beginning of the observation period, 97% in year 10, and increases the rate of adoption by over 350% in year 20.

  14. 14.

    According to the Cox model with time-varying coefficients (Table 8), voting alignment with Russia has a positive and statistically significant relationship with highly-restrictive laws (0.22, p < 0.01). A standard deviation increase in the voting alignment with Russia (approximately 11.75 points on a 100-point scale) at the beginning of the observation period, while all other variables are held constant, yields a hazard ratio equal to exp(0.22*11.75) = 13.26. Thus, the rate of adoption increases by 1226% with a standard deviation increase in voting alignment with Russia at the beginning of the observation period. The time-varying component has a negative and statistically significant relationship for highly-restrictive laws only (0.18, p < 0.05). This suggests the voting alignment with Russia is a propellant for adoption for highly-restrictive laws declines with every unit of time. A standard deviation in the factor, while all other variables are held constant, yields a hazard ratio equal to exp(0.22*11.75 + − 0.02*11.75*10) = 1.264. Thus, the rate of adoption increases by 26.4%. While holding all other variables constant, this discrete change at year 20 yields a hazard ratio equal to exp(0.22*11.75 + − 0.02*11.75*20) = 0.1206, which decreases the rate of adoption by (100% − 12.06%) = 87.93%. Holding all else equal, a positive and discrete change in voting alignment with Russia increases the rate of adoption by more than 1000% at the beginning of the observation period, 26% at year 10, and decreases the rate of adoption by 88% at year 20.

  15. 15.

    The dictators’ dilemma of comparative politics stresses undemocratic leaders to use their tools of office—such as elections, laws, and policy implementation—to gather information that prolongs their stay in power and maximizes their gains from office.

References

  1. African Charter on Human and Peoples’ Rights (ACHPR) 1981 Organization of African Unity.

  2. Aligica, P. D. (2018). Public entrepreneurship, citizenship, and self-governance. Cambridge: Cambridge University Press.

    Google Scholar 

  3. Allison, P. D. (2014). Event history and survival analysis (2nd ed.). Los Angeles: SAGE.

    Google Scholar 

  4. American Convention on Human RightsPact of San Jose, Costa Rica” (B-32) (ACHR) 1969 Organization of American States.

  5. Amnesty International. (2019). Laws designed to silence: The global crackdown on civil society organizations. Online.

  6. Amrhein, V., Greenland, S., & McShane, B. (2019a). Scientists rise up against statistical significance. Nature,567, 305–307.

    Google Scholar 

  7. Amrhein, V., Trafimow, D., & Greenland, S. (2019b). Inferential statistics as descriptive statistics: There is no replication crisis if we don’t expect replication. The American Statistician,73(sup1), 262–270.

    Google Scholar 

  8. Anheier, H. K., & Salamon, L. M. (1998). The nonprofit sector in the developing world: A comparative analysis. Manchester: Manchester University Press.

    Google Scholar 

  9. Anheier, H. K., & Toepler, S. (2019). Civil society and the G20: Towards a review fo regulatory models and approaches. Retrieved from https://t20japan.org/wp-content/uploads/2019/03/t20-japan-tf6-10-civil-society-g20.pdf.

  10. Arhin, A. A., Kumi, E., & Adam, M.-A. S. (2018). Facing the bullet? Non-governmental organisations (NGOs’) responses to the changing aid landscape in Ghana. VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations,29(2), 348–360.

    Google Scholar 

  11. Bailey, M. A., Strezhnev, A., & Voeten, E. (2017). Estimating dynamic state preferences from United Nations voting data. Journal of Conflict Resolution,61(2), 430–456.

    Google Scholar 

  12. Baldwin, E., Carley, S., & Nicholson-Crotty, S. (2019). Why do countries emulate each others’ policies? A global study of renewable energy policy diffusion. World Development,120, 29–45.

