The Autocratic Welfare State: Resource distribution,
credible commitments and regime survival
November 22, 2015
Abstract
We investigate the adoption, characteristics and consequences of social policies in autocracies.
Our argument on why autocrats implement such policies, particularly pension programs, highlights that they can be considered club goods that I) are targeted to critical supporting groups and
II) solve credible commitment problems on promises of future distribution, thereby mitigating the
probability of regime breakdown. We test different implications from the argument by combining
extant regime data with novel data on welfare state programs, for 140 countries with time series
from the 1880s. We find that autocracies are as likely as democracies to have pension systems.
But, autocracies have less universal pension (and other) programs. Further, we find that such
programs reduce probabilities of autocratic breakdown and of democratization.
1
Introduction
The welfare state literature has considered social policy programs either as mechanisms for providing
individuals with insurance against risks such as job-loss or illness, or as tools for progressive redistribution (see Moene and Wallerstein 2001). Hence, different theoretical models are well suited to
explain the existence of social policy programs in democracies; most voters are risk averse (Chetty
2006), thus preferring even costly insurance, and the (relatively poor) median voter often has incentives to vote for parties promising income redistribution (Meltzer and Richard 1981). Yet, social
policy programs are adopted not only in democracies, but also in autocracies. The perhaps most
famous examples are the sickness, accident, disability and old-age pension programs adopted in Bismarck’s Germany, but these are not exceptions. Mares and Carnes (2009, 97) count that the vast
majority of the 38 “non-industrialized” countries for which they have data actually adopted their first
old-age pension system under autocracy. Indeed, analyzing more than 140 countries, with time series
from the 1880s, we show below that autocracies are no less likely than democracies to adopt or have
old-age pension systems. But, why would autocrats adopt such programs and spend resources on
pensions rather than keeping tax revenues for private consumption, or for investment in repressive
capacity to ensure regime survival?
We propose that pension programs in autocracies are, indeed, used for political survival purposes,
and our statistical tests estimate a clear and sizable causal effect of pension programs on regime survival once adjusting for the endogenous adoption of such programs. Social policies, and notably
old-age pensions, involve targeted distribution of resources to particular groups, and thus constitute a
special form of co-optation. While the role of formalized social policy programs has received comparatively little attention in the literature on co-optation in autocracies, which often focuses on discretionary distribution and patronage networks, this insight is not new. Mares and Carnes (2009), for
instance, explicitly argue that social policies are used for targeted distribution to critical groups. However, neither Mares and Carnes nor the wider literature have elaborated on a second critical feature
of such programs, namely how they allow autocrats to make credible promises of future distribution.
We thus respond to the theoretical challenge raised by, for example, Acemoglu and Robinson (2006);
why would any group trust a dictator’s current promise of future distribution (see also, e.g., Svolik
2012)?
In brief, we conceptualize social policy programs in autocracies as club-goods, and highlight that
1
they combine two characteristics – i) inter-temporal credibility and ii) potential for targeting – which
make them vital for autocratic survival. First, in contrast with discretionary distribution of private
goods, social policy programs, such as old-age pension programs, are fairly transparent, stable and
predictable, and not easy to reverse without substantial costs for rulers. Hence, distribution through
these programs – not unlike establishing particular institutions such as ruling parties (Magaloni 2008;
Gehlbach and Keefer 2011, 2012; Svolik 2012; Boix and Svolik 2013) – may solve the critical credible commitment problem. Second, in contrast with pure public goods, pension programs can be
targeted towards identifiable groups. Hence, autocrats can design programs channeling resources to
groups they need support from to stay in power, rather than spending precious resources on irrelevant groups. Although implementing and sustaining, for example, old-age pension programs could
be more resource-demanding than discretionary distribution, it is thus preferable under some circumstances, for instance when the critical supporting groups of the regime are fairly large. We thereby
also address the important question of why (and under what circumstances) dictators would opt to
distribute through targeted social policy programs rather than targeted discretionary spending.
The argument generates various empirical implications – including on the typical characteristics
of autocratic pension programs, and how they affect regime survival. We test these implications using
our new Social Policies Around the World (SPAW) dataset, with global coverage and time series
from the 19th century. By detailing how and why autocrats employ pension programs, we bridge
(and contribute to) the welfare state literature, which has mainly focused on (developed) democratic
contexts, and the rapidly growing literature on autocratic politics.
After having briefly reviewed relevant literature on social policies and survival strategies in autocracies (Section 2), we present our argument (Section 3). We thereafter describe our novel SPAW
dataset (Section 4). In the empirical analysis (Section 5), we report evidence that autocracies are as
likely as democracies to adopt and have pension systems, although autocracies have less universal
systems. Finally, we present the first systematic empirical study of its kind on how social policies affect regime survival and change, finding clear evidence that pension systems reduce the probability of
autocratic regime breakdown. This estimated effect, which holds in models trying to account for the
implementation of pensions being endogenous to the security threats facing the regime, is statistically
significant and substantial in size. In extension, we report evidence suggesting that implementing
pension systems may not only make autocracies less likely to break down, in general, but also reduce
2
the probability of democratic transitions.
2
Literature
Most welfare state researchers have focused on OECD democracies (Flora and Heidenheimer 1981;
Korpi 1983; Esping-Andersen 1990; Huber and Stephens 2001; Iversen 2005). Different theoretical
accounts then also suggest that social policies – and the provision of public services such as basic
health-care and primary education (e.g., Lake and Baum 2001; Stasavage 2005; Harding and Stasavage 2014) – should be more widespread, better financed, and have broader coverage in democracies
than in autocracies (Boix 2003; Acemoglu and Robinson 2006; Haggard and Kaufman 2008; Ansell
2010). There is evidence that Western countries expanded welfare spending due to franchise extensions from the 19th century onwards (Lindert 2005; Kim 2007).
Yet, some empirical studies report that democracies are not, on net, associated with higher social
spending (e.g., Mulligan, Sala-i Martin and Gill 2003).1 Former Communist dictatorships presided
over extensive social policy programs (Milanovic 1998), but so did autocracies such as PRI-Mexico
(Magaloni 2006). Desai, Olosgård and Yousef (2009) find that – among autocracies – those with
harsher repression provide more welfare spending. Famously, a historical pioneer in implementing
welfare policy programs was Imperial Germany, under the strategic guidance of Chancellor Bismarck
(see Rimlinger 1971). Different strands of literature provide two particularly important insights about
autocratic politics. We combine these two insights in our argument on how autocrats conduct distributive politics, and apply them more specifically to old-age pension systems: First, policies are
intentionally used by many autocrats to remain in power. Second, ensuring that promises made to the
regime’s critical supporters are credible is of utmost importance:
Regarding the first insight, some autocrats might be genuinely concerned with the welfare of
their citizens (see, e.g., Wintrobe 1998, 95–96), and provide old-age public pensions or unemployment insurance to improve their life quality. However, the literature often considers policy-making
as a strategic tool used mainly for for enhancing regime survival (e.g., Bueno de Mesquita et al.
2003; Svolik 2012). Several studies show how autocratic regimes rely on particular groups for their
political survival, and design economic policies to satisfy these groups. Bueno de Mesquita et al.
(2003) theorize that whereas large-coalition rulers should find it cost-effective to spend resources on
1
There are likely systematic differences among democracies; proportional representation (PR)
electoral systems may enhance social spending and induce more universal programs than pluralmajoritarian (e.g., Persson and Tabellini 2003; Iversen and Soskice 2009).
3
public goods, small-coalition rulers should allocate private goods directly to their few vital supporters. Democracies, because of larger coalitions, should observe more spending on public goods than
dictatorships (see also Lake and Baum 2001). Still, there are differences in coalition size between
dictatorships – for example, party-based autocracies generally have larger coalitions (Fjelde 2010) –
and thus also in optimal resource-distribution strategies. Also regarding social policies, more specifically, regime survival was widely considered a core motivation for Bismark’s reforms (Rimlinger
1971), and Mares and Carnes (2009) and Haggard and Kaufman (2008) outline how (targeted) social policies were designed and implemented, at least in part, for political-survival reasons in various
autocracies.
Regarding the second insight, the theoretical argument in Acemoglu and Robinson (2006) highlights how autocratic elites often cannot credibly commit to future redistribution, even when faced
with a revolutionary threat; they would expectedly stop redistributing (i.e., stop “wasting resources”)
once the threat is removed. The dictator may ensure survival through repression, but this is costly
(Wintrobe 1998), and co-optation is often preferable (see Gandhi 2008). Although autocrats may
face difficulties in making promises of future distribution credible, this is far from impossible. Different studies (North and Weingast 1989; Geddes 1999; Magaloni 2006; Gandhi 2008; Wright 2008;
Gehlbach and Keefer 2011, 2012; Svolik 2012; Boix and Svolik 2013, for democracies, see Iversen
2005) point out that autocrats can reduce credible commitment problems through establishing regime
parties and other institutions, such as legislatures. By allowing, for example, party institutions to
check the ruler, the latter cannot easily alter distributive policies favoring the current supporting
coalition. Hence, promises of future distribution are more credible within party-based regimes, in
turn reducing risks of regime splits, non-constitutional leader removal, or even regime collapse (Geddes 1999). However, we take this one step further by proposing that not only institutions, but also the
design of distributive policies matter; formalized social policy programs – and particularly old-age
pension programs – mitigate credible commitment problems.
3
Why would autocrats implement social policies?
We argue that autocrats often have strong incentives to implement social policy programs to please
their supporters, and this effectively helps them prolong their rule. Institutionalizing the distribution
of resources through such programs makes future distribution to particular groups more credible than
promises of discretionary private-goods distribution. When distributive policy is clearly codified, the
4
autocrat cannot as easily backtrack on promises once the relevant group no longer poses an imminent
threat. Changing social policy programs may be particularly difficult in institutionalized autocracies
with considerable checks on the ruler, as reversing programs require the co-operation of multiple
veto players. Implementing generous welfare programs does imply unwelcome future costs. Yet,
this is something the autocrat willingly accepts if the alternative is losing power; when promises of
future distribution are considered credible, the benefiting groups will more likely support the regime.
Program benefits should, however, be targeted towards groups the regime needs support from – be it
urban workers or soldiers – whereas “irrelevant” citizens should receive little. By appeasing groups
that might otherwise contest the regime, the introduction of such programs should bolster autocratic
regime survival, and even worsen prospects for democratization.
Old-age pension systems should be particularly suitable as such instruments of regime survival.
First, the amount of resources channeled through pension programs often dwarfs the amount channeled through other programs, such as unemployment benefit programs. Second, old-age pension
systems can easily be designed to target various groups. Pension schemes are more flexible cooptation devices than, say, maternity leave programs; pensions are not restricted to pre-given groups,
such as women of child-bearing age, that may or may not be relevant for political survival. Also,
unemployment benefits programs advantage those with high risks of losing their jobs, and autocracies
have typically focused on ensuring the job safety of critical supporting groups through employment
protection legislation and other means (e.g., Mares and Carnes 2009). Further, disability insurance
might come in handy for members of the autocrat’s critical supporting groups, but the risk of disabilitating work injury is relatively small in comparison with that of growing old. Third, old-age pension
systems speak directly to the temporal dimension of the credible commitment problem. For most
concerned individuals, pensions will be paid out in the not-too-near future – i.e., when current coalition members are retired from their positions, and thus perhaps less capable of holding the autocrat
accountable. Absent old-age pension schemes, current coalition members may not trust dictators to
provide future resources.
Let us explore the argument and its core assumptions in more detail. To simplify, we divide the
population in dictatorships into two groups, the regime’s critical supporters (S) and all other citizens
(C). In addition, we have the regime (R), i.e. the dictator and his/her very close collaborators. S may,
for instance, consist of soldiers, land-owners, civil servants, former army veterans, or workers in vital
5
industrial sectors. One key aspect, however, is that members of S are identifiable according to some
feature (e.g. “Participated in war of independence” or “Works in the oil industry”). Leaving repression
aside, we consider co-optation through current and (promised) future distribution of resources the key
policy tool R employs to stay in power.2
Since, R needs the approval of S, but not C, to remain in power, R is concerned only with pleasing
the former. Members of S care about present and future consumption, and thus require resources
distributed to them today and credible promises of resources tomorrow. If not, S may seek replacing
R with another regime more willing to comply with their demands, or even pursue democratization
which has some benefits also to economic and other elites (Ansell and Samuels 2014). R can basically
co-opt S through three strategies: 1) lump-sum provision of private goods; 2) public goods provision
benefiting both S and C; and, 3) provision of club goods to S. Club goods are non-rivalrous; their
consumption have externalities, and club goods thus affect a group of individuals. But, in contrast to
public goods, they are excludable. Pension checks are posted to individuals and consumed by them
privately – distribution through pensions therefore has some private-good characteristics – but there
are good reasons to consider segmented pension programs as club goods:
3.1
Social policies as club goods – targeting
One key aspect of social policies is their insurance function (Moene and Wallerstein 2001). Individuals cannot perfectly predict whether they will be unemployed five years from now or capable
of working when they turn 70, and many would sacrifice some current income to insure against the
low income levels accompanying such events. Payments should, importantly, be implemented under
clearly specified conditions and in a rule-following manner. If the government could simply decide
on a case-by-case basis which members should (not) receive payments, this would undermine the program’s general credibility and reduce expected utility for all program members. Hence, the insurance
aspects of welfare programs have non-rivalrous properties.
Regarding excludability, this requires the presence of observable, codifiable criteria for distinguishing program members from non-members. The excludability aspect of pension programs, combined with R’s incentives, indicates that such programs will be targeted (only) towards members of S
(see also Mares and Carnes 2009). Haggard and Kaufman (2008) describe the targeted character of
welfare states in numerous Asian and Latin American countries. In Thailand and South Korea, for
2
Yet, repressive strategies are likely endogenous to autocrats’ co-optation strategies (Desai,
Olosgård and Yousef 2009; Frantz and Kendall-Taylor 2014).
6
example, social insurance was limited “to core constituents in the state itself: the military and civil
servants” (p. 140), whereas in pre-Perón Argentina “social-insurance protection was confined mainly
to the military, white-collar workers, and a few strategically situated blue-collar unions” (p. 95).
R has strong incentives to under-provide truly public goods, since these are only cost-effective
in ensuring political survival when the critical supporting groups are very large (Bueno de Mesquita
et al. 2003). This leaves R with the two targeted co-optation strategies, discretionary distributing
private-goods to individuals in S or establishing club goods – notably targeted pension programs –
for S collectively. The former has some important benefits from R’s point of view. First, if R, at
some point, strongly consolidates power, R can easily reduce the amount of private goods distributed,
leaving more resources for R’s personal consumption. Formalized social policy programs are, as
discussed below, more difficult to adjust at will. Thus, dictators (and their closest collaborators)
who consider themselves very safe in power should prioritize private-goods distribution. Second,
discretionary distribution of private goods is arguably easier to direct exclusively to members of S; in
practice, social policy programs may lead to “leakage” or “waste”; some citizens that are unimportant
for the dictator’s survival may receive benefits. For instance, former servicemen with little political
influence are covered by veteran benefits, or “irrelevant” parts of the manufacturing sector producing
in the periphery are covered by pension schemes for manufacturing workers, even if these are mainly
intended for powerful firms and workers in the capital.
3.2
Social policies and solving commitment problems
Yet, R, particularly when the future grip on power is insecure, may substitute private goods distribution with formalized and fairly stable social policies targeting critical groups. The reason is the
exact same stability of such programs making them monetarily more costly. R wants to avoid making
S uncertain about future distribution; such uncertainty may induce S to replace R. Acemoglu and
Robinson (2006), among others, highlight the fundamental problems of time consistency and credibility of promises facing autocrats. They propose that a liberalizing regime change, thus giving away
political decision-making powers, is the only viable solution for autocrats facing imminent threats.
Still, recent contributions have highlighted how particular institutions in autocracies alleviate this
problem (Svolik 2012; Boix and Svolik 2013). We further propose that the way in which distributive
policies are designed contribute to determine whether promises of future distribution are credible or
not. If the future reversal of policies entail high economic or political costs, autocrats can credibly
7
promise their continuation. Thus, autocrats may want to “tie their own hands” when designing policies that channel resources to critical groups. One way of doing this is through codifying pension
programs.