    Google Scholar 

  13. Barber, P., & Farwell, M. M. (2017). The Relationships between State and Nonstate Interventions in Charitable Solicitation Law in the United States. In O. B. Breen, A. Dunn, & M. Sidel (Eds.), Regulatory waves: Comparative perspectives on state regulation and self-regulation policies in the nonprofit sector (pp. 199–220). Cambridge: Cambridge University Press.

    Google Scholar 

  14. Bauerle Danzman, S., Winecoff, W. K., & Oatley, T. (2017). All crises are global: Capital cycles in an imbalanced international political economy. International Studies Quarterly,61(4), 907–923.

    Google Scholar 

  15. Baumgartner, F. R., & Jones, B. D. (1991). Agenda dynamics and policy subsystems. The Journal of Politics,53(04), 1044–1074.

    Google Scholar 

  16. Baumgartner, F. R., & Jones, B. D. (1993). Agendas and instability in American politics. Chicago, IL: University of Chicago Press.

    Google Scholar 

  17. Benevolenski, V. B., & Toepler, S. (2017). Modernising social service delivery in Russia: Evolving government support for non-profit organisations. Development in Practice,27(1), 64–76.

    Google Scholar 

  18. Bloodgood, E. A., Tremblay-Boire, J., & Prakash, A. (2014). National styles of NGO regulation. Nonprofit and Voluntary Sector Quarterly,43(4), 716–736.

    Google Scholar 

  19. Boettke, P. J., & Candela, R. A. (2019). Productive specialization, peaceful cooperation and the problem of the predatory state: Lessons from comparative historical political economy. Public Choice. https://doi.org/10.1007/s11127-019-00657-9.

    Article  Google Scholar 

  20. Box-Steffensmeier, J. M., & Jones, B. S. (2004). Event history modeling: A guide for social scientists. Cambridge: Cambridge University Press.

    Google Scholar 

  21. Brass, J. N. (2012). Why do NGOs go where they go? Evidence from Kenya. World Development,40(2), 387–401.

    Google Scholar 

  22. Brass, J. N. (2016). Allies or adversaries? NGOs and the state in Africa. Cambridge: Cambridge University Press.

    Google Scholar 

  23. Brass, J. N., Longhofer, W., Robinson, R. S., & Schnable, A. (2018). NGOs and international development: A review of thirty-five years of scholarship. World Development,112, 136–149.

    Google Scholar 

  24. Breen, O. B., Dunn, A., & Sidel, M. (Eds.). (2017). Regulatory waves: Comparative perspectives on state regulation and self-regulation policies in the nonprofit sector. Cambridge: Cambridge University Press.

    Google Scholar 

  25. Breen, O. B., Dunn, A., & Sidel, M. (2019). Riding the regulatory wave: Reflections on recent explorations of the statutory and nonstatutory nonprofit regulatory cycles in 16 jurisdictions. Nonprofit and Voluntary Sector Quarterly, 48(4), 691–715.

    Google Scholar 

  26. Brennan, G., & Buchanan, J. M. (1985). The reason of rules: Constitutional political economy. Cambridge: Cambridge University Press.

    Google Scholar 

  27. Breslin, B. (2009). From words to worlds: Exploring constitutional functionality. Baltimore: Johns Hopkins University Press.

    Google Scholar 

  28. Brown, D. S., Brown, J. C., & Desposato, S. W. (2008). Who gives, who receives, and who wins? Transforming capital into political change through nongovernmental organizations. Comparative Political Studies,41(1), 24–47.

    Google Scholar 

  29. Buchanan, J. M. (1975). The Limits of Liberty: Between Anarchy and Leviathan. Chicago: University of Chicago Press.

    Google Scholar 

  30. Buchanan, J. M., & Tullock, G. (1961). The calculus of consent. Ann Arbor: University of Michigan Press.

    Google Scholar 

  31. Cameron, A. C., & Trivedi, P. K. (2005). Microeconometrics: Methods and applications. Cambridge: Cambridge University Press.

    Google Scholar 

  32. Cammett, M. C., & MacLean, L. M. (2014). Introduction. In L. M. MacLean & M. C. Cammett (Eds.), The politics of non-state social welfare (pp. 1–16). Ithaca: Cornell University Press.

    Google Scholar 

  33. Carothers, T. (2006). The backlash against democracy promotion. Foreign Affairs,85(2), 55–68.

    Google Scholar 

  34. Carothers, T. (2015). The closing space challenge: How are funders responding? Washington, DC: Carnegie Endowment for International Peace.

    Google Scholar 

  35. Carothers, T., & Brechenmacher, S. (2014). Closing space: Democracy and human rights support under fire. Washington, DC: Carnegie Endowment for International Peace.

    Google Scholar 

  36. Chahim, D., & Prakash, A. (2013). NGOization, foreign funding, and the Nicaraguan civil society. VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations,25(2), 487–513.

    Google Scholar 

  37. Chapman, T. L. (2009). Audience beliefs and international organization legitimacy. International Organization,63(4), 733–764.

    Google Scholar 

  38. Christensen, D., & Weinstein, J. M. (2013). Defunding dissent: Restrictions on aid to NGOs. Journal of Democracy,24(2), 77–91.

    Google Scholar 

  39. CIVICUS. (2018). CIVICUS monitor. Washington, DC: World Alliance for Citizen Participation.

    Google Scholar 

  40. Cole, D. H. (2017). Laws, norms, and the institutional analysis and development framework. Journal of Institutional Economics,13(4), 829–847.

    Google Scholar 

  41. Cole, D. H., Epstein, G., & McGinnis, M. D. (2014). Toward a new institutional analysis of social-ecological system (NIASES): Combining Elinor Ostrom’s IAD and SES frameworks. Indiana Legal Studies Research Paper No. 299, August 2015.

  42. Coppedge, M., Gerring, J., Knutsen, C. H., Lindberg, S. I., Skaaning, S.-E., Teorell, J., Altman, D., Bernhard, M., Fish, M. S., Cornell, A., Dahlum, S., Gjerløw, H., Glynn, A., Hicken, A., Krusell, J., Lührmann, A., Marquardt, K. L., McMann, K., Mechkova, V., Medzihorsky, J., Olin, M., Paxton, P., Pemstein, D., Pernes, J., von Römer, J., Seim, B., Sigman, R., Staton, J., Stepanova, N., Sundström, A., Tzelgov, E., Wang, Y.-t., Wig, T., Wilson, S., & Ziblatt, D. (2018). V-Dem [Country-Year/Country-Date] Dataset v8.

  43. Cox, D. R. (1972). Regression models and life tables. Journal of the Royal Statistical Society,34, 187–220.

    Google Scholar 

  44. Cox, D. R. (1975). Partial likelihood. Biometrika,62(2), 269–276.

    Google Scholar 

  45. Crack, A. M. (2018). The regulation of international NGOS: Assessing the effectiveness of the INGO accountability charter. VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations,29(2), 419–429.

    Google Scholar 

  46. de Tocqueville, A. (1840). Democracy in America (English ed.). Indianapolis: Liberty Fund.

    Google Scholar 

  47. DeMattee, A. J. (2019). Toward a coherent framework: A typology and conceptualization of CSO regulatory regimes. Nonprofit Policy Forum,9(4), 1–17.

    Google Scholar 

  48. Donnelly, J. (2013). Universal human rights in theory and practice (3rd ed.). Ithaca: Cornell University Press.

    Google Scholar 

  49. Dupuy, K., Ron, J., & Prakash, A. (2016). Hands off my regime! governments’ restrictions on foreign aid to non-governmental organizations in poor and middle-income countries. World Development,84, 299–311.

    Google Scholar 

  50. Edwards, M. (2004). Civil society. Malden, MA: Polity Press.

    Google Scholar 

  51. Elkins, Z., Ginsburg, T., & Melton, J. (2009). The endurance of national constitutions. Cambridge: Cambridge University Press.