Implementing such programs stabilizes distributive policy through multiple mechanisms. For instance, R initially makes investments in administrative and monitoring capacity, thus reducing (the
relative) costs of continuing the program once up and running, creating lock-in effects (Pierson 2000).
Further, the reversal of pension programs constitute easily identifiable events which could function
as signals, or “sparks” (Kuran 1989), alleviating anti-regime collective action problems. The relative ease with which members of S can identify “breaches to co-operation” reduces R’s incentives to
reverse such programs to begin with. Thus, pension programs, once implemented, constitute “guarantees” of future distribution, thereby reducing coalition incentives to remove the regime. Indicatively,
Ponticelli and Voth (2011), using European data from 1919–2008, find that cuts in public spending generate unrest and anti-regime collective action. Our SPAW data also show that social policy
programs are durable and seldom retracted or scaled down substantially. For instance, only 10% of
changes to old-age pension systems in our dataset, in terms of expanding or reducing the number of
groups covered, are reductions. Also following expectations, such reductions in pension programs are
far less likely in established autocracies than immediately after regime change when autocrats may
cut programs benefitting supporters of the old regime (see also Albertus and Menaldo 2012) – the
relative frequency of such pension reductions is almost three times higher in the first three years after
regime change. Withdrawing pension benefits from S is likely very dangerous for R. Anticipating
this, members of S perceive promises of future resource distribution as credible if linked to pension
programs. This is the case particularly if members of S expect the regime to be long-lived – which is
the case especially for party-based autocracies (e.g., Geddes 1999). Hence, S will also have incentives
to support R’s hold on power once a pension program is in place.
3.3
Empirical implications
The argument produces quite diverse implications, and we highlight three hypotheses. First, since
costs of setting up and monitoring policy programs are likely increasing but concave in number of
recipients, small-coalition dictatorships may more often employ non-formalized distribution of private goods to their supporters, and dictatorships with larger winning coalitions should, everything
8
else equal, more often operate social programs than small-coalition dictatorships.3 However, our argument does not imply that the systems with the largest coalitions, namely democracies, have more
programs. Although large coalitions increase incentives to spend through formalized programs, this
could very well be balanced by democratic leaders not needing to solve the theoretically highlighted
commitment problems through such mechanisms. Leaders and voters know the former may more
easily be removed from office in the next election if they do not redistribute (e.g., Acemoglu and
Robinson 2006). On the other hand, such commitment issues were at the core of our argument for
why autocracies operate welfare programs.
More specifically, our expectations on (lack of) clear and systematic differences between democracies and autocracies mainly concern old-age pension systems. Pension schemes are, as discussed,
more flexible autocratic co-optation devices than, for example, maternity leave programs, and for most
concerned individuals pensions will be paid out only after several years – i.e., when current coalition
members are retired from their positions, and thus less capable of holding the autocrat accountable.
Absent old-age pension schemes, current “coalition members” may not trust dictators to provide future resources. Thus, the coalition size and commitment mechanisms may very well balance each
other out, in terms of inducing leaders to provide pension systems in autocracies and democracies.
One plausible hypothesis is therefore:
H1) Autocracies are not (systematically) less likely than are democracies to adopt/have old-age
pension systems.
Yet, our argument clearly indicated that the social policies, notably including pension programs,
pursued in autocracies should, generally, be more targeted than in democracies. This mainly stems
from autocracies having narrower supporting coalitions. This yields
H2) Old-age pension systems are less universal in autocracies than in democracies.
In extension we also test whether, party-based autocracies, which often have larger coalitions
(but which are also associated with other factors that may affect welfare characteristics in different
3
Extant measures only allow us to test this indirectly: Party-based autocracies typically have larger
coalitions (see, e.g., Fjelde 2010), and one might therefore expect them to more often operate welfare
programs. We test this empirically below. However, one theoretically relevant moment that may
attenuate this net association is that some ruling party institutions could be so strong that they, in
themselves, ensure promises of future distribution being credible (e.g., Svolik 2012), reducing the
need to formalize distributive policies through pension programs.
9
directions), have more universal programs than monarchies or military regimes.
Factors influencing the risk of regime breakdown might incentivize autocrats to initiate (or expand) pension programs. However, forward-looking autocrats may implement such programs as
safety-nets long before the economy crashes or protesters assemble in the streets (when it may be
too late to save the situation by promising benefits). Nevertheless, the notion that such programs –
once in place – boost autocratic survival prospects is a testable proposition following our argument.
Since we highlight that adoption of pension programs is endogenous we must adjust for selection
mechanisms when estimating their effects on survival. Hence, our final hypothesis is
H3) Old-age pension systems reduce the risk of autocratic regime breakdown
Further, if social policy programs reduce commitment problems for autocrats, as we argue, they
should expectedly also mitigate the (more specific) probability of democratizing regime transitions
(see Acemoglu and Robinson 2006). As another extension we thus estimate the effect of old-age
pension systems on democratization.4 Also, despite our focus on old-age pensions, we elaborate on
various nuances of the “Autocratic Welfare State” by briefly discussing resembling tests on the five
other types of social policy programs coded in SPAW.
4
The SPAW dataset
Our new Social Policies around the World (SPAW) dataset covers 223 countries, with time series for
some variables from 1790–2010. SPAW draws on various sources, notably the ILO Legislative series
(1919-) and US Labor Department SSPTW-reports (1937-).5 As detailed below, it contains variables
measuring first year of legislation and existence of programs and which groups are covered for all
major transfer programs, allowing us to test whether autocrats implement particular social policies,
such as targeted pension programs, and whether these, in turn, systematically affect regime survival.
Previous analysis of determinants or effects of social policies largely rely on aggregated spending measures. While interesting for some purposes, they are often poor proxies for the welfare-state
characteristics of proposed causal relevance (see, e.g., Esping-Andersen 1990; Scruggs 2006). We
are here interested in capturing to what degree benefits are targeted to specific groups, or welfare
state segmentation. Since spending measures tell very little about the structure of benefits, they are
4
This empirical implication contrasts directly with, for instance, that following from the argument
in Ansell and Samuels (2014), proposing that targeted social welfare spending should enhance the
probability of democratization (see p.144).
5
For details, see the Appendix and the SPAW codebook at REMOVED WEBPAGE/REFERENCE.
10
unfitting for this purpose; two countries can have identical spending levels but very different welfarestate structures (Esping-Andersen 1990; Allan and Scruggs 2004). Likewise, an important alternative
measure – the population share insured – is also problematic. For example, insurance statistics will
not distinguish countries where only one small group, say urban industrial workers, is insured, from
another where benefits accrue to all workers but the required contribution period is so high that only
some end up eligible. In the first case benefits are targeted to a specific group, whereas in the latter
eligibility is more “random”. Consequently, neither spending nor insurance measures are very appropriate for analyzing how autocrats use targeted benefits to survive by reducing regime threats from
particular groups, such as urban workers. Different variables in SPAW alleviate such problems. The
variables are grouped into two main categories, I) legislative changes and II) program characteristics.
For both I) and II), SPAW contains information on sickness leave, maternity leave, unemployment,
disability/work injury, family allowances, and old-age pensions.
For legislative-changes, we identify first year of major reform and presence of (above-minimumthreshold) programs. Contrary to previous attempts to capture first year of legislation (e.g., Mares
and Carnes 2009), we distinguish between major and special programs (see Appendix for closer
discussion). Several countries legislated special public servant pensions during the 19th century, way
before introducing old-age pension programs in their modern form. In Argentina, for instance, the
first civil servant pension (originally just for the judiciary and treasury) was enacted in 1803, whereas
the first law covering industrial workers passed 143 years later in 1946 (Mesa-Lago 1978). Our
operationalization thus only considers a (national-level) pension system to exist when at least one
major social/occupational group (see below) is covered.
The variables in SPAW improve on previous codings in several regards. First, much effort has
gone into using multiple sources – both cross-country reports and country-specific sources – to crosscheck the validity and reliability of the codings. While this introduces and element of subjective
judgement even for such de jure measures, this is highly preferable to employing only one of the
sources at hand, both for increasing coverage and for avoiding different issues associated with the
individual sources. Second, the main source is the ILO Legislative series, which we deem as the most
reliable. Previous attempts to measure “first laws” have tended to use SSPTW reports. The latter
are associated with different problems, for instance operating without a consistent criterion over time
(or between countries) on what constitutes a first law. Depending on the edition, the SSPTW reports
11
switch between using special programs or major programs. Thus, we have employed the entire archive
of SSPTW reports, alongside the other sources, and not only a restricted sample.
Regarding program characteristics, these may be conceptualized as varying along two dimensions, IIa) eligibility and IIb) distributive potential. Eligibility refers to rules defining who can partake, whereas distributive potential concerns the program’s inherent capacity to allocate benefits to
members. Therefore, eligibility relates to who is covered against risks, and distributive potential to
the benefits program members can expect. SPAW measures eligibility by first classifying whether
programs are means-tested, employment-based or fully universal, and make further nuances by constructing Universalism indices (see below) – with the longest time series starting with the German
accident liability insurance of 1871. For distributive potential, SPAW contains numerous variables,
including measures of whether program eligibility is constrained by income ceilings; number of days
recipients must wait before receiving benefits; number of weeks program benefits extend over; and,
whether the state subsidizes programs.
We use two types of variables from SPAW below, measuring particularly relevant aspects for our
argument. The first are dichotomous measures of whether pension (and other social policy) programs
– passing a minimum threshold – exist or not. This minimum threshold is operationalized as the program covering ≥ 1 of the following 8 major social/occupational groups: agricultural workers; industrial/production workers; small-firm workers; self-employed; students; employers; temporary/casual
workers; family/domestic workers.6
The second are Universalim Indices, capturing what parts of the population are included on equal
terms in the same program.7 The baseline versions range from 0–9, and are scored 0 if programs are
absent and 1 if programs are means-tested based on some property criteria (income-based exclusions
are not counted as means-tested programs; as robustness tests we run models excluding means-tested
6
Even laws (generally) prescribing broad coverage might simultaneously exclude specified groups
such as ethnic minorities (or majorities, as in Apartheid South Africa) or certain occupational groups.
SPAW thus includes a measure capturing number of explicitly excluded groups. Second, although
covered in general systems, groups might receive special benefits from additional programs. In SPAW,
the (country-year) average is 3 special programs, with a maximum of 63 special programs (Greece
1983).
7
Below, we do not distinguish voluntary from compulsory coverage, or social insurance from private schemes (provident funds), but SPAW includes variables coding, e.g., whether or not systems are
private.
12
programs and our results remain almost identical). Further, contribution- or employment-based programs covering one major group are scored 2. 3-scores indicate that programs are contribution- or
employment-based and cover two major groups, and so on, up to 8-scores where seven groups are
covered. Maximum 9-scores indicate all residents are automatically entitled to benefits – a fully universal system.8 Appendix Figure A.1 shows the baseline Universalism Index distributions for the
six major programs. Although this index is not of interval scale, we employ easy-to-interpret linear
regression models for our baselines. We also tested ordered logit and probit models, and logit and
probit models run on dummy variables (constructed by using different cut-offs) and our results are
robust (see Appendix).
While the SPAW data allow us to better measure theoretically important characteristics, there are
potential reliability and validity issues also with these data. First, while coding de jure programs
allows for constructing an extensive database with comparable measures across time and space, some
programs may be “empty shells”. Hence, our data could, for instance, include pension programs that
exist on paper only, but do not distribute resources in practice. While we would have loved to conduct
systematic tests on de facto program characteristics, these are very difficult to code in a valid manner.
Further, if many programs are simply empty shells, they would expectedly not fulfill the abovetheorized roles and have a clear causal effect on regime survival. Hence, our strong results below may
actually underestimate the “true effect” of pension programs on regime survival. Likewise, de facto
dismantling of existing programs (see Mares and Carnes 2009, 94) should attenuate our estimates.
Second, we do not differentiate groups according to size. In principle, one could weight scores
by group size, but there is no available, sufficiently disaggregated information for coding groups in a
comparable manner between countries or over time. We therefore use non-weighted measures. In any
case, our theoretical argument evolves around whether benefits are targeted to few or many (clearly
identifiable) groups. Third, our Universalism Indices do not make explicit qualitative distinctions
between the particular groups covered. Expectedly, some major groups – such as domestic workers –
are constituencies autocrats are unlikely to grant benefits, whereas other – such as industrial workers
– are critical supporting groups in many contexts. In democracies, leaders may need to cater to
both of these groups to retain power, and autocracies should thus have less universal systems than
democracies (while both should expectedly have pension programs covering at least one group). Our
results below – indeed indicating higher universalism scores among democracies – would not reflect
8
If, e.g., different pension programs exist, all are coded and we average over these programs.
13
our argument well if pension systems were first enacted covering only domestic or family workers
(while excluding, e.g., industrial workers). This is, however, not the case in practice. During our
coding, we found no instances of coverage being extended to domestic/family workers without urban
manufacturing/industrial workers also receiving benefits.9 In sum, the measures used capture our
theoretical concepts quite well.
5
Empirical analysis
We organize our tests in three sub-sections. First, we test H1 on regimes and adoption of pension
systems. We then test H2 on how targeted or universal pension programs are under different regimes.
Finally, we test H3 by investigating how pension programs causally affect autocratic regime breakdown. Corroborating our expectations, we, for instance, find that autocracies are no less likely to
have pension programs than democracies, but that democracies generally have more universal such
programs. To the the extent that our models allow for causal identification (which is, of course, always
difficult to ensure in a cross-country time-series framework), we find strong evidence of the expected
effect of pension programs on autocratic survival, but also on democratization more in particular.
5.1
Does regime type matter for probability of adopting pension programs?
We argued that autocrats use social policies for targeting and credibly promising resources to groups
that are critical to regime survival. In particular, old-age pension schemes can be effectively targeted
towards groups capable of affecting regime survival. Hence, while democracies and autocracies may
adopt pension programs for different reasons, we do not expect any clear and systematic differences
in the probabilities of adopting or having such programs between regime types.
To test H1, we mainly use the dichotomous Boix, Miller and Rosato (2013) regime measure
(henceforth BMR). BMR distinguishes between democracies and autocracies according to existence
of free and fair elections and whether more than half of the male population can vote. The dependent variable in Models A1–A3, Table 1, is a dummy recording whether an old-age pension system
(covering ≥ 1 major occupational/social group) exists (whereas Models A4–A6 have adoption of
pension system as dependent variable). BMR and existence of pension system correlate by .26 (9166
obs. from Model A1); democracies more often have pension systems than autocracies, although the
9
Benefits were usually first extended to urban employees such as industrial, commercial workers,
or salaried employees, with or without restrictions on firm size. Where agriculture was also included,
a distinction was often made between workers using and not using electric power, ensuring only
workers close to urban centers were covered.
14
difference has dwindled in more recent decades. This is illustrated by the upper-left graphic in Figure
1, which shows shares of democracies and autocracies with pension programs.
.8
.4
.2
0
0
.2
.4
.6
.8
.6
.8
.6
.4
.2
0
Sickness Benefits
1
Maternity Benefits
1
1
Old-age Benefits
1900 1920 1940 1960 1980 2000
Accident Benefits
Unemployment Benefits
Familiy Benefits
1900 1920 1940 1960 1980 2000
0
.2
.4
.6
.8
1
0
0
.2
.2
.4
.4
.6
.6
.8
.8
1
1900 1920 1940 1960 1980 2000
1
1900 1920 1940 1960 1980 2000
1900 1920 1940 1960 1980 2000
Democracies
1900 1920 1940 1960 1980 2000
Autocracies
Figure 1: Shares of democracies and autocracies, as operationalized by Boix, Miller and Rosato
(2013), with programs in six social policy areas during the 20th century.
Nevertheless, this correlation ignores other factors affecting adoption of pension systems. Pension
systems expanded during the 20th century, although this occurred earlier in some regions. Since also
democratization has clustered in temporal waves, differently timed across regions (see Huntington
1991), controlling for time period and geographic neighborhood is important. Our models therefore
always include year dummies and eight region dummies (from Miller 2013). We also include log
GDP per capita (from Maddison 2007); richer countries have more resources to spend on pension
systems, and income correlates with democracy. For our baselines, we run logit models on existence
or adoption of pension system, with BMR as the key independent variable, and with errors clustered on country to account for panel-specific autocorrelation (which is also alleviated by the year
dummies). However, we also tested Cox survival models and logit models using a dummy variable
BCSTS specification following Beck, Katz and Tucker (1998) or cubic splines following Carter and
Signorino (2010) to further account for temporal dependence. Our baseline logit results are fairly
robust to these and other plausible changes to the model set-up (see Appendix).10
10
Ideally, we would control for country-specific effects in our logit regressions, but this is infeasible.
We include country dummies below when investigating degree of universalism, which provides far
more within-country variation.
15
Democracy (BMR)
Ln GDP p.c.
Ln population
Ethnic fractionalization
Urbanization
Military size
Resource dependence
Region dummies
Year dummies
N
Countries
A1
A2
A3
DV: Existence pension program
b/(t)
b/(t)
b/(t)
0.128
0.095
-0.010
(0.32)
(0.24)
(-0.03)
0.498* 0.693***
-0.123
(1.75)
(2.59)
(-0.42)
0.445*** 0.424**
(2.85)
(2.41)
-0.615
-1.040
(-0.62)
(-1.09)
0.063***
(3.94)
-0.010
(-0.05)
0.001
(0.06)
Y
Y
Y
Y
Y
Y
9166
8330
7881
140
137
137
A4
A5
A6
DV: Adoption pension program
b/(t)
b/(t)
b/(t)
0.353
0.437
0.350
(1.11)
(1.32)
(1.09)
0.098
0.190
-0.178
(0.62)
(1.03)
(-0.59)
0.305**
0.317**
(2.55)
(2.47)
-0.273
-0.842
(-0.46)
(-1.27)
0.029***
(2.70)
0.045
(0.51)
0.004
(0.25)
Y
Y
Y
Y
Y
Y
1504
1344
1252
114
111
107
Table 1: Democracies, autocracies, and existence or adoption of pension systems.
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Logit regressions with errors clustered on country. Existence of minimum old-age pension system
is dependent variable in A1–A3 and adoption in A4–A6 (here, countries already having systems are omitted). Year dummies, region dummies and
constant are omitted from table. Maximum time series extend from late 1880s–2004.
Model A1, using these controls, is our most parsimonious specification. When accounting for
income and region- and year-specific effects, the correlation between democracy and pension system
vanishes. BMR is positive in A1, but t = 0.3. Hence, we find initial support for H1 – there is no
systematic difference between democracies and autocracies regarding existence of pension systems.
Model A2 adds two potentially relevant demographic controls. One is the (time-invariant) ethnic
fractionalization index from Alesina et al. (2003); a fractionalized society may make it more difficult,
or less desirable, to implement pension systems (see, e.g., Alesina, Baquir and Easterly 1999). The
other is log population size, from Maddison (2007). Although the latter is statistically significant –
larger countries more often have pension systems, possibly due to the increased complexity of running distributive schemes though patronage networks in larger countries, necessitating formal policy
programs such as old-age pension systems – BMR does not change much. In Model A3 we further
add urbanization, which correlates with regime type and pensions, to mitigate potential omitted variable bias. We also add two variables capturing available resources for, and alternative strategies of,
co-optation (and repression). These are revenues from oil, gas, coal and metals as percentage of GDP
and military personnel as percentage of the population.11 BMR turns negative in A3, but remains
11
Urbanization, resource dependence, and military size are from Miller (2013). The results do not
change when, e.g., rather using natural resource income measures from Haber and Menaldo (2011).
16
insignificant. These (non-)results are robust to adding other potentially relevant controls (taken from
Haber and Menaldo 2011; Miller 2013), such as civil war, British ex-colony status, literacy rates, or
dependence on agriculture (Appendix Table A.2). We also controlled for trade to account for stronger
incentives to adopt social policies in open economies (e.g., Mares 2005), and our results are robust.
The results are retained when lagging all independent variables, substituting BMR with the continuous Polity index, controlling for many autocracies being Communist regimes, and running the models
on post-1945 samples only (see Appendix).12
Models A4–A6 resemble A1–A3, but only include countries without pension system in t − 1 and
have introduction of pension system in t as dependent variable. These models show that autocracies
are no less likely to adopt pension systems than democracies; while BMR is always positive, the
t-values range from 1.1–1.3. Thus, there is no robust evidence that democracy is systematically
associated with having or adopting old-age pension systems. This corroborates H1.
These results do not, however, imply that the typical “Autocratic Welfare State” is identical to the
democratic. Indeed, our argument provided several indications on how they may differ. First, since
autocratic coalitions often only comprise particular groups, some programs, other than pensions, may
be less common in autocracies. For example, autocracies may less often implement unemployment
insurance programs, because the main beneficiaries – unemployed workers – are presumably often
outside the autocrat’s coalition. Instead, employment protection legislation might fulfill the autocrat’s
need to insure and win over targeted groups of high income workers, without redistributing to groups
posing little risk to regime security.
Models B1–B5, Table 2 then also show less support for a “generalized version” of H1 applying
to all types of social policies rather than pensions only. When running models resembling A3 from
Table 1, but on existence of (above minimum-threshold) unemployment benefits-, maternity leave-,
family allowance-, work injury-, and sickness benefits programs, democracy is positively associated
with three of them: BMR is significant at 5 percent for unemployment benefits (B1), work injury (B4),
and sickness leave (B5). These policies are presumably less easy (than pension programs) to use for
12
However, there are indications that democracies more often have pension systems once using the
DD measure from Cheibub, Gandhi and Vreeland (2010). This partly reflects the shorter sample (post1945) – t-values are higher (though still insignificant) for BMR post-1945. We also tested various
Cox duration model specifications. Here, the regime coefficient was sensitive to minor specification
changes, with many models showing no effect and some showing a positive effect of democracy.
17
Program
Democracy (BMR)
Ln GDP p.c.
Ln population
Ethnic fraction.
Urbanization
Size military
Resource dependence
Region dummies
Year dummies
N
Countries
B1
Unemployment
b/(t)
1.002**
(2.45)
1.184**
(2.45)
0.746***
(3.62)
-1.676
(-1.25)
0.015
(0.83)
0.260
(1.38)
-0.011
(-0.33)
Y
Y
7039
130
B2
Maternity
b/(t)
0.272
(0.91)
-0.209
(-0.58)
0.279
(1.26)
-1.052
(-1.03)
0.024**
(2.05)
0.106
(0.48)
-0.004
(-0.33)
Y
Y
7825
136
B3
Family allowance
b/(t)
0.289
(0.82)
0.272
(0.75)
0.092
(0.42)
-0.787
(-0.74)
0.039***
(2.84)
-0.303
(-1.42)
-0.049***
(-2.58)
Y
Y
6724
136
B4
Work injury
b/(t)
1.036***
(2.79)
-0.446
(-0.91)
0.002
(0.01)
-0.396
(-0.29)
0.004
(0.18)
0.347
(0.78)
0.011
(0.47)
Y
Y
7765
135
B5
Sickness
b/(t)
0.650**
(2.50)
0.002
(0.00)
0.560**
(1.97)
-2.390**
(-2.12)
-0.005
(-0.41)
0.020
(0.11)
0.009
(0.55)
Y
Y
7894
136
Table 2: Democracies, autocracies, and the existence of various welfare programs.
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Logit regressions with errors clustered on country, with existence of program above minimum standard
as dependent variables. Maximum time series extend from late 1880s–2004.
directing resources towards autocratic winning coalitions. However, democracy is not systematically
related to maternity leave or family allowance programs. These results are not predicted by our
theoretical argument (but does not contradict either). However, they could relate to other features
of many autocratic regimes, such as the active pro-natalist policies pursued in many autocracies to
increase child births and expand the population (and, autocracies do indeed have higher population
growth, see Przeworski et al. 2000, chapter 5).
Our argument also indicated that autocracies, depending on how they are institutionalized, differ
in incentives and capabilities to construct and commit to distributive social policy schemes. We employ data from Geddes, Wright and Frantz (2014), distinguishing regimes after who controls access
to offices and policy-making decisions. One could, for example, expect dominant-party regimes to
more often have pension systems than the other autocracy types (personalist, military, monarchies),
due to their typically larger supporting coalitions. We thus ran models on probability of having and
adopting pension systems, but substituting BMR with the Geddes et al. dummies. These models
(Appendix Table A.8) have “pure” dominant-party regimes as reference category (hybrid dominantparty regimes, such as “dominant-party–military”, receive their own dummy). All autocratic regime
dummies are, indeed, consistently negative. Also the democracy dummy is negative more often than
not. Thus, the point estimates indicate that dominant-party regimes more often have, or adopt, pen-
18
sion programs. Surprisingly, however, personalist regimes are not systematically less likely to have
pension programs than dominant-party regimes. One potential reason, alluded to above, is that some
ruling parties may be so strongly institutionalized, providing clear checks on the leadership (e.g., Svolik 2012), that promises from regime leaders are credible even absent formalized pension programs.
In personalist regimes, fewer institutional checks on the ruler exist, and formalized pension programs
may be among the only available options for credibly committing to future distribution, even if they
are costly to set up for relatively small ruling coalitions. In contrast, the results are often statistically significant at conventional levels for the military regime and monarchy dummies, also when
controlling for dominant-party regimes more often being Communist.13
We also tested whether dominant-party autocracies more often have other social policy programs.
While the results are not clear, dominant-party regimes, generally, seem to correlate also with such
programs (Appendix Table A.11). For instance, dominant-party regimes more often have sickness
benefits than monarchies and personalist regimes, but there are no significant differences on unemployment.
5.2
Do autocracies adopt less universal pension programs than democracies?
Despite similar propensities for democracies and autocracies to have pension systems, our argument
indicates that these pension systems should differ in how they are structured. We expect such systems
to cover more groups in democracies. Indeed, the mean Universalism-Index score for pensions is
3.4 for democracies and 1.5 for autocracies. In post-WWII samples, the respective numbers are 4.4
and 3.0. Yet, income level, or country-specific historical factors, could affect both regime type and
pension features. We thus run different models, controlling both for year- and country-specific effects,
to test H2. We use linear models for our baseline specifications, but in the appendix we report various
robustness tests, including ordinal logit and probit specifications and logit and probit models using
dummies based on various cut-offs for the index, and these provide the same result (i.e., support for
H2).
Models C1–C2, Table 3, report standard fixed effects specifications with errors clustered on country. C1 controls also for income level, and the point estimate indicates that going from autocracy to
democracy increases the pension system Universalism Index (ranging from 0–9) with 0.5 (t=2.03), a
13
The results are retained for existence of pension programs, although not for adoption, when us-
ing the original dominant-party dummy (including hybrid regimes) from Geddes et al. as reference
category.
19
Democracy (BMR)
Ln GDP p.c.
C1
Fixed Effects
b/(t)
0.528**
(2.03)
0.333
(1.31)
Ln population
Urbanization
Size military
Resource dep.
C2
Fixed Effects
b/(t)
0.128
(0.61)
-0.005
(-0.02)
-0.340
(-1.22)
0.056***
(5.36)
0.025
(0.25)
-0.008
(-1.17)
Lag dep. var.
Year dummies
Y
N
Countries
Y
Y
10727
139
Y
Y
8128
138
C3
ECM
b/(t)
0.068***
(2.70)
0.051**
(2.18)
-0.069***
(-9.13)
Y
Y
10600
139
C4
ECM
b/(t)
0.052*
(1.82)
0.040
(1.28)
-0.009
(-0.34)
0.005***
(3.71)
0.025
(1.57)
-0.000
(-0.06)
-0.091***
(-9.81)
Y
Y
8057
138
C5
System GMM
b/(t)
0.088**
(2.10)
0.031
(0.78)
0.932***
(53.22)
Y
Y
10600
139
C6
System GMM
b/(t)
0.092**
(1.98)
0.017
(0.26)
0.005
(0.30)
0.001
(0.42)
-0.000
(-0.01)
0.003**
(2.07)
0.923***
(44.79)
Y
8057
138
Table 3: Democracies, autocracies, and universalism of pension systems.
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Universalism Index for pensions is dependent variable. Errors are clustered on country. Maximum
time series extend from late-1880s–2004.
modest increase. We also tested a random effects specification, and the result was very similar (see
Appendix). However, BMR loses significance in C2, which adds population, urbanization, military
size and resource dependence as controls.
However, to further limit endogeneity concerns (for example due to targeted pension programs
being conducive to autocracy), we run Error Correction Models (ECM). In addition to country-fixed
effects, these models control for the lagged dependent variable (and lag the independent). Hence, C3
and C4 investigate whether regime type in t − 1 affects changes in universalism from t − 1 to t. This
is, indeed, the case, both according to C3 (t=2.70) and the more extensive C4 (t=1.82).14 We also
run a more complex dynamic panel model, namely the system GMM estimator from Blundell and
Bond (1998), which is particularly suitable for analyzing slow-moving variables (such as regime type
14
Nevertheless, social policy programs are fairly stable. For pensions, 94% of our observations show
no change in the Universalism index. Eligibility is also extended more often than withdrawn. 90.9
percent of all changes in democracies were expansions, and 90.1 percent in autocracies. As expected
from our argument, few regimes dare scale down pension benefits to groups already receiving them.
These instances then tend to appear right after a new regime takes power (our data show that the
relative frequency is almost three times higher in the first three years after regime inception), and
likely concern groups only perceived critical to the former regime’s survival (see also Albertus and
Menaldo 2012).
20
and pension programs). As for ECM, system GMM models account for country-specific effects, and
include lagged dependent variables. These models, both the parsimonious (C5; t=2.10) and extensive
(C6; t=1.98), basically replicate the ECM results.15 Thus going from autocracy to democracy is
associated with subsequent increases in the universalism of old-age pensions.
In sum, democracy is positively associated with level of universalism (fixed effects) and even
more clearly with subsequent positive changes in universalism score (ECM and GMM). Autocracies
thus target their pension spending towards fewer social groups than democracies do. However, on
this account, our argument did not imply that the association is likely restricted to pensions; the
electoral and other political dynamics in democracies may induce them to promote more universal
programs across the board. To provide an overall evaluation, we construct a simple, additive Welfare
State Universalism Index (WSUI), which sums the Universalism Indices for all 6 program types (thus
ranging from 0–54). Although not completely robust, the result is generally corroborated also for
WSUI. Appendix Table A.12 presents six models equivalent to those in Table 3 but with WSUI
as dependent variable. The results are qualitatively similar to the pension results, with all models,
except the extensive fixed effects model (C8), yielding significant, positive BMR coefficients. Thus,
democracies have more universal social policies, in general, than autocracies.
In extension, we investigated whether dominant-party autocracies have less universal pension
programs than other autocracies, and we further tested models using WSUI to check for differences
between dictatorships regarding overall level of universalism. The results are reported in Appendix
Table A.19 (pensions) and A.20 (WSUI). We do not find solid evidence of systematic differences
between the autocracy types. This may be considered a surprising null-finding not only in light of our
argument, but also since the wider literature on varieties of autocracy highlight the mass support base
and broader supporting coalitions of party-based autocracies (e.g., Geddes 1999; Magaloni 2006;
Fjelde 2010; Levitsky and Way 2010). To be fair, the point estimates suggest that dominant-party
regimes, on average, have more universal policies, including pension programs, but the coefficients
15
The results are fairly robust to lagging all independent variables one additional year. We also
tested system GMM models instrumenting for levels of the endogenous independent variable (regime
type) by using lagged differences, and for differences in regime type by using lagged levels (see
Roodman 2009). To mitigate the “too-many-instruments” problem, we limited the number of lags
used for instrumentation to 4. The point-estimates were basically similar to the more efficient C5 and
C6, but the t-values dropped to 1.5–1.6 (see Appendix).
21
are mostly non-significant. There are some exceptions; System GMM models treating regime type
as endogenous report that pure dominant-party regimes have more universal pension programs than
monarchies, personalist regimes and hybrid-party regimes. Nevertheless, the lack of robustness means
that we cannot conclude on any systematic difference between different types of dictatorships in terms
of welfare state universalism.
5.3
Do pension programs reduce risks of autocratic regime breakdown?