    Google Scholar 

  52. Elkins, Z., Ginsburg, T., & Melton, J. (2012). Constitutional constraints on executive lawmaking. Retrieved from http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.434.9346.

  53. Elkins, Z., Ginsburg, T., & Melton, J. (2014). Comparative constitutions project: Characteristics of national constitutions, Version 2.0 (2014).The Comparative Constitutions Project (CCP).

  54. European Convention on Human Rights as amended by Protocols Nos. 11 and 14, and supplemented by Protocols Nos. 1, 4, 6, 7, 12, 13, and 16 2010 Council of Europe.

  55. Frantz, T. R. (1987). The role of NGOs in the strengthening of civil society. World Development,15(Supplement), 121–127.

    Google Scholar 

  56. Fukuyama, F. (1995). Trust: The social virtues and the creation of prosperity. New York: Free Press.

    Google Scholar 

  57. Gibelman, M., & Gelman, S. R. (2004). A loss of credibility: Patterns of wrongdoing among nongovernmental organizations. VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations,15(4), 355–381.

    Google Scholar 

  58. Gibney, M., Cornett, L., Wood, R., Haschke, P., Arnon, D., & Pisanò, A. (2017). The Political Terror Scale 19762016. (Political Terror Scale website: http://www.politicalterrorscale.org/.) Created online: http://www.politicalterrorscale.org/.

  59. Gormley, W. T. (1986). Regulatory issue networks in a federal system. Polity,18(4), 595–620.

    Google Scholar 

  60. Greenland, S. (2017). Invited commentary: The need for cognitive science in methodology. American Journal of Epidemiology,186(6), 639–645.

    Google Scholar 

  61. Hadenius, A. & Teorell, J. (2005). Assessing Alternative Indices of Democracy. C&M Working Papers, IPSA.

  62. Hanmer, M. J., & Ozan Kalkan, K. (2013). Behind the curve: Clarifying the best approach to calculating predicted probabilities and marginal effects from limited dependent variable models. American Journal of Political Science,57(1), 263–277.

    Google Scholar 

  63. Hathaway, O. A. (2002). Do human rights treaties make a difference? Faculty Scholarship Series at Yale Law School (Paper 839) (pp. 1935–2042).

  64. Hellwig, T., & Samuels, D. (2008). Electoral accountability and the variety of democratic regimes. British Journal of Political Science,38(1), 65–90.

    Google Scholar 

  65. Henkin, L. (2000). Human rights: Ideology and aspiration, reality and prospect. In S. Power & G. T. Allison (Eds.), Realizing human rights: Moving from inspiration to impact (1st ed., pp. 3–38). New York: St. Martin’s Press.

    Google Scholar 

  66. Hyde, S. D., & Marinov, N. (2012). Which elections can be lost? Political Analysis,20(2), 191–210.

    Google Scholar 

  67. ICNL. (2009). Global philanthropy in a time of crisis. In Global trends in NGO law: A quarterly review of NGO legal trends around the world (Vol. 1, no. (2), pp. 1–10). www.icnl.org.

  68. ICNL. (2015). The right to freedom of expression: Restrictions on a foundational right. In Global trends in NGO law: A quarterly review of NGO legal trends around the world (Vol. 6, no. 1, pp. 1–28). www.icnl.org.

  69. Ingram, H., & Schneider, A. (1990). Improving implementation through framing smarter statutes. Journal of Public Policy,10(01), 67–88.

    Google Scholar 

  70. Ingram, H., & Schneider, A. (1991). The Choice of Target Populations. Administration & Society,23(3), 333–356.

    Google Scholar 

  71. International Covenant on Civil and Political Rights (ICCPR) 1966 United Nations General Assembly.

  72. Jones, B. S. (1994). A longitudinal perspective on congressional elections. Ph.D. dissertation, State University of New York at Stony Brook.

  73. Kajese, K. (1987). An agenda of future tasks for international and indigenous NGOs: Views from the south. World Development,15(Supplement), 79–85.

    Google Scholar 

  74. Kameri-Mbote, P. (2002). The operational environment and constraints for NGOs in Kenya: Strategies for good policy and practice. Switzerland: International Environmental Law Research Centre. in Geneva.

    Google Scholar 

  75. Keck, M. E., & Sikkink, K. (1999). Transnational advocacy networks in international and regional politics. International Social Science Journal,15(159), 89–101.

    Google Scholar 

  76. Kiai, M. (2012). Report of the special rapporteur on the rights to freedom of peaceful assembly and association. United Nations General Assembly.

  77. Kiai, M., Stern, C., Simons, D., Anderson, G., & Kaguongo, W. (2017). Full text the Freedom of Association Chapter of FOAA Online!—The world’s most user-friendly collection of legal arguments on assembly and association rights. United Nations High Commissioner for Human Rights.

  78. King, G., & Zeng, L. (2001). Logistic regression in rare events data. Political Analysis,9(2), 137–163.

    Google Scholar 

  79. Kingdon, J. W. (1984). Agendas, alternatives, and public policies. Boston, MA: Little, Brown.

    Google Scholar 

  80. Lindblom, C. E. (1959). The science of ‘muddling through’. Public Administration Review,19(2), 79–88.

    Google Scholar 

  81. Linzer, D. A., & Staton, J. K. (2015). A global measure of judicial independence, 1948–2012. Journal of Law and Courts,3(2), 223–256.

    Google Scholar 

  82. Long, J. S. (1997). Regression models for categorical and limited dependent variables. Thousand Oaks: Sage Publications.

    Google Scholar 

  83. Long, J. S., & Freese, J. (2014). Regression models for categorical dependent variables using stata (3rd ed.). College Station, TX: Stata Press Publication, StataCorp LLP.

    Google Scholar 

  84. Lowi, T. J. (1964). American business, public policy, case-studies, and political theory. World Politics,16(4), 677–715.

    Google Scholar 

  85. Lowi, T. J. (1972). Four systems of policy, politics, and choice. Public Administration Review,32(4), 298–310.

    Google Scholar 

  86. MacLean, L. M. (2011). State retrenchment and the exercise of citizenship in Africa. Comparative Political Studies,44(9), 1238–1266.

    Google Scholar 

  87. Mahajan, V., & Peterson, R. A. (1985). Models for innovation diffusion. Beverly Hills: Sage Publications.

    Google Scholar 

  88. Malesky, E., & Schuler, P. (2011). The single-party dictator’s dilemma: Information in elections without opposition. Legislative Studies Quarterly,36(4), 491–530.

    Google Scholar 

  89. Marshall, M. G., Gurr, T. R., & Jaggers, K. (2017). Polity IV Project: Political regime characteristics and transitions, 18002016. (Center for Systemic Peace). www.systemicpeace.org.

  90. Maru, M. T. (2017). Legal frameworks governing non-governmental organizations in the horn of Africa. Kampala,

  91. Mayhew, S. H. (2005). Hegemony, politics and ideology: The role of legislation in NGO–government relations in Asia. The Journal of Development Studies,41(5), 727–758.

    Google Scholar 

  92. McGinnis, M. D., & Ostrom, E. (2014). Social-ecological system framework: Initial changes and continuing challenges. Ecology and Society,19(2), 30–42.

    Google Scholar 

  93. Mettler, S., & Soss, J. (2004). The consequences of public policy for democratic citizenship: Bridging policy studies and mass politics. Perspectives on Politics,2(01), 55–73.

    Google Scholar 

  94. Morgan, P. (2018). Ideology and relationality: Chinese aid in Africa revisited. Asian Perspective,42(2), 207–238.

    Google Scholar 

  95. Murphy, W. F. (1993). Constitutions, constitutionalism, and democracy. In D. Greenberg, S. N. Katz, M. B. Oliviero, & S. C. Wheatley (Eds.), Constitutionalism and democracy: Transitions in the contemporary world (pp. 3–25). New York: Oxford University Press.