If pension systems constitute the survival tool that we propose, they should reduce the probability
of autocratic breakdown. We thus test H3 by employing data from Geddes et al. (2014) on autocratic
regime failure. These data draw on a regime definition emphasizing formal and informal rules for selecting policies and leaders, and the measure captures changes from autocracy to democracy, between
different autocracy types, and even between regimes of similar type (e.g., between the Personalist,
but distinct, Mobutu and Kabila regimes in Zaire/Congo). Initially, we employ a naive assumption
and consider pension programs as exogenous – despite the evidence in Table 1 showing they are endogenous to several factors, and our argument indicating that fragile autocratic regimes should have
stronger incentives to adopt (costly) pension systems. We run logit and probit models with regime
failure as dependent variable, controlling for ethnic fractionalization, GDP per capita, population,
military size and resource dependence. We also include the Geddes regime dummies, since autocracy
type may affect probability of adopting pension systems, as seen above, and regime survival (Geddes
1999; Hadenius and Teorell 2007). We include log regime duration in some models; older regimes are
generally at lower risk of being deposed (e.g., Svolik 2012). However, tenure length may partly stem
from historical existence of pension programs, which is strongly correlated with current programs.
Hence, including regime duration may induce post-treatment bias, and we thus also report models
without it.
These models are reported in Appendix Table A.22. Given the plausible argument on how pension programs bolster autocratic regimes through serving as a co-optation device allowing both for
targeting and credible commitments, the lack of statistically significant association between pension
programs and regime failure might, at first glance, be surprising. Is there no effect on autocratic
survival, after all? Such a conclusion would, however, not only be premature, but incongruent with
the argument H3 is drawn from. We expected autocrats to implement expensive pension programs
where the regime perceives its future survival is threatened. In such situations, the regime may initi-
22
ate pension programs to shore up coalition support, or co-opt new groups. Pension programs should
thus tend to exist in contexts where probability of regime survival is, initially, low. Even if pension
programs should have a positive causal effect on regime survival, they may have zero correlation.
Hence, we must model the “selection” of pension programs and the effect of these programs
on regime survival simultaneously. There is no established duration model set-up that allows us to
properly model endogenous independent variables, and we did therefore not pursue any tests using
such models here. We rather employ Instrumental Variables Probit (IVProbit) models to deal with
the vital endogeneity question. Since this model can be fairly sensitive to specification choices, we
test different combinations of controls and instruments, and also run (more robust) 2SLS models.
To produce consistent estimates of the causal effect, we must identify instruments that are strongly
correlated with pension programs, and that are not directly related to regime failure. Drawing on
an insight from the literature on institutions and economic outcomes (Persson and Tabellini 2003;
Knutsen 2011; Huber, Ogorzalek and Gore 2012; Acemoglu et al. 2014) we construct instruments
tapping exogenous regional (and global) variation in the endogenous independent variable.
The underlying notion – backed up by literature on cross-border learning and other diffusion
mechanisms pertaining to social policy design (e.g., Obinger, Schmitt and Starke 2012) – is that
regimes residing in neighborhoods where most countries have pension programs are more likely to
adopt and continue such programs, everything else equal. While some cross-country variation in
adoption, as argued, stems from perceived regime threats and strategic calculation on the part of the
regime, this is, of course, not the only source of variation. We aim at isolating the (exogenous) share
of variation not affected by strategic adoption; Weyland (2005, 2007), for instance, highlights how
various cognitive biases may make leaders more likely to adopt regionally “common models” of such
policies, inducing cross-border “learning”. Further, once controlling for the other covariates (with one
caveat related to neighboring regime breakdowns accounted for below), neighboring country pension
programs should not directly impact on regime failure. The latter point point is important, as the
exclusion restriction is conditional on the control variablesincluded, and we test models controlling
for different covariate sets, for instance including time-trend/year-fixed effects, region-/country-fixed
effects and regime age. Hence, our instrument – percentage share of other countries in the region (as
defined by Miller 2013) having a pension program in that year – is expectedly a strong instrument
and – while this can never be fully ensured – should stand a good chance of satisfying the exclusion
23
restriction. Importantly, these propositions are clearly supported by diagnostics tests on instrument
strength and the exclusion restriction, which are presented below. In some models, we add share
of (other) countries globally with pension programs, to pick up exogenous global trends, but this
instrument is weaker.
24
25
3248
98
R
147.3
16.4
3248
98
R
148.6
16.4
0.000
(0.08)
-0.867**
(-2.33)
0.724***
(5.66)
-0.374*
(-1.71)
0.444***
(4.39)
0.389
(1.25)
-0.093
(-1.16)
-0.044
(-1.25)
-0.069
(-0.40)
0.008*
(1.91)
-0.145*
(-1.86)
-0.000
(-0.06)
0.095**
(2.15)
D2
IVProbit
b/(t)
3248
98
R+G
0.30
81.0
19.9
0.002
(0.41)
-0.735**
(-2.18)
0.650***
(6.10)
-0.275
(-1.39)
0.398***
(4.46)
0.479*
(1.75)
-0.079
(-1.08)
-0.035
(-1.08)
-0.090
(-0.56)
0.007*
(1.80)
-0.134*
(-1.78)
-0.001
(-0.14)
D3
IVProbit
b/(t)
3248
98
R+G
0.27
80.1
19.9
0.000
(0.04)
-0.831**
(-2.19)
0.731***
(5.82)
-0.360*
(-1.67)
0.445***
(4.42)
0.395
(1.27)
-0.094
(-1.20)
-0.045
(-1.30)
-0.073
(-0.42)
0.008*
(1.86)
-0.145*
(-1.86)
-0.000
(-0.08)
0.094**
(2.12)
D4
IVProbit
b/(t)
3247
98
R+G
0.48
81.3
19.9
2.758
(1.43)
1.226**
(2.00)
0.003
(0.69)
-0.720**
(-2.11)
0.656***
(6.24)
-0.256
(-1.30)
0.411***
(4.55)
0.468*
(1.72)
-0.083
(-1.12)
-0.028
(-0.85)
-0.096
(-0.60)
0.007*
(1.75)
-0.113
(-1.50)
-0.001
(-0.19)
D5
IVProbit
b/(t)
3247
98
R+G
0.45
80.5
19.9
-0.809**
(-2.13)
0.730***
(5.94)
-0.335
(-1.57)
0.454***
(4.47)
0.393
(1.28)
-0.097
(-1.22)
-0.037
(-1.07)
-0.080
(-0.47)
0.008*
(1.80)
-0.124
(-1.59)
-0.001
(-0.13)
0.087**
(2.00)
2.644
(1.38)
1.191*
(1.96)
0.001
(0.33)
D6
IVProbit
b/(t)
3247
98
R+G
0.48
81.3
19.9
Y
1.750
(1.02)
0.655
(1.18)
0.017***
(3.44)
-2.481***
(-4.50)
0.182
(0.68)
-0.427
(-1.64)
0.148
(0.80)
0.130
(0.47)
-0.055
(-0.49)
0.152**
(1.98)
0.169
(0.58)
0.010*
(1.78)
-0.011
(-0.13)
0.002
(0.55)
D7
IVProbit
b/(t)
3247
98
R+G
0.45
80.5
19.9
Y
-2.551***
(-4.74)
0.251
(0.84)
-0.501*
(-1.90)
0.185
(0.91)
0.057
(0.20)
-0.065
(-0.55)
0.145*
(1.86)
0.181
(0.60)
0.010*
(1.81)
-0.020
(-0.24)
0.002
(0.55)
0.080
(1.45)
1.566
(0.93)
0.618
(1.14)
0.015***
(2.98)
D8
IVProbit
b/(t)
53.3
16.4
3120
98
R
Y
Y
-3.119***
(-14.37)
-0.190
(-0.63)
-0.475*
(-1.94)
0.004
(0.02)
-0.147
(-0.50)
-0.000
(-0.00)
0.222***
(3.31)
0.190
(0.55)
0.013**
(2.25)
0.008
(0.08)
0.001
(0.22)
D9
IVProbit
b/(t)
52.6
16.4
3120
98
R
Y
Y
-3.141***
(-16.21)
-0.147
(-0.44)
-0.520**
(-2.11)
0.025
(0.11)
-0.203
(-0.70)
-0.005
(-0.04)
0.216***
(3.19)
0.194
(0.56)
0.013**
(2.25)
0.003
(0.03)
0.001
(0.26)
0.054
(0.85)
D10
IVProbit
b/(t)
3283
97
R+G
0.42
82.5
19.9
Y
0.380*
(1.88)
0.127*
(1.77)
0.003***
(2.98)
0.002**
(2.54)
-0.026***
(-2.61)
0.001
(1.32)
-0.196**
(-2.50)
0.077***
(3.63)
0.075**
(2.40)
0.032*
(1.70)
0.132
(1.30)
-0.036**
(-2.34)
-0.051
(-1.08)
D11
FE2SLS
b/(t)
3283
97
R+G
0.37
81.7
19.9
Y
0.002***
(2.63)
-0.028***
(-2.78)
0.001
(1.08)
0.028***
(4.56)
0.360*
(1.78)
0.130*
(1.82)
0.002*
(1.88)
-0.207***
(-2.63)
0.101***
(4.64)
0.032
(0.99)
0.043**
(2.24)
0.136
(1.34)
-0.044***
(-2.88)
-0.041
(-0.88)
D12
FE2SLS
b/(t)
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. R=share other countries in region having pension program that year; G=share other countries globally having pension program that year. All diagnostic tests for IVProbit models
are conducted on structurally equivalent (fixed effects) 2SLS models. The first-stage regressions are reported in Appendix Table A.23. Errors are clustered on country in IVProbit models. Maximum time series are 1946–2004.
Table 4: Second-stage IV regressions with autocratic regime failure as dependent variable and existence of old-age pension programs as endogenous
independent.
N
Countries
Instruments
Sargan p-value
Cr.-Donald–Wald F-stat
Weak Id. crit.value (10%)
Year dummies
Region dummies
Country dummies
Time trend
Region regime failures
Global regime failures
Ln regime duration
Resource dependence
Size of military
Urbanization
Ethnic fractionaliz.
Ln population
Ln GDP p.c.
Other autocracy
Personalist regime
Monarchy
0.002
(0.46)
-0.770**
(-2.32)
0.643***
(5.90)
-0.288
(-1.43)
0.397***
(4.43)
0.474*
(1.74)
-0.077
(-1.04)
-0.034
(-1.03)
-0.086
(-0.53)
0.007*
(1.85)
-0.133*
(-1.78)
-0.001
(-0.12)
Pension program
Military regime
D1
IVProbit
b/(t)
Model
Table 4 reports the second-stage results from twelve IV specifications. D1 and D2 (adding log
regime duration to D1) only use the “region-share” instrument. This instrument is, quite naturally,
highly collinear with the region dummies, but also with the year dummies. D1 and D2 omit these
dummies, but include a linear time trend. Both models report a clear negative causal effect (significant
5 percent) of having pension programs on regime failure. Further, the region-share t-value in the first
stage is very high (> 5), and weak identification tests show the instrument is sufficiently strong.16
Yet, including only one instrument disallows overidentification tests for checking the exclusion
restriction. Thus, we add the “global-share” instrument in D3 and D4. The results are retained, and
the Sargan p-values (around .3) indicate the exclusion restriction holds, suggesting that we may report
consistent estimates of a causal effect.
However, one could still, on theoretical grounds, expect the exclusion restriction to be violated
if pension programs regionally and globally correlate with incidences of regime failure regionally
and globally, and regime failures abroad directly impact on risk of regime failure domestically. To
exclude this causal pathway affecting our specification and results, we control for shares of (other)
countries regionally and globally experiencing regime failure (in given year) in D5 and D6. This does
not alter the estimated effect of pension programs on regime survival, but, as expected, the Sargan
p-values increase even further, thereby increasing our confidence that the exclusion restriction holds.
D7 and D8 include region-fixed effects. While this clearly weakens the first-stage t-value of our
regional instrument (to, respectively, 1.1 and 2.1) due to collinearity, the absolute t-values for pensions
on survival increase, to 4.5 and 4.7 respectively. We also tried substituting the time trend in D7 and D8
with year dummies, but these models did not converge. After omitting the regional and global regime
failure variables, and using only the regional instrument, the IVprobit regressions ran also with both
region and year dummies. These models (D9 and D10) lead to exceedingly high collinearity with the
instrument in the first stage, and we thus put less trust in these results. Still, if we are to believe them,
there is a very strong effect of pension programs on regime survival (t < −14). Finally, D11 and
D12 are linear fixed effects 2SLS specification using both the regional and global instruments. These
are conservative models, controlling for country-specific effects, and they report a causal effect of
16
The first-stage results (Appendix Table A.23) also indicate that urbanization correlates with pen-
sion programs. This coheres with our argument; urban societies present more difficult challenges to
autocrats, as the concentration of people within large cities eases different opposition coordination
problems. Hence, autocrats have stronger incentives to adopt pension programs in urban societies.
26
Italy 1925
Argentina 1952
Thailand 1973
China 2000
0,5
0,45
0,4
0,35
0,3
0,25
0,2
0,15
0,1
0,05
0
Figure 2: Predicted probabilities of regime breakdown (from Model D1) with 90% CIs.
All covariates set at actual values, except hypotheticals on pensions (Thailand-1973 did not have pension system, but the
other three did) and Italy-1925 (not coded by GWF) is coded as dominant-party regime.
pension programs significant at 1 percent. While this result does not hold when substituting the time
trend with year dummies, we did find a significant effect in other FE2SLS specifications. Although
the results are somewhat weakened (-1.2<t<-17.2 for D1–D12; see Appendix) when lagging all
independent variables one additional year, the finding is fairly robust. Ideally, we would test whether
withdrawing a pension program has a different effect on regime survival than introducing one. Our
argument indicated that withdrawing a pension program is particularly dangerous for autocrats. This
may well explain why so few autocracies have abolished pension programs, with the unfortunate sideeffect that we cannot estimate separate models on implementing and abolishing pensions. With this
caveat in mind, the results above still suggest that pension programs guard against regime breakdown
in autocracies, supporting H3.
Our models indicate not only statistically significant, but substantively large effects (although the
confidence intervals around the point predictions are admittedly wide, see Figure 2). For illustration,
Model D1 predicts the Mussolini regime had about 1.7 percent chance of breakdown in Italy-1925.
Had the regime not posited a pension program, however, the best estimate is 8.8 percent. To consider
another well-known example, our model indicates the personalist Perón regime was less consolidated,
27
with a 7.3 percent breakdown probability in Argentina-1952. However, our best guess of this probability is 24.8 percent if Perón had removed the Argentine pension system. Figure 2 also reports
the predicted probability in a (non-pension) observation which actually observed a breakdown of its
military regime, Thailand-1973, following mass protests and intervention by the Thai King. Model
D1 indeed predicts 24.9 percent chance of breakdown. If the Thai military had instituted a pension
program, D1 predicts only a 7.3 percent chance (remember that we interpret the estimated effects as
symmetrical only since we have insufficient variation to conduct separate analysis on abolishment
and introduction of pensions). Finally, China-2000 had about 2 percent chance of regime failure, but
this would have been above 10 percent without a Chinese pension system. Although these estimates
are very uncertain – evidenced by the wide confidence intervals – there is seemingly a large effect of
pension programs on regime survival.
Finally, we estimated the relationship between pension systems and democratic transitions. Acemoglu and Robinson’s (2006) model highlight the lack of tools for credible commitment as the key
explanation of democratization. According to our argument, however, pension programs allow autocrats to credibly commit to future distribution – thereby avoiding the fate of the pressured dictator
of Acemoglu and Robinson, who has to allow political liberalization. Whereas, for instance, Ansell
and Samuels (2014) expect targeted social policy programs to enhance the probability of democratization, we rather expect pension programs to reduce this probability. Indeed, we do find that
pension programs in autocracies inhibit democratization. We employ similar specifications as for
autocratic breakdown, modeling pension programs as endogenous using the same instrumentation
strategy. Models E1–E12 in Appendix Table A.26 thus resemble D1–D12 in Table 4 except for the
dependent variable; “democratization” is coded as changes from any autocracy to democracy in the
following year according to Geddes et al. (2014). 184 autocratic breakdowns were included in D1,
but only 68 democratization instances enter E1. Nonetheless, the results are strikingly similar; E1
indicates a clear negative effect of pension programs on democratization (t=-2.5). Again, we find
large substantive estimated effects – if pension programs had been removed from Mussolini Italy and
Perón Argentina, for example, the predicted probabilities of democratization increase from 0.3 to 4.4
percent (Italy-1925) and 2.3 to 18.2 percent (Argentina-1952).