    Google Scholar 

  96. Musila, G. M. (2019). Freedoms under threat: The spread of anti-NGO measures in Africa. Retrieved from www.freedomhouse.org.

  97. Mutunga, W. (1999). Constitution-making from the middle: Civil society and transition politics in Kenya, 1992–1997. Nairobi: SAREAT.

    Google Scholar 

  98. Ndegwa, S. N. (1996). The two faces of civil society: NGOs and politics in Africa. West Hartford, CT: Kumarian Press.

    Google Scholar 

  99. North, D. C., & Weingast, B. R. (1989). Constitutions and commitment: The evolution of institutions governing public choice in seventeenth-century England. The Journal of Economic History,49(4), 803–832.

    Google Scholar 

  100. Ostrom, E. (1990). Governing the commons: The evolution of institutions for collective action. Cambridge: Cambridge University Press.

    Google Scholar 

  101. Ostrom, V. (1997). The meaning of democracy and the vulnerabilities of democracies: A response to Tocqueville’s challenge. Ann Arbor: University of Michigan Press.

    Google Scholar 

  102. Ostrom, E. (2005). Understanding institutional diversity. Princeton: Princeton University Press.

    Google Scholar 

  103. Ostrom, E. (2011). background on the institutional analysis and development framework. Policy Studies Journal,39(1), 7–27.

    Google Scholar 

  104. Ostrom, E., & Cox, M. (2010). Moving beyond panaceas: A multi-tiered diagnostic approach for social-ecological analysis. Environmental Conservation,37(4), 451–463.

    Google Scholar 

  105. Ostrom, E., & Ostrom, V. (2004). The quest for meaning in public choice. American Journal of Economics and Sociology,63(1), 105–147.

    Google Scholar 

  106. Pallas, C., Anderson, Q., & Sidel, M. (2018). Defining the scope of aid reduction and its challenges for civil society organizations: Laying the foundation for new theory. VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations,29(2), 256–270.

    Google Scholar 

  107. Pierson, P. (1993). When effect becomes cause: Policy feedback and political change. World Politics,45(4), 595–628.

    Google Scholar 

  108. Pierson, P. (1994). Dismantling the welfare state? Reagan, Thatcher, and the politics of retrenchment. Cambridge, UK: Cambridge University Press.

    Google Scholar 

  109. Pierson, P. (2000). Increasing returns, path dependence, and the study of politics. The American Political Science Review,94(2), 251–267.

    Google Scholar 

  110. Pitkin, H. F. (1987). The idea of a constitution. Journal of Legal Education,37, 167–170.

    Google Scholar 

  111. Popplewell, R. (2018). Civil society, legitimacy and political space: Why some organisations are more vulnerable to restrictions than others in violent and divided contexts. VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations,29(2), 388–403.

    Google Scholar 

  112. Powell, E. J., & Staton, J. K. (2009). Domestic judicial institutions and human rights treaty violation. International Studies Quarterly,53(1), 149–174.

    Google Scholar 

  113. Putnam, R. D. (1993). Making democracy work: Civic traditions in modern Italy. Princeton: Princeton University Press.

    Google Scholar 

  114. Rakner, L. (2019). Democratic rollback in Africa. Oxford Research Encyclopedia of Politics. Oxford University Press. Retrieved July 30, 2019, from https://oxfordre.com/politics/view/10.1093/acrefore/9780190228637.001.0001/acrefore-9780190228637-e-887.

  115. Reddy, M. (2018). Do good fences make good neighbours? Neighbourhood effects of foreign funding restrictions to NGOs. St Antony’s International Review,13(2), 109–141.

    Google Scholar 

  116. Reimann, K. D. (2006). A view from the top: International politics, norms and the worldwide growth of NGOs. International Studies Quarterly,50(1), 45–67.

    Google Scholar 

  117. Rutzen, D. (2015). Aid barriers and the rise of philanthropic protectionism. International Journal of Not-for-Profit Law,17(1), 1–42.

    Google Scholar 

  118. Sabatier, P. A. (1988). An advocacy coalition framework of policy change and the role of policy-oriented learning therein. Policy Sciences,21(2), 129–168.

    Google Scholar 

  119. Sabatier, P. A., & Jenkins-Smith, H. C. (1993). Policy change and learning: An advocacy coalition approach. Boulder, CO: Westview Press.

    Google Scholar 

  120. Sabatier, P. A., & Weible, C. M. (2007). The advocacy coalition framework: Innovations and clarifications. In P. A. Sabatier (Ed.), Theories of the policy process (2nd ed., pp. 189–222). Boulder, CO: Westview Press.

    Google Scholar 

  121. Salamon, L. M. (Ed.). (2002). The tools of government: A guide to the new governance. Oxford: Oxford University Press.

    Google Scholar 

  122. Salamon, L. M., Benevolenski, V. B., & Jakobson, L. I. (2015). Penetrating the dual realities of government-nonprofit relations in Russia. VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations,26(6), 2178–2214.

    Google Scholar 

  123. Salamon, L. M., & Toepler, S. (1997). The international guide to nonprofit law. New York: Wiley.

    Google Scholar 

  124. Salamon, L. M., & Toepler, S. (2000). The influence of the legal environment on the nonprofit sector. No. 17 [Lecture]. The Johns Hopkins Center for Civil Society Studies, unpublished.

  125. Salamon, L. M., & Toepler, S. (2012). The impact of law on nonprofit development: A framework for analysis. In C. Overes & W. van Ween (Eds.), Met Recht Betrokken (pp. 276–284). Amsterdam: Kluwe.

    Google Scholar 

  126. Sartori, G. (1976). Parties and party systems: A framework for analysis. Cambridge: Cambridge University Press.

    Google Scholar 

  127. Schnable, A. (2015). New American relief and development organizations: Voluntarizing global aid. Social Problems,62(2), 309–329.

    Google Scholar 

  128. Schneider, A., Ingram, H., & deLeon, P. (2014). Democratic policy design: Social construction of target populations. In P. A. Sabatier & C. M. Weible (Eds.), Theories of the policy process (3rd ed., pp. 105–150). New York, NY: Westview Press.

    Google Scholar 

  129. Sidel, M. (2017). State regulation and the emergence of self-regulation in the Chinese and Vietnamese nonprofit and philanthropic sectors. In O. B. Breen, A. Dunn, & M. Sidel (Eds.), Regulatory waves: Comparative perspectives on state regulation and self-regulation policies in the nonprofit sector (pp. 92–112). Cambridge: Cambridge University Press.

    Google Scholar 

  130. Smyth, R. (2019). Winning hybrid elections: Organized opposition, incumbent regimes, and threat of popular engagement, world politics research seminar. Indiana University Department of Political Science.

  131. Smyth, R., & Turovsky, R. (2018). Legitimising victories: Electoral authoritarian control in Russia’s gubernatorial elections. Europe-Asia Studies,70(2), 182–201.

    Google Scholar 

  132. Stremlau, C. (1987). NGO coordinating bodies in Africa, Asia, and Latin America. World Development,15(Supplement), 213–225.

    Google Scholar 

  133. Toepler, S., Pape, U., & Benevolenski, V. (2019). Subnational variations in government-nonprofit relations: A comparative analysis of regional differences within Russia. Journal of Comparative Policy Analysis: Research and Practice. https://doi.org/10.1080/13876988.2019.1584446.

    Article  Google Scholar 

  134. U.N. Human Rights Committee. (2006). UN Doc. CCPR/C/88/D/1274/2004: Views of the Human Rights Committee under article 5, paragraph 4, of the Optional Protocol to the International Covenant on Civil and Political Rights (88th session). United Nations. http://hrlibrary.umn.edu/undocs/1274-2004.html.