Also the democratization result is robust; the 12 t-values in Appendix Table A.26 vary between
-2.0 and -16.4. The Sargan-test p-values are lower for the democratization models, but the hypothesis
28
that the exclusion restriction holds is never rejected at 5 percent. Although not robust, we also find
indications that pension programs inhibit democratization when using a BMR-based democratization
measure and running models (omitting controls with short time series) on the entire sample from late1880s–2004 (116 countries, 103 democratization experiences; see Appendix Table A.27). Hence,
pension systems in autocracies seem to lower chances of democratization. This result should be of
interest to the many scholars studying democratization, who, to our knowledge, have not yet systematically investigated the role of social policies in affecting transitions from autocracy to democracy.
6
Conclusion
Extensive social policy programs are not restricted to democracies; our new SPAW dataset allows
us to provide systematic evidence on the extent to which autocracies provide such programs. We
also present an argument for why they do so: Autocrats use social policy programs – and pension
programs in particular – to direct resources to critical supporting groups. Targeted pension programs
are often favored over private goods distribution, because they enable autocrats to credibly commit
also to future distribution, thus reducing the incentives of benefitting groups to overthrow the regime.
We tested implications from this argument on almost 140 countries with time series from the
1870s. We find clear patterns in the data that corroborate our argument. First, we do not find any
differences between democracies and autocracies in probabilities of adopting or having pension systems. However, democracies operate more universal programs, and autocracies are less likely to run
other programs, such as unemployment insurance, less suitable for directing resources to autocratic
coalitions. Extensions of this analysis provided indications that dominant-party autocracies more
often operate pension programs than military regimes and monarchies, but not more often than personalist regimes. Most notably, we find robust evidence for our perhaps most important theoretical
expectation, namely that pension programs have a positive effect on autocratic regime survival. In
extension, we also find evidence that having pension systems make autocracies less likely to democratize. Hence, implementing pension programs constitute an effective political survival strategy for
incumbent autocrats.
29
References
Acemoglu, Daron and James A. Robinson. 2006. Economic Origins of Dictatorship and Democracy.
New York: Cambridge University Press.
Acemoglu, Daron, Suresh Naidu, Pascual Restrepo and James A. Robinson. 2014. “Democracy Does
Cause Growth.” Cambridge, MA.: NBER Working Paper No. 20004.
Albertus, Michael and Victor Menaldo. 2012. “If you’re against them you’re with us: the effect of
expropriation on autocratic survival.” Comparative Political Studies 45(8):973–1003.
Alesina, Alberto, Arnaud Devleeschauwer, William Easterly, Sergio Kurlat and Romain Wacziarg.
2003. “Fractionalization.” Journal of Economic Growth 8(2):155–194.
Alesina, Alberto, Reza Baquir and William Easterly. 1999. “Public Goods and Ethnic Divisions.” The
Quarterly Journal of Economics 114(4):1243–1284.
Allan, James P. and Lyle Scruggs. 2004. “Political Partisanship and Welfare State Reform in Advanced Industrial Societies.” American Journal of Political Science 48(3):496–512.
Ansell, Ben W. 2010. From the Ballot to the Blackboard: The Redistributive Political Economy of
Education. Cambridge: Cambridge University Press.
Ansell, Ben W. and David J. Samuels. 2014. Inequality and Democratization: An Elite-Competition
Approach. Cambridge: Cambridge University Press.
Beck, Nathaniel, Jonathan N. Katz and Richard Tucker. 1998. “Taking Time Seriously: Time-SeriesCross-Section Analysis with a Binary Dependent Variable.” American Journal of Political Science
42(4):1260–1288.
Blundell, Richard and Stephen Bond. 1998. “Initial conditions and moment restrictions in dynamic
panel data models.” Journal of Econometrics 87(1):115–143.
Boix, Carles. 2003. Democracy and Redistribution. Cambridge: Cambridge University Press.
Boix, Carles, Michael Miller and Sebastian Rosato. 2013. “A Complete Data Set of Political Regimes,
1800-2007.” Comparative Political Studies 46(12): 1523–1554.
Boix, Carles and Milan W. Svolik. 2013. “The Foundations of Limited Authoritarian Government:
Institutions, Commitment, and Power-Sharing in Dictatorships.” Journal of Politics 75(2):300–316.
Bueno de Mesquita, Bruce, Alistair Smith, Randolph M. Siverson and James D. Morrow. 2003. The
Logic of Political Survival. Cambridge, Ma.: MIT Press.
Carter, David B. and Curtis S. Signorino. 2010. “Back to the Future: Modeling Time Dependence in
30
Binary Data.” Political Analysis 18(3):271–292.
Cheibub, José Antonio, Jennifer Gandhi and James Vreeland. 2010. “Democracy and dictatorship
revisited.” Public Choice 143:67–101.
Chetty, Rai. 2006. “A New Method of Estimating Risk Aversion.” American Economic Review
96(5):1821–1834.
Desai, Raj M., Anders Olosgård and Tarik M. Yousef. 2009. “The Logic of Authoritarian Bargains.”
Economics & Politics 21(1):93–125.
Esping-Andersen, Gøsta. 1990. The three worlds of welfare capitalism. Princeton: Princeton University Press.
Fjelde, Hanne. 2010. “Generals, Dictators, and Kings: Authoritarian Regimes and Civil Conflict,
1973–2004.” Conflict Management and Peace Science 27(3):195–218.
Flora, Peter and Arnold Joseph Heidenheimer. 1981. The development of welfare states in Europe
and America. Transaction Publishers.
Frantz, Erica and Andrea Kendall-Taylor. 2014. “A dictator’s toolkit: Understanding how co-optation
affects repression in autocracies.” Journal of Peace Research Forthcoming.
Gandhi, Jennifer. 2008. Political Institutions under Dictatorship. Cambridge: Cambridge University
Press.
Geddes, Barbara. 1999. “What Do We Know About Democratization After Twenty Years?” Annual
Review of Political Science 2:115–144.
Geddes, Barbara, Joseph Wright and Erica Frantz. 2014. “Autocratic Breakdown and Regime Transitions: A New Data Set.” Perspectives on Politics 12(2):313–331.
Gehlbach, Scott and Philip Keefer. 2011. “Investment without democracy: ruling-party institutionalization and credible commitment in autocracies.” Journal of Comparative Economics 39(2):123–
139.
Gehlbach, Scott and Philip Keefer. 2012. “Private Investment and the Institutionalization of Collective
Action in Autocracies: Ruling Parties and Legislatures.” Journal of Politics 74(2):621–635.
Haber, Stephen and Victor Menaldo. 2011. “Do natural resources fuel authoritarianism? A reappraisal
of the resource curse.” American Political Science Review 105(1):1–26.
Hadenius, Axel and Jan Teorell. 2007. “Pathways from Authoritarianism.” Journal of Democracy
18(1):143 –156.
31
Haggard, Stephan and Robert R. Kaufman. 2008. Development, Democracy, and Welfare States:
Latin America, East Asia, and Eastern Europe. Princeton: Princeton University Press.
Harding, Robin and David Stasavage. 2014. “What Democracy Does (and Doesnât Do) for Basic
Services: School Fees, School Inputs, and African Elections.” Journal of Politics 76(1):229–245.
Huber, Evelyne and John D. Stephens. 2001. Development and crisis of the welfare state: Parties and
policies in global markets. Chicago: University of Chicago Press.
Huber, John D., Thomas K. Ogorzalek and Radhika Gore. 2012. “Democracy, targeted redistribution
and ethnic inequality.” New York: Columbia University. Working Paper.
Huntington, Samuel P. 1991. The Third Wave: Democratization in the Late Twentieth Century. Norman: University of Oklahoma Press.
Iversen, Torben. 2005. Capitalism, democracy, and welfare. Cambridge: Cambridge University Press.
Iversen, Torben and David Soskice. 2009. “Distribution and redistribution: The shadow of the nineteenth century.” World Politics 61(3):438–486.
Kim, Wonik. 2007. “Social Insurance Expansion and Political Regime Dynamics in Europe, 1880–
1945.” Social Science Quarterly 88(2):494–514.
Knutsen, Carl Henrik. 2011. “Democracy, Dictatorship and Protection of Property Rights.” Journal
of Development Studies 47(1):164–182.
Korpi, Walter. 1983. The democratic class struggle. London: Routledge & Kegan Paul.
Kuran, Timur. 1989. “Sparks and Prairie Fires: A Theory of Unanticipated Political Revolution.”
Public Choice 61(1):41–74.
Lake, David A. and Matthew A. Baum. 2001. “The Invisible Hand of Democracy: Political Control
and the Provision of Public Services.” Comparative Political Studies 34(6):587–621.
Levitsky, Steven and Lucan A Way. 2010. Competitive Authoritarianism: Hybrid Regimes After the
Cold War. Cambridge: Cambridge Universtiy Press.
Lindert, Peter H. 2005. Growing Public: Social Spending and Economic Growth since the Eighteenth
Century. Volume 1. Cambridge: Cambridge University Press.
Maddison, Angus. 2007. Contours of the World Economy, 1-2030 AD. Oxford: Oxford University
Press.
Magaloni, Beatriz. 2006. Voting for Autocracy: Hegemonic Party Survival and its Demise in Mexico.
New York: Cambridge University Press.
32
Magaloni, Beatriz. 2008. “Credible Power-Sharing and the Longevity of Authoritarian Rule.” Comparative Political Studies 41(4-5):715–741.
Mares, Isabela. 2005. “Social Protection Around the World: External Insecurity, State Capacity, and
Domestic Political Cleavages.” Comparative Political Studies 38(6):623–651.
Mares, Isabela and Matthew E. Carnes. 2009. “Social Policy in Developing Countries.” Annual Review of Political Science 12:93–113.
Meltzer, Allan and Scott Richard. 1981. “A rational theory of the size of government.” Journal of
Political Economy 89(5):914–927.
Mesa-Lago, Carmelo. 1978. Social Security in Latin America: pressure groups, stratification, and
inequality. Pittsburgh: University of Pittsburgh Press.
Milanovic, Branko. 1998. Income, Inequality, and Poverty During the Transition from Planned to
Market Economy. Washington D.C.: The World Bank.
Miller, Michael K. 2013. “Democratic Pieces: Autocratic Elections and Democratic Development
since 1815.” British Journal of Political Science FirstView:1–30.
Moene, Kalle and Michael Wallerstein. 2001. “Inequality, Social Insurance, and Redistribution.”
American Political Science Review 95(4):859–874.
Mulligan, Casey B., Xavier Sala-i Martin and Ricard Gill. 2003. “Do Democracies Have Different
Public Policies than Nondemocracies?” Jounal of Economic Perspectives 18(1):51–74.
North, Douglass C. and Barry R. Weingast. 1989. “Constitutions and Commitment: The Institutions Governing Public Choice in Seventeenth Century Englad.” Journal of Economic History
44(4):803–832.
Obinger, Herbert, Carina Schmitt and Peter Starke. 2012. “Policy Diffusion and Policy Transfer in
Comparative Welfare State Research.” Social Policy & Administration 47(1):111–129.
Persson, Torsten and Guido Tabellini. 2003. The Economic Effect of Constitutions. Cambridge, MA:
MIT Press.
Pierson, Paul. 2000. “Increasing Returns, Path Dependence, and the Study of Politics.” American
Political Science Review 94(2):251–267.
Ponticelli, Jacopo and Hans-Joachim Voth. 2011. “Austerity and Anarchy: Budget Cuts and Social
Unrest in Europe, 1919-2008.” Working paper.
Przeworski, Adam, Michael E. Alvarez, José Antonio Cheibub and Fernando Limongi. 2000. Democ-
33
racy and Development. Cambridge: Cambridge University Press.
Rimlinger, Gaston Victor. 1971. Welfare Policy and Industrialization in Europe, America and Russia.
New York: Wiley.
Roodman, David. 2009. “How to do xtabond2: An introduction to difference and system GMM in
Stata.” STATA Journal 9(1):86–136.
Scruggs, Lyle. 2006. “The generosity of social insurance, 1971–2002.” Oxford Review of Economic
Policy 22(3):349–364.
Stasavage, David. 2005. “Democracy and Educational Spending in Africa.” American Journal of
Political Science 49(2):343–358.
Svolik, Milan W. 2012. The Politics of Authoritarian Rule. Cambridge: Cambridge University Press.
Weyland, Kurt. 2005. “Theories of Policy Diffusion: Lessons from Latin American Pension Reform.”
World Politics 57(2):262–295.
Weyland, Kurt. 2007. Bounded Rationality and Policy Diffusion: Social Sector Reform in Latin
America. Princeton: Princeton University Press.
Wintrobe, Ronald. 1998. The Political Economy of Dictatorship. Cambridge: Cambridge University
Press.
Wright, Joseph. 2008. “Do Authoritarian Institutions Constrain? How Legislatures Affect Economic
Growth and Investment.” American Journal of Political Science 52(2):322–343.
34
A
Online Appendix for The Autocratic Welfare State: Resource distribution,
credible commitments and political survival
This appendix contains additional information about the coding rules and sources used for con-
structing the variables from the SPAW dataset, and a number of robustness tests and extensions of
the empirical analysis presented in the paper. After presenting some more detail on the coding rules
and sources used for the coding of social policy programs, the remaining sections of the appendix are
organized such that they cover the robustness tests and extensions found in the different subsections
of the empirical analysis.
Hence, after the data discussion, we present (descriptive statistics and) additional analysis related
to the subsection that addresses whether regime type matters for the probability of having pension (and
other social policy) programs. These tests include models with alternative sets of control variables,
models run on shorter time series and models using alternative measures of democracy than the BoixMiller-Rosato measure that we use as our baseline. We also report results on autocracy types and
probability of having or adopting pensions, and of having the other major social policy programs.
Thereafter, we present additional analysis for the subsection on whether autocracies and democracies differ in terms of how universal or targeted their social policy programs are. This sub-section
include versions of the models on pension systems reported in the paper, and on the broader Universalism Index constructed over all the major social policy programs, using alternative estimation
techniques, such as random effects specifications and system GMM models that use instrumentation
strategies for modelling regime type as endogenous. Further, the sub-section contains different models that robustness test whether our assumptions regarding the measurement level of the universalism
index matters, including ordinal logit and probit models. We also report the tables on differences
between autocracy types in terms of universalism of pension programs and other social policies.
We finally present descriptive statistics and results for the subsection addressing whether pension
programs affect probability of regime breakdown and democratization. More specifically, we first
present the “naive” models that do not account for the fact that the adoption of pension programs is
endogenous, and which thus yield biased results. This sub-section also contains the first-stage results
from the IV probit and FE 2SLS models treating pension programs as endogenous, and thus providing
insight into which factors enhance the likelihood of having pension programs. Finally, it includes the
IV regression results on democratization.
i
A.1
SPAW - notes on coding rules and sources
In this section, we outline, in more detail, how our coding in SPAW differs from previous attempts
to code the introduction of major social policy laws. Further, we provide more information on how the
information used for constructing the Universalism Indices is coded. Finally, we report and discuss
the main sources used for coding the SPAW variables used in the paper.
The main differences in the scoring of existence of major social policy programs between the
SPAW and previous coding efforts (Cutright 1965; Collier and Messick 1975; Hicks 1995; Mares and
Carnes 2009) is the more precise definition of what constitutes a major law, what type of programs
are considered as social policy programs, and how to deal with federal states. We discuss how we
dealt with these and other issues below.
As discussed in the paper, we only consider a program as a major welfare law if it covers (at
least) one major social group. These groups are agricultural workers; industrial/production workers;
small-firm workers; self-employed; students; employers; temporary/casual workers; family/domestic
workers. The SPAW operationalization thus excludes from being considered as an above minimumlevel program the early social legislation that tended to be restricted to, for instance, miners, sailors,
postal workers, transport workers, and those employed in various forms of public service (military,
civil servants, judges and so on). As noted, we employ this criterion for coding social policy programs
divided into six areas: sickness leave; maternity leave; unemployment; disability/work injury; family
allowances; and, old-age pensions.
Another source of difference between the SPAW coding and other datasets concerns how to deal
with programs in federal states. Social policy legislation might be absent at the federal level, but be
prevalent at the state-level, but often far more so in some states (within a country) than others. It is
difficult to ensure cross-country comparability in the coding of federal-state legislation (in part due to
lack of comparable sources). It is not clear how previous databases have solved this problem. In order
to be consistent, we have thus only employed national legislation, meaning that state or provincial
legislation is not included in the coding. Quite naturally, one might worry that this underestimate
the extent of social regulation in federal states compared to in unitary states. For example, federal
family benefits in Switzerland was for a long time restricted to farmers only, with the various Cantons
having introduced measures to cover most urban workers. So, in effect, most categories of workers
ended up with coverage even if not trough the same legislation. At the same time, differences between
ii
local and national arrangements also exist in unitary states. For example, in Norway local pensions
covered most of the population long before a national pension system was put in place in 1936. It
is therefore not entirely certain that this omission only impacts federal states, as also unitary states
had encompassing local or regional initiatives. Coding (and aggregating) local regulations both for
federal and unitary states, down to the municipality level, is a tremendous task, and we thus consider
the choice to only include national legislation the best alternative, in practice.
One common source of disagreement on the introduction of major laws relates to previous studies
not separating benefits in kind systems from transfer programs. In order to provide a clear coding and
avoid any resulting conflation and misinterpretation, SPAW only codes the latter. It should be noted
that even previous datasets that have claimed to measure only transfer programs are sometimes doing
otherwise. For example, New Zealand is sometimes coded as having an unemployment program in
1930. Unfortunately, closer inspection with primary sources shows that New Zealand at that time had
only introduced an unemployment relief program, where unemployed workers were given public work
in exchange for basic assistance. Even if one accepts this as an unemployment insurance program,
benefits where not in payments, but instead in kind. New Zealand did not introduce a transfer program
before in 1938, an eight year difference between our score and the score from the previous measures.
Further, in the SPAW coding, we have focused on whether groups have the right to become members of an insurance, and not the stronger requirement that they are actively forced so by the government (as is the case in compulsory insurance). Hence, in our main variables used in the paper,
we make no difference between coverage of major groups resulting from coverage in a voluntary
program or whether they are given the option to insure voluntary in a program that is obligatory for
other groups. This explains why, for instance, our coded coverage level for Japanese programs is
more generous than what is sometime assumed; we take into account that even as compulsory insurance was restricted to urban workers in large firms, other categories of workers could voluntary opt
into coverage. We have, however, left out voluntary coverage programs without some form of state
subsidies.
In creating the Universalism Indexes, we drew on various sources, notably the ILO Legislative series (1919-), US Labor Department SSPTW-reports (1937-), and various national sources such as national law-databases (Scandinavian and Anglo-Saxon countries) and statistical yearbooks (Australian
Bureau of Statistics, Various; New Zealand Statistics, Various; Statistics Norway, Various). Previous
iii
data collection efforts such as De Mesa and Mesa-Lago (2006); Flora, Alber and others (1983,?);
Mesa-Lago (1978, 2007, 2008), and particularly the early M-series reports by the ILO (1922-1936)
have also been consulted. The distributions of the Universalism Indices for the six major programs
0
2
4
6
8
10
Universalism Accident Benefits
60
10 20 30 40 50 60 70 80
0
2
4
6
8
10
Universalism Sickness Benefits
0
10 20 30 40 50 60 70 80
0
2
4
6
8
10
Universalism Maternity Benefits
0
0
Percent
10 20 30 40 50 60 70 80
0
2
4
6
8
10
Universalism Old-Age Benefits
0
10
20
30
40
50
60
20
0
0
10
10
20
30
40
Percent
30 40
50
50
60
are shown in Figure A.1.
0
2
4
6
8
10
Universalism Unemployment Benefits
0
2
4
6
8
10
Universalism Familiy Benefits
Figure A.1: Distribution on Universalism Index for the six major social policy programs, all SPAW
observations.
In coding the Universalism Indices we had to make choices regarding what was to be considered as
proper coverage. In coding the extent of coverage, the SPAW variables make no distinction between
whether this coverage results from voluntary or compulsory insureance or private or social insurance,
or whether groups are covered in the same program or have their own program. The latter is important,
as coverage was usually extended to urban and rural workers in different programs in continental
European countries, with programs for the rural sector coming later than for the urban workers. One
issue of concern is that some countries – the present day unemployment insurance in the UK being
one example – have enacted two, or even up to three, major programs covering the same risk, for
instance with the first being a social insurance program and the second a means-tested program. In
order to deal with the presence of such dual-programs, we follow the same strategy as Mares (2005).
First, we code each independent program using the universalism metric, and then we take the average
of these program scores to arrive at the final universalism score (as observant readers will note from
Figure A.1, the universalism scores are not all integer values, and this is the reason). It is important to
note that in order to be considered as a dual or triple system, entitlements for the second program must
iv
be independent of the first program. A (second) program will not be coded as an independent program
if payment in this program is dependent on having received payments from the first program. This
follows from the fact that eligibility for the continuation benefit is dependent on first having qualified
for the first program. It is therefore hard to consider these as separate programs, which our scoring
reflects.
In addition, if a country has enacted a social insurance program covering industrial workers and
small-firm workers in urban areas and a second social insurance program for rural workers, we decided to count these groups as if they were covered in the same system. In this instance (if no other
groups were covered), the program would be given a universalism score of 4. An alternative solution
would have been to first score the two programs independently, and then take the average of two
programs as the final score. What this approach takes into account is the degree of segmentation, as
coverage was extended through different programs with different conditions and benefits instead of
covering both groups in the same program. Following the alternative approach would have resulted in
a universalism score of 2.5 (rural program = universalism score of 2; urban program = universalism
score of 3; final universalism score 2+3/2 = 2.5). The two approaches obvious leads to quite important differences in scoring, and while there may be good reasons to correct for segmentation, the
alternative options has larger drawbacks than benefits. In our example, a country would score higher
on universalism when not covering rural workers than when covering them under a separate program,
which clearly is too penalizing. We therefore opted for the first approach in coding the number of
groups covered to arrive at the final universalism score.
The perhaps primary source of bias in our coding is that “de jure sources” (and especially the
SSPTW) are not always explicit as to how many groups are covered. Sometimes the law specifies “all
employees” as eligible, with temporary and domestic workers included in the definition of employees,
whereas other times, “all employees” only refer to employees under long-term contracts, and not those
in domestic employment. Sometimes additional country sources can be used to determine which
definition of employees was used – and we have gone lengths to identify and assess such sources
– but in many instances the law itself is ambiguous. In these instances, we therefore developed the
following rule of thumb: when “only employees” are stated as covered, we coded the system as
covering wage and salaried workers in both industry and agriculture independent of firm size. When
instead the law states that “all employees” are covered, we coded the system as covering all forms
v
of wage and salaried work, including temporary, students and domestic workers. The only excluded
categories in these systems are self-employed and employers.
As a final word of caution we want to highlight that even if we find our coding procedures clearer
and more consistent than previous coding efforts conducted over the same numbers of countries, the
SPAW dataset can still be considered “work in progress”. As more sources become obtainable, and as
scholars and others correct and point out errors in our coding, we will update the SPAW dataset and
make these improved versions available to the public.
vi
A.2
Robustness tests and extensions for Section 5.1
Variable\Statistic
Mean
Standard deviation
Minimum
Maximum
Old-age pension system
Democracy (BMR)
Ln GDP per capita
Ln population
Ethnic fractionalization
Urbanization
Size of military
Resource dependence
0.71
0.39
8.11
9.00
0.42
37.19
0.63
4.12
0.45
0.48
1.02
1.48
0.25
24.25
0.69
9.55
0
0
5.38
5.33
0.00
0
0
0
1
1
11.08
14.06
0.93
100
12.73
100
Table A.1: Descriptive statistics for the 7881 observations (and variables) entering Model A3, Table
1 in the paper.
vii
Democracy (BMR)
Ln GDP p.c.
Ln population
Ethnic fractionaliz.
Urbanization
Size military
Total resources inc.
b/(t)
0.023
(0.06)
-0.010
(-0.04)
0.449**
(2.53)
-0.956
(-1.02)
0.066***
(4.04)
-0.003
(-0.02)
-0.000***
(-2.73)
Resource dependence
Civil War
b/(t)
-0.025
(-0.07)
-0.140
(-0.49)
0.431**
(2.43)
-1.029
(-1.07)
0.062***
(3.96)
-0.007
(-0.04)
b/(t)
0.080
(0.22)
-0.106
(-0.36)
0.413**
(2.36)
-0.857
(-0.91)
0.065***
(3.92)
-0.061
(-0.30)
b/(t)
0.055
(0.13)
-0.201
(-0.64)
0.414**
(2.28)
-1.004
(-1.06)
0.063***
(3.45)
-0.058
(-0.28)
b/(t)
-0.076
(-0.19)
-0.417
(-1.16)
0.423**
(2.16)
-0.940
(-0.92)
0.055***
(2.66)
-0.093
(-0.44)
b/(t)
-0.016
(-0.04)
-0.251
(-0.79)
0.342*
(1.84)
-1.330
(-1.49)
0.062***
(3.76)
-0.019
(-0.10)
0.001
(0.05)
-0.434
(-1.30)
0.003
(0.17)
0.002
(0.11)
0.002
(0.12)
-0.002
(-0.08)
British colony
-0.715
(-1.53)
Literacy
0.002
(0.12)
Agriculture (%)
-0.016
(-1.17)
Total trade
Region dummies
Year dummies
N
Countries
Y
Y
7586
137
Y
Y
7881
137
Y
Y
7881
137
Y
Y
7773
137
Y
Y
7556
137
0.000*
(1.79)
Y
Y
7457
136
Table A.2: Democracies, autocracies, and the existence of pension systems. Robustness testing by
including different control variables in Model A3 from Table 1.
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Logit regressions with errors clustered on country, with existence of minimum old-age pension system
as dependent variable. Year dummies, region dummies and constant are omitted from the table. The longest time series extend from the late 1880s
to 2004. Control variables are taken either from Miller (2013) or Haber and Menaldo (2011). The regions are: Eastern Europe/Soviet Union; Latin
America; North Africa/Middle East, sub-Saharan Africa; Western Europe/British settler colonies; East Asia; Southeast Asia/Pacific; South Asia
viii
Model
Democracy
Ln GDP p.c.
A1
b/(t)
0.189
(0.45)
0.515*
(1.74)
A2
b/(t)
0.213
(0.51)
0.689**
(2.45)
0.437***
(2.72)
-0.606
(-0.59)
Y
Y
9081
Y
Y
8224
Ln population
Ethnic fraction.
Urbanization
Size military
Resource dependence
Region dummies
Year dummies
N
A3
b/(t)
0.097
(0.24)
-0.193
(-0.62)
0.407**
(2.25)
-1.033
(-1.05)
0.067***
(3.93)
0.004
(0.02)
0.004
(0.22)
Y
Y
7770
A4
b/(t)
0.415
(1.22)
0.130
(0.76)
A5
b/(t)
0.565
(1.56)
0.193
(0.96)
0.305**
(2.34)
-0.261
(-0.42)
Y
Y
2108
Y
Y
1817
A6
b/(t)
0.486
(1.38)
-0.199
(-0.61)
0.297**
(2.11)
-0.685
(-0.97)
0.029**
(2.53)
0.119
(1.11)
0.002
(0.09)
Y
Y
1660
Table A.3: Democracies, autocracies, and existence (A1–A3) or adoption (A4–A6) of pension systems. Robustness testing when lagging all independent variables 1 year.
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Logit regressions with errors clustered on country. Existence of minimum old-age pension system
is dependent variable in A1–A3 and adoption in A4–A6 (here, countries already having systems are omitted). Year dummies, region dummies and
constant are omitted from table. Maximum time series (independent variables) extend from late 1880s–2003.
ix
Model
Democracy
Ln GDP p.c.
A1
b/(t)
0.559
(1.42)
0.203
(1.07)
A2
b/(t)
0.919**
(2.24)
0.401*
(1.66)
0.471***
(2.62)
-0.558
(-0.69)
Y
Y
Y
Y
A3
b/(t)
0.668
(1.58)
0.169
(0.44)
0.479**
(2.39)
-1.266
(-1.54)
0.032*
(1.84)
-0.131
(-0.49)
-0.013
(-0.50)
Y
Y
Y
Y
Y
824
724
668
Ln population
Ethnic fractionalization
Urbanization
Size military
Resource dependence
Region dummies
Year dummies
Linear time trend
Years w.o. pension dummies
Year w.o. pension cubic splines
N
A4
b/(t)
0.497
(1.45)
0.198
(1.02)
0.335***
(2.67)
-0.228
(-0.38)
Y
Y
Y
1316
A5
b/(t)
0.369
(1.10)
-0.099
(-0.30)
0.306**
(2.26)
-0.713
(-1.03)
0.031***
(2.58)
-0.045
(-0.45)
-0.010
(-0.45)
Y
Y
Y
1225
A6
b/(t)
0.309
(0.94)
0.059
(0.37)
A7
b/(t)
0.415
(1.27)
0.098
(0.53)
0.227**
(2.02)
0.443
(0.75)
Y
Y
A8
b/(t)
0.361
(1.13)
-0.142
(-0.45)
0.188
(1.63)
0.340
(0.52)
0.014
(1.32)
0.142
(1.31)
0.002
(0.12)
Y
Y
Y
Y
Y
5286
Y
4795
Y
3202
Table A.4: Democracies, autocracies, and adoption of pension systems. Robustness testing when
including dummies for years without pension systems or cubic splines on number of years without
pension systems to further account for temporal dependence.
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Logit regressions with errors clustered on country. Adoption of minimum old-age pension system
is dependent variable (countries already having systems are omitted). Year dummies, linear time trend region dummies, dummies for year without
pensions, cubic splines for year without pensions and constant are omitted from table. Maximum time series (independent variables) extend from late
1880s–2003. Please note that the models using dummies for year without pensions have very few observations when also including year dummies. We
also tested such models substituting the year dummies with a linear time trend, but these models did not converge. Also, the most parsimonious model
(only BMR, GDP p.c. and region dummies) that included year dummies and cubic splines for years without pension did not converge.
x
Model
Polity 2
Ln GDP p.c.
A1
b/(t)
0.019
(0.73)
0.562*
(1.89)
A2
b/(t)
0.017
(0.68)
0.747***
(2.65)
0.387**
(2.47)
-0.565
(-0.57)
Y
Y
9091
139
Y
Y
8316
137
Ln population
Ethnic fraction.
Urbanization
Size military
Resource dependence
Region dummies
Year dummies
N
Countries
A3
b/(t)
0.009
(0.34)
-0.039
(-0.13)
0.355**
(2.03)
-1.039
(-1.11)
0.060***
(3.91)
-0.045
(-0.21)
0.003
(0.18)
Y
Y
7886
137
A4
b/(t)
0.024
(1.05)
0.315
(1.64)
0.258**
(2.14)
-0.267
(-0.44)
Y
Y
1364
111
Table A.5: Democracies, autocracies, and the existence (A1–A3) or adoption (A4) of pension systems
(the two other models on adoption, resembling A1 and A3 did not converge). Robustness testing using
Polity 2 rather than BMR.
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Logit regressions with errors clustered on country, with existence of minimum old-age pension system
as dependent variable in A1–A3 and introduction of such a system as dependent in A4 (here, countries already having such systems are omitted). Year
dummies, region dummies and constant are omitted from the table.
xi
Model
Democracy (BMR)
Ln GDP p.c.
A1
b/(t)
0.331
(0.61)
0.409
(1.44)
A2
b/(t)
0.195
(0.41)
0.758***
(2.68)
0.537***
(3.19)
0.165
(0.15)
Y
Y
6366
Y
Y
6239
Ln population
Ethnic fraction.