  135. U.N. Human Rights Committee. (2007). UN Doc. CCPR/C/90/D/1296/2004 : Views of the Human Rights Committee under article 5, paragraph 4, of the Optional Protocol to the International Covenant on Civil and Political rights (9th session). United Nations. http://www.un.org/en/ga/search/view_doc.asp?symbol=CCPR/C/90/D/1296/2004.

  136. U.N. Human Rights Committee. (2015). UN Doc. CCPR/C/115/D/2011/2010: Views of the Human Rights Committee under article 5(4) of the Optional Protocol to the International Covenant on Civil and Political Rights (115th session). United Nations.

  137. Union of International Associations. Yearbook of International Organizations. (Online: https://uia.org/yearbook). Created online: https://uia.org/yearbook.

  138. United Nations Office of Legal Affairs. (2018). Status of ratification of a core international human rights treaty or its optional protocol. (Online)

  139. Voeten, E. (2013). Data and analyses of voting in the UN general assembly. In Reinalda, B. (Ed.), Routledge handbook of international organization. https://doi.org/10.2139/ssrn.2111149.

  140. Wasserstein, R. L., Schirm, A. L., & Lazar, N. A. (2019). Moving to a world beyond “p < 0.05”. The American Statistician,73(sup1), 1–19.

    Google Scholar 

  141. Wintrobe, R. (1998). The political economy of dictatorship. Cambridge: Cambridge University Press.

    Google Scholar 

  142. Woldense, J. (2018). The ruler’s game of musical chairs: Shuffling during the reign of Ethiopia’s last emperor. Social Networks,52, 154–166.

    Google Scholar 

  143. World Bank. (1997). Handbook on good practices for laws relating to non-governmental organizations. Washington, DC: The World Bank.

    Google Scholar 

  144. World Bank. (2018). World development indicators (WDI). Washington, DC. https://datacatalog.worldbank.org/dataset/world-development-indicators.

  145. Zahariadis, N. (2014). Ambiguity and Multiple Streams. In P. A. Sabatier & C. M. Weible (Eds.), Theories of the policy process (3rd ed., pp. 25–58). New York, NY: Westview Press.

    Google Scholar 

Download references

Acknowledgements

I am grateful to Jennifer Brass, Terrance Chapman, Sean Nicholson-Crotty, Michelle Reddy, Chrystie Swiney, Joey Carroll, Martin Delaroche, Renzo de la Riva Agüero, Laura Montenovo, the reviewers, and the journal’s editorial team for their written comments and constructive criticisms. My appreciation also extends to those who offered early input on the project at the International Society for Third-Sector Research, Brass Club, the Emory Conference on Institutions and Lawmaking, the Midwest Political Science Association, the International Spring School on Public Policy, and the Workshop on the Ostrom Workshop (WOW6).

Author information

Affiliations

Authors

Corresponding author

Correspondence to Anthony J. DeMattee.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Appendix

Appendix

See Tables 5, 6, 7, 8, 9 and 10.

Table 5 Descriptive statistics of estimation sample
Table 6 Pooled event history analysis (EHA) with logistic regression
Table 7 Competing risk model with logistic regression (full table)
Table 8 Cox with time-varying coefficients CRM (full model)
Table 9 Rare events CRM (logistic regression, random 35% of sample)
Table 10 Assessing conceptual equivalence and distinctness (all modeling strategies)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

DeMattee, A.J. Covenants, Constitutions, and Distinct Law Types: Investigating Governments’ Restrictions on CSOs Using an Institutional Approach. Voluntas 30, 1229–1255 (2019). https://doi.org/10.1007/s11266-019-00151-2

Download citation

Keywords

  • CSOs
  • NGOs
  • Civil society
  • Regulations
  • Constitutions
  • International treaties