Urbanization
Size military
Resource dependence
Region dummies
Year dummies
N
A3
b/(t)
0.126
(0.27)
0.149
(0.50)
0.583***
(2.97)
-0.288
(-0.28)
0.058***
(3.40)
-0.007
(-0.03)
-0.001
(-0.06)
Y
Y
6073
A4
b/(t)
0.692**
(2.06)
0.013
(0.07)
A5
b/(t)
0.628*
(1.78)
0.278
(1.36)
0.380***
(2.61)
0.068
(0.09)
Y
Y
772
Y
Y
772
A6
b/(t)
0.562
(1.61)
0.132
(0.41)
0.462***
(3.09)
-0.432
(-0.53)
0.019
(1.39)
0.012
(0.09)
0.001
(0.06)
Y
Y
760
Table A.6: Democracies, autocracies, and the existence (A1–A3) or adoption (A4–A6) of pension
systems. Robustness testing on restricted sample from 1946.
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Logit regressions with errors clustered on country, with existence of minimum old-age pension system
as dependent variable in A1–A3 and introduction of such a system as dependent in A4–A6 (here, countries already having such systems are omitted).
Year dummies, region dummies and constant are omitted from the table.
xii
Model
Democracy (BMR)
Communist regime
Ln GDP p.c.
A1
b/(t)
0.403
(0.74)
1.916
(0.91)
0.478
(1.62)
A2
b/(t)
0.262
(0.54)
1.330
(0.74)
0.794***
(2.69)
0.519***
(2.98)
0.316
(0.28)
Y
Y
6330
Y
Y
6203
Ln population
Ethnic fraction.
Urbanization
Size military
Resource dependence
Region dummies
Year dummies
N
A3
b/(t)
0.178
(0.38)
1.637
(0.95)
0.172
(0.56)
0.558***
(2.76)
-0.116
(-0.11)
0.064***
(3.45)
-0.140
(-0.55)
-0.002
(-0.15)
Y
Y
6038
A4
b/(t)
0.746**
(2.11)
0.981*
(1.77)
0.029
(0.16)
A5
b/(t)
0.687*
(1.84)
0.941*
(1.86)
0.302
(1.43)
0.387***
(2.60)
0.240
(0.32)
Y
Y
765
Y
Y
765
A6
b/(t)
0.654*
(1.72)
1.146**
(2.12)
0.164
(0.50)
0.468***
(3.09)
-0.257
(-0.31)
0.020
(1.41)
-0.075
(-0.53)
0.001
(0.08)
Y
Y
753
Table A.7: Democracies, autocracies, and the existence (A1–A3) or adoption (A4–A6) of pension
systems. Robustness testing when controlling for Communist regimes (post-1946 sample, using coding from Cheibub, Gandhi and Vreeland 2010).
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Logit regressions with errors clustered on country, with existence of minimum old-age pension system
as dependent variable in A1–A3 and introduction of such a system as dependent in A4–A6 (here, countries already having such systems are omitted).
Year dummies, region dummies and constant are omitted from the table.
xiii
Model
Democracy (DD)
Ln GDP p.c.
A1
b/(t)
1.025**
(2.06)
0.316
(1.09)
A2
b/(t)
0.807*
(1.71)
0.676**
(2.14)
0.497***
(2.89)
0.421
(0.36)
Y
Y
6232
Y
Y
6105
Ln population
Ethnic fractionaliz.
Urbanization
Size military
Resource dependence
Region dummies
Year dummies
N
A3
b/(t)
0.595
(1.33)
0.123
(0.39)
0.559***
(2.73)
-0.096
(-0.09)
0.057***
(3.46)
-0.100
(-0.37)
0.001
(0.03)
Y
Y
5969
A4
b/(t)
0.943***
(2.66)
0.047
(0.26)
A5
b/(t)
0.870**
(2.41)
0.333
(1.55)
0.412**
(2.54)
0.242
(0.32)
Y
Y
753
Y
Y
753
A6
b/(t)
0.795**
(2.05)
0.174
(0.54)
0.500***
(2.96)
-0.243
(-0.29)
0.019
(1.34)
0.012
(0.09)
0.003
(0.19)
Y
Y
741
Table A.8: Democracies, autocracies, and the existence (A1–A3) or adoption (A4–A6) of pension
systems. Robustness testing on restricted sample from 1946 and using the DD measure of democracy.
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Logit regressions with errors clustered on country, with existence of minimum old-age pension system
as dependent variable in A1–A3 and introduction of such a system as dependent in A4–A6 (here, countries already having such systems are omitted).
Year dummies, region dummies and constant are omitted from the table.
xiv
Model
Military
Monarchy
Personalist
Hybrid-party
Democracy (GWF)
Other/Non-regime
Ln GDP p.c.
A1
b/(t)
-0.842
(-1.30)
-1.683**
(-2.29)
-0.271
(-0.43)
-0.308
(-0.53)
0.455
(0.73)
0.009
(0.01)
0.405
(1.40)
A2
b/(t)
-1.320**
(-2.23)
-1.922***
(-2.63)
-0.427
(-0.73)
-0.468
(-0.81)
0.053
(0.09)
-0.243
(-0.39)
0.856***
(2.76)
0.737***
(3.65)
0.760
(0.63)
Y
Y
4789
110
Y
Y
4703
108
Ln population
Ethnic fract.
Urbanization
Size military
Resource dep.
Region dummies
Year dummies
N
Countries
A3
b/(t)
-1.162**
(-2.07)
-1.326*
(-1.68)
-0.290
(-0.50)
-0.351
(-0.62)
-0.053
(-0.10)
-0.522
(-0.92)
0.322
(1.04)
0.839***
(3.57)
-0.055
(-0.05)
0.054***
(3.42)
-0.026
(-0.10)
-0.000
(-0.02)
Y
Y
4594
108
A4
b/(t)
-1.150*
(-1.73)
-1.040
(-1.62)
-0.405
(-0.77)
-0.632
(-1.27)
-0.064
(-0.13)
-0.000
(-0.00)
0.058
(0.26)
A5
b/(t)
-1.246*
(-1.90)
-1.226*
(-1.88)
-0.429
(-0.78)
-0.843
(-1.61)
-0.246
(-0.51)
-0.221
(-0.20)
0.358
(1.46)
0.426**
(2.57)
0.579
(0.71)
Y
Y
743
75
Y
Y
743
75
A6
b/(t)
-1.062
(-1.57)
-0.956
(-1.42)
-0.260
(-0.46)
-0.724
(-1.38)
-0.213
(-0.41)
-0.854
(-0.56)
0.118
(0.31)
0.499***
(2.91)
0.172
(0.19)
0.018
(1.12)
0.016
(0.12)
0.008
(0.52)
Y
Y
731
74
Table A.9: Autocracy types and existence (A1-A3) or adoption (A4-A6) of pension systems.
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Logit regressions with errors clustered on country. Year dummies, region dummies and constant are
omitted from table. “Pure” dominant-party regime (hybrids categorized as separate dummy) is reference category. Maximum time series is 1946–2004.
xv
Model
Communist regime dummy
Hybrid dominant-party
Military
Monarchy
Personalist
Other/non-democracy
Democracy
Ln GDP p.c.
A1
b/(t)
0.955
(0.43)
-0.259
(-0.47)
-0.754
(-1.25)
-1.639**
(-2.24)
-0.184
(-0.30)
0.067
(0.10)
0.535
(0.89)
0.444
(1.53)
A2
b/(t)
0.395
(0.27)
-0.444
(-0.79)
-1.279**
(-2.17)
-1.902***
(-2.59)
-0.387
(-0.66)
-0.214
(-0.35)
0.093
(0.16)
0.867***
(2.80)
0.733***
(3.57)
0.778
(0.64)
4776
4690
Ln population
Ethnic fraction.
Urbanization
Size military
Resource dependence
N
A3
b/(t)
1.554
(1.11)
-0.195
(-0.35)
-0.952*
(-1.70)
-1.188
(-1.45)
-0.097
(-0.16)
-0.376
(-0.69)
0.127
(0.23)
0.350
(1.11)
0.829***
(3.51)
-0.026
(-0.02)
0.058***
(3.39)
-0.065
(-0.27)
-0.002
(-0.08)
4581
A4
b/(t)
0.505
(0.77)
-0.607
(-1.24)
-1.053
(-1.52)
-0.983
(-1.47)
-0.332
(-0.62)
0.045
(0.04)
0.000
(0.00)
0.070
(0.32)
A5
b/(t)
0.471
(0.75)
-0.816
(-1.58)
-1.156
(-1.62)
-1.183*
(-1.75)
-0.360
(-0.62)
-0.182
(-0.16)
-0.181
(-0.34)
0.374
(1.50)
0.426**
(2.56)
0.571
(0.69)
743
743
A6
b/(t)
0.790
(1.31)
-0.695
(-1.36)
-0.924
(-1.33)
-0.897
(-1.30)
-0.155
(-0.27)
-0.856
(-0.56)
-0.110
(-0.20)
0.159
(0.41)
0.503***
(2.94)
0.161
(0.18)
0.018
(1.07)
-0.016
(-0.12)
0.007
(0.52)
731
Table A.10: Autocracy types and existence (A1-A3) or adoption (A4-A6) of pension systems: Controlling for Communist regimes (using coding from Cheibub, Gandhi and Vreeland 2010).
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Logit regressions with errors clustered on country. Year dummies, region dummies and constant are
omitted from table. “Pure” dominant-party regime (hybrids categorized as separate dummy) is reference category. Maximum time series is 1946–2004.
xvi
Program (dep. var.)
Monarchy
Personalist
Hybrid party reg.
Military regime
Other/non-regime
Democracy
Ln GDP p.c.
Ln population
Ethnic fraction.
Urbanization
Size military
Resource dependence
Region dummies
Year dummies
N
Countries
Unemployment
b/(t)
-1.326
(-1.31)
0.133
(0.22)
0.225
(0.37)
0.210
(0.36)
-0.028
(-0.04)
0.370
(0.84)
0.035
(0.09)
0.350
(1.08)
0.059
(0.05)
0.020
(1.42)
0.009
(0.03)
-0.004
(-0.32)
Y
Y
5532
126
Maternity
b/(t)
-1.912**
(-2.04)
-1.199**
(-1.96)
-0.762
(-1.19)
-0.088
(-0.13)
-0.133
(-0.16)
-0.018
(-0.04)
0.038
(0.07)
0.686
(1.58)
-1.270
(-0.91)
-0.002
(-0.13)
-0.165
(-0.81)
0.008
(0.54)
Y
Y
5599
127
Family allowance
b/(t)
-5.716***
(-4.19)
-0.364
(-0.38)
-0.377
(-0.32)
0.962
(1.28)
1.200
(0.95)
1.161*
(1.77)
1.533***
(3.31)
0.662**
(2.39)
-0.497
(-0.29)
0.014
(0.61)
-0.066
(-0.32)
0.023
(1.07)
Y
Y
5316
121
Work injury
b/(t)
-1.420
(-0.70)
0.581
(0.44)
0.006
(0.00)
-0.500
(-0.40)
0.000
(.)
0.925
(0.87)
-0.593
(-0.82)
-0.304
(-0.72)
0.459
(0.25)
0.001
(0.03)
0.655
(0.85)
-0.021*
(-1.77)
Y
Y
5388
125
Sickness
b/(t)
0.223
(0.19)
0.729
(1.19)
0.140
(0.21)
0.494
(0.86)
0.738
(0.95)
0.944*
(1.73)
0.326
(0.83)
-0.007
(-0.03)
-0.641
(-0.56)
0.037**
(2.33)
-0.370
(-1.51)
-0.049**
(-2.51)
Y
Y
5645
128
Table A.11: Types of autocracies and the existence of various welfare programs.
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Logit regressions with errors clustered on country, with existence of program above minimum standard
as dependent variables. Year dummies, region dummies and constant are omitted from table. Maximum time series extend from 1946–2004.
xvii
Robustness tests and extensions for Section 5.2
0
5
10
15
20
25
A.3
1900
1920
1940
1960
1980
2000
Years
Universalism in Democracies
Universalism in All Regimes
Universalism in Autocracies
Figure A.2: Average score on Welfare State Universalism Index (aggregated over the six major social
policy programs; ranging from 0–54) – for democracies, for autocracies, and average for all observations – over time.
xviii
xix
Y
8128
138
0.128
(0.61)
-0.005
(-0.02)
-0.340
(-1.22)
0.056***
(5.36)
0.025
(0.25)
-0.008
(-1.17)
C2
Fixed Eff.
b/(t)
-0.069***
(-9.13)
Y
10600
139
0.068***
(2.70)
0.051**
(2.18)
C3
ECM
b/(t)
0.052*
(1.82)
0.040
(1.28)
-0.009
(-0.34)
0.005***
(3.71)
0.025
(1.57)
-0.000
(-0.06)
-0.091***
(-9.81)
Y
8057
138
C4
ECM
b/(t)
0.932***
(53.22)
Y
10600
139
0.088**
(2.10)
0.031
(0.78)
C5
Sys. GMM
b/(t)
0.092**
(1.98)
0.017
(0.26)
0.005
(0.30)
0.001
(0.42)
-0.000
(-0.01)
0.003**
(2.07)
0.923***
(44.79)
Y
8057
138
C6
Sys.GMM
b/(t)
Y
7781
129
3.042***
(3.33)
1.448
(1.29)
C7
Fixed Eff
b/(t)
Y
5857
127
1.029
(1.45)
-0.045
(-0.05)
-3.184*
(-1.89)
0.240***
(5.35)
-0.073
(-0.23)
-0.048**
(-2.12)
C8
Fixed Eff.
b/(t)
-0.046***
(-3.98)
Y
7616
129
0.237***
(3.15)
0.112
(1.41)
C9
ECM
b/(t)
0.144*
(1.94)
0.043
(0.44)
-0.154
(-1.08)
0.020***
(3.50)
0.033
(0.78)
-0.004
(-1.31)
-0.076***
(-4.33)
Y
5736
126
C10
ECM
b/(t)
0.892***
(19.35)
Y
7616
129
0.367**
(2.22)
1.056*
(1.87)
C11
Sys. GMM
0.335**
(1.97)
0.999
(1.29)
0.079
(0.56)
0.014
(1.36)
-0.079
(-0.98)
-0.014*
(-1.92)
0.876***
(13.67)
Y
5736
126
C12
Sys. GMM
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Universalism Index for pensions is dependent variable for C1–C6, and Welfare State Universalism Index (additive universalism score over all six policy programs) for C7–C12.
Errors are clustered on country. Maximum time series extend from late-1880s–2004.
Table A.12: Democracies, autocracies, and universalism of pension systems (C1–C6) or of social policy programs more generally (C7–C12).
Year dummies
N
Countries
Lag dep. var.
Resource dep.
Size military
Urbanization
Ln population
Y
10727
139
0.528**
(2.03)
0.333
(1.31)
Democracy (BMR)
Ln GDP p.c.
C1
Fixed Eff
b/(t)
Model
xx
Y
10727
139
0.532**
(2.09)
0.395*
(1.79)
A1
Random Eff.
b/(t)
Y
8128
138
0.141
(0.69)
0.001
(0.01)
-0.224
(-1.26)
0.054***
(5.40)
0.016
(0.16)
-0.008
(-1.28)
A2
Random Eff.
b/(t)
Y
10727
139
0.528**
(2.03)
0.333
(1.31)
A3
Fixed Eff
b/(t)
Y
8128
138
0.128
(0.61)
-0.005
(-0.02)
-0.340
(-1.22)
0.056***
(5.36)
0.025
(0.25)
-0.008
(-1.17)
A4
Fixed Eff.
b/(t)
-0.069***
(-9.13)
Y
10600
139
0.068***
(2.70)
0.051**
(2.18)
A5
ECM
b/(t)
0.052*
(1.82)
0.040
(1.28)
-0.009
(-0.34)
0.005***
(3.71)
0.025
(1.57)
-0.000
(-0.06)
-0.091***
(-9.81)
Y
8057
138
A6
ECM
b/(t)
0.932***
(53.22)
Y
10600
139
0.088**
(2.10)
0.031
(0.78)
A7
System GMM
b/(t)
0.092**
(1.98)
0.017
(0.26)
0.005
(0.30)
0.001
(0.42)
-0.000
(-0.01)
0.003**
(2.07)
0.923***
(44.79)
Y
8057
138
A8
System GMM
b/(t)
0.904***
(53.50)
Y
10600
139
0.070
(1.59)
0.083**
(2.39)
A9
System GMM
b/(t)
0.075
(1.51)
0.040
(0.71)
-0.030*
(-1.69)
0.001
(0.43)
0.011
(0.45)
0.003**
(2.39)
0.904***
(50.43)
Y
8057
138
A10
System GMM
b/(t)
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Regressions with universalism index score for old-age pensions as dependent variables. Errors are clustered on country. Year dummies, region dummies, country dummies and
constant are omitted from the table. In A9 and A10, democracy is treated as endogenous, with four lags used for instrumentation (see Roodman 2009). A3–A8 are also reported in the paper, but the other models are not. The
maximum time series extend from the late 1880s to 2004.
Table A.13: Democracies, autocracies, and the universalism of pension systems.
Year dummies
N
Countries
Lagged dep. var.
Resource dep.
Size military
Urbanization
Ln population
Ln GDP p.c.
Democracy (BMR)
Model
xxi
Y
7781
129
3.068***
(3.39)
1.636
(1.56)
A1
Random Eff.
b/(t)
Y
5857
127
1.053
(1.49)
0.015
(0.02)
-2.609*
(-1.89)
0.241***
(5.52)
-0.069
(-0.22)
-0.051**
(-2.26)
A2
Random Eff.
b/(t)
Y
7781
129
3.042***
(3.33)
1.448
(1.29)
A3
Fixed Eff
b/(t)
Y
5857
127
1.029
(1.45)
-0.045
(-0.05)
-3.184*
(-1.89)
0.240***
(5.35)
-0.073
(-0.23)
-0.048**
(-2.12)
A4
Fixed Eff.
b/(t)
-0.046***
(-3.98)
Y
7616
129
0.237***
(3.15)
0.112
(1.41)
A5
ECM
b/(t)
0.144*
(1.94)
0.043
(0.44)
-0.154
(-1.08)
0.020***
(3.50)
0.033
(0.78)
-0.004
(-1.31)
-0.076***
(-4.33)
Y
5736
126
A6
ECM
b/(t)
0.892***
(19.35)
Y
7616
129
0.367**
(2.22)
1.056*
(1.87)
A7
System GMM
b/(t)
0.335**
(1.97)
0.999
(1.29)
0.079
(0.56)
0.014
(1.36)
-0.079
(-0.98)
-0.014*
(-1.92)
0.876***
(13.67)
Y
5736
126
A8
System GMM
b/(t)
0.884***
(22.37)
Y
7616
129
0.162
(0.97)
1.129**
(2.48)
A9
System GMM
b/(t)
0.135
(0.73)
0.964*
(1.69)
0.129
(1.24)
0.018**
(2.01)
-0.057
(-0.77)
-0.013*
(-1.94)
0.874***
(17.04)
Y
5736
126
A10
System GMM
b/(t)
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Regressions with Welfare State Universalism Index (additive universalism score, aggregated over all six major policy programs) as dependent variable. Errors are clustered on
country. Year dummies, region dummies, country dummies and constant are omitted from the table. In A9 and A10, democracy is treated as endogenous, with four lags used for instrumentation (see Roodman 2009). E3–E8 are
also reported in the paper, but the other models are not. The maximum time series extend from the late 1880s to 2004.
Table A.14: Democracies, autocracies, and universalism of social policy programs.
Year dummies
N
Countries
Lagged dep. var.
Resource dep.
Size military
Urbanization
Ln population
Ln GDP p.c.
Democracy (BMR)
Model
xxii
Y
10345
0.564**
(2.12)
0.370
(1.64)
A1
Random Eff.
b/(t)
Y
7838
0.192
(0.90)
-0.045
(-0.20)
-0.212
(-1.19)
0.055***
(5.64)
0.015
(0.15)
-0.009
(-1.40)
A2
Random Eff.
b/(t)
Y
10345
0.558**
(2.06)
0.300
(1.16)
A3
Fixed Eff
b/(t)
Y
7838
0.176
(0.81)
-0.068
(-0.27)
-0.316
(-1.12)
0.057***
(5.53)
0.025
(0.24)
-0.009
(-1.31)
A4
Fixed Eff.
b/(t)
-0.066***
(-8.63)
Y
10216
0.068***
(2.68)
0.050**
(2.21)
A5
ECM
b/(t)
0.053*
(1.84)
0.038
(1.25)
-0.010
(-0.36)
0.005***
(3.53)
0.022
(1.43)
0.000
(0.06)
-0.087***
(-8.88)
Y
7761
A6
ECM
b/(t)
0.889***
(36.98)
Y
10216
0.127**
(2.53)
0.047
(0.98)
A7
System GMM
b/(t)
0.140**
(2.50)
0.026
(0.38)
-0.008
(-0.27)
0.002
(0.51)
0.009
(0.31)
0.001
(0.57)
0.877***
(32.45)
Y
7761
A8
System GMM
b/(t)
0.853***
(37.41)
Y
10216
0.302**
(2.47)
0.044
(0.89)
A9
System GMM
b/(t)
0.310**
(2.30)
-0.003
(-0.05)
-0.047
(-1.64)
0.001
(0.40)
0.030
(1.06)
0.003*
(1.65)
0.847***
(33.23)
Y
7761
A10
System GMM
b/(t)
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Regressions with universalism index score for old-age pensions as dependent variables. Errors are clustered on country. Year dummies, region dummies, country dummies and
constant are omitted from the table. In A9 and A10, democracy is treated as endogenous, with four lags used for instrumentation (see Roodman 2009). The maximum time series extend from the late 1880s to 2004.
Table A.15: Democracies, autocracies, and the universalism of pension systems. Robustness testing when excluding observations with means-tested
programs from the analysis.
Year dummies
N
Lag dep. var.
Resource dep.
Size military
Urbanization
Ln population
Ln GDP p.c.
Democracy (BMR)
Model
xxiii
0.638**
(2.47)
0.204
(0.89)
-0.212**
(-2.05)
0.019*
(1.69)
0.323*
(1.78)
-0.010
(-0.90)
Y
7994
Univ.≥ 5
b/(t)
Y
13221
0.188
(0.71)
0.070
(0.40)
Univ.≥ 6
b/(t)
0.515*
(1.82)
0.296
(1.15)
-0.276***
(-2.66)
-0.001
(-0.09)
0.249
(1.08)
-0.003
(-0.25)
Y
7994
Univ.≥ 6
b/(t)
Y
13221
0.243
(0.88)
-0.031
(-0.17)
Univ.≥ 7
b/(t)
0.729***
(2.73)
0.316
(1.15)
-0.354***
(-3.35)
-0.012
(-1.11)
0.369
(1.57)
0.001
(0.10)
Y
7994
Univ.≥ 7
b/(t)
Y
13221
0.211
(0.73)
-0.086
(-0.46)
Univ.≥ 8
b/(t)
0.722***
(2.66)
0.266
(0.94)
-0.414***
(-3.93)
-0.014
(-1.20)
0.386
(1.63)
0.001
(0.13)
Y
7994
Univ.≥ 8
b/(t)
Y
13221
-0.125
(-0.37)
-0.314
(-1.52)
Univ.≥ 9
b/(t)
0.362
(1.27)
-0.153
(-0.55)
-0.510***
(-4.64)
-0.010
(-0.75)
0.452**
(2.11)
0.008
(0.66)
Y
7994
Univ.≥ 9
b/(t)
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10.
Table A.16: Democracies, autocracies, and the universalism of pension systems. Robustness testing by running logit regressions with dummy dependent
variable based on different cut-offs on universalism index (see top row).
Year dummies
N
Resource dep
Size military
Urbanization
Ln population
Y
13221
0.347
(1.41)
0.295*
(1.78)
Democracy (BMR)
Ln GDP p.c.
Univ.≥ 5
b/(t)
Cut-off dep. var. = 1
Democracy (BMR)
Ln GDP p.c.
Ordinal logit
b/(t)
Ordinal logit
b/(t)
Ordinal Probit
b/(t)
Ordinal Probit
b/(t)
RE Ord. Probit
b/(t)
RE Ord. Probit
b/(t)
0.478**
(2.01)
0.866***
(5.48)
0.359
(1.56)
0.441**
(1.97)
-0.035
(-0.40)
0.025**
(2.30)
0.069
(0.44)
-0.014
(-1.30)
Y
8128
0.328**
(2.42)
0.493***
(5.88)
0.256**
(2.01)
0.259**
(2.24)
-0.013
(-0.29)
0.014**
(2.48)
0.038
(0.45)
-0.007
(-1.30)
Y
8128
0.245***
(5.48)
0.041
(0.86)
0.137***
(2.77)
0.208***
(3.71)
1.187***
(12.43)
0.033***
(12.07)
0.069**
(2.09)
-0.003
(-1.53)
Y
8128
Ln population
Urbanization
Size military
Resource dependence
Year dummies
N
Y
10727
Y
10727
Y
10727
Table A.17: Democracies, autocracies, and the universalism of pension systems. Robustness testing
by running ordinal logit, ordinal probit, and random effects ordinal probit regressions.
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. The maximum time series extend from the late 1880s to 2004. Constants and year dummies are
omitted from the table. We also tried running random effects ordinal logit regressions, but these did not converge.
xxiv
xxv
Y
8002
0.192
(0.88)
0.027
(0.10)
-0.334
(-1.19)
0.056***
(5.29)
0.051
(0.49)
-0.007
(-0.99)
C2
Fixed Eff.
b/(t)
-0.064***
(-9.28)
Y
10434
0.076***
(2.95)
0.049**
(2.21)
C3
ECM
b/(t)
0.067**
(2.31)
0.037
(1.34)
-0.013
(-0.52)
0.004***
(3.34)
0.024*
(1.82)
-0.000
(-0.37)
-0.081***
(-9.02)
Y
7905
C4
ECM
b/(t)
0.932***
(52.71)
Y
10544
0.036
(1.14)
0.065
(1.60)
C5
Sys. GMM
b/(t)
0.051
(1.45)
0.091
(1.25)
-0.009
(-0.42)
0.000
(0.06)
0.051**
(2.29)
0.003
(0.84)
0.909***
(33.78)
Y
7975
C6
Sys.GMM
b/(t)
Y
7699
3.166***
(3.41)
1.455
(1.29)
C7
Fixed Eff
b/(t)
Y
5769
1.097
(1.51)
-0.063
(-0.07)
-3.048*
(-1.80)
0.244***
(5.47)
-0.062
(-0.19)
-0.049**
(-2.14)
C8
Fixed Eff.
b/(t)
-0.041***
(-4.00)
Y
7421
0.217***
(3.21)
0.094
(1.39)
C9
ECM
b/(t)
0.115*
(1.72)
0.045
(0.53)
-0.131
(-1.01)
0.017***
(3.37)
0.026
(0.61)
-0.003
(-1.15)
-0.066***
(-4.23)
Y
5560
C10
ECM
b/(t)
0.895***
(20.38)
Y
7571
0.251
(1.40)
1.138**
(2.25)
C11
Sys. GMM
0.156
(0.88)
1.244**
(2.06)
0.096
(0.71)
0.005
(0.53)
-0.033
(-0.39)
-0.011
(-1.60)
0.883***
(16.15)
Y
5677
C12
Sys. GMM
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Universalism Index for pensions is dependent variable for C1–C6, and Welfare State Universalism Index (additive universalism score over all six policy programs) for C7–C12.
Errors are clustered on country. Maximum time series extend from late-1880s–2004.
Table A.18: Democracies, autocracies, and universalism of pension systems (C1–C6) or of social policy programs more generally (C7–C12). Robustness
testing when lagging all independent variables 1 additional year.
Year dummies
N
Lag dep. var.
Resource dep.
Size military
Urbanization
Ln population
Y
10620
0.580**
(2.17)
0.359
(1.39)
Democracy
Ln GDP p.c.
C1
Fixed Eff
b/(t)
Model
xxvi
5251
0.037
(0.11)
-0.156
(-0.59)
-0.302
(-0.48)
0.030
(0.09)
0.030
(0.10)
0.016
(0.07)
0.266
(1.35)
A1
Random Eff.
b/(t)
508
0.028
(0.10)
-0.215
(-0.92)
-0.260
(-0.58)
-0.105
(-0.38)
-0.272
(-1.25)
-0.128
(-0.66)
0.059
(0.29)
0.353**
(2.04)
0.046***
(4.25)
0.066
(0.51)
-0.001
(-0.29)
A2
Random Eff.
b/(t)
5251
0.116
(0.35)
-0.188
(-0.71)
-0.014
(-0.02)
0.005
(0.02)
-0.020
(-0.07)
-0.063
(-0.28)
-0.025
(-0.10)
A3
Fixed Eff
b/(t)
5084
0.105
(0.36)
-0.188
(-0.83)
-0.047
(-0.10)
-0.080
(-0.28)
-0.091
(-0.39)
-0.097
(-0.50)
0.152
(0.66)
1.340***
(3.03)
0.042***
(3.12)
0.019
(0.15)
-0.001
(-0.29)
A4
Fixed Eff.
b/(t)
-0.136***
(-10.16)
5212
-0.066
(-1.01)
-0.072
(-1.34)
-0.169
(-1.54)
-0.073
(-1.14)
-0.063
(-0.81)
-0.031
(-0.61)
0.034
(0.74)
A5
ECM
b/(t)
-0.049
(-0.76)
-0.062
(-1.23)
-0.156*
(-1.80)
-0.083
(-1.34)
-0.106
(-1.59)
-0.026
(-0.52)
0.065
(1.40)
0.193**
(2.22)
0.008***
(3.26)
0.042
(1.57)
-0.000
(-0.00)
-0.151***
(-11.01)
5046
A6
ECM
b/(t)
0.917***
(41.17)
5212
0.039
(0.50)
0.023
(0.29)
-0.060
(-0.57)
0.078
(0.81)
-0.076
(-0.69)
-0.027
(-0.31)
0.050
(1.06)
A7
System GMM
b/(t)
0.028
(0.35)
0.003
(0.03)
-0.099
(-0.89)
0.079
(0.82)
-0.152
(-1.38)
-0.043
(-0.46)
0.029
(0.36)
-0.017
(-0.71)
0.002
(0.53)
-0.002
(-0.04)
0.002*
(1.88)
0.913***
(40.23)
5046
A8
System GMM
b/(t)
0.872***
(47.63)
5212
-0.237*
(-1.86)
-0.145
(-1.26)
-0.382*
(-1.87)
-0.161
(-1.40)
-0.130
(-0.72)
-0.131
(-1.15)
0.081**
(2.41)
A9
System GMM
b/(t)
-0.241*
(-1.88)
-0.151
(-1.28)
-0.342*
(-1.89)
-0.195*
(-1.68)
-0.254
(-1.48)
-0.135
(-1.17)
0.053
(0.98)
0.011
(0.28)
0.002
(0.91)
0.006
(0.16)
0.002
(1.55)
0.867***
(44.61)
5046
A10
System GMM
b/(t)
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Regressions with universalism index score for old-age pensions as dependent variable. “Pure” dominant-party systems (excluding the hybrid party regimes) is the reference
category. Errors are clustered on country. Year dummies, region dummies, country dummies and constant are omitted from the table. In A9 and A10, the Geddes et al. regime dummies are treated as endogenous, with four lags
used for instrumentation (see Roodman 2009). The maximum time series extend from 1946 to 2004.
Table A.19: Types of autocracies and the universalism of pension systems.
N
Lag dep. var.
Resource dep.
Size military
Urbanization
Ln population
Ln GDP p.c.
Democracy (GWF)
Other/non-regime
Personalist
Monarchy
Military regime
Hybrid party reg.
Model
xxvii
3918
-0.114
(-0.09)
-0.312
(-0.24)
0.900
(0.37)
-0.014
(-0.01)
0.297
(0.18)
0.212
(0.20)
0.767
(0.90)
A1
Random Eff.
b/(t)
3791
-0.086
(-0.07)
-0.251
(-0.20)
0.854
(0.41)
-0.071
(-0.04)
0.163
(0.11)
0.148
(0.15)
0.458
(0.51)
0.477
(0.41)
0.080*
(1.86)
0.284
(0.57)
-0.014
(-1.01)
A2
Random Eff.
b/(t)
3918
-0.086
(-0.06)
-0.428
(-0.32)
1.218
(0.50)
-0.147
(-0.08)
0.056
(0.03)
0.023
(0.02)
0.327
(0.37)
A3
Fixed Eff
b/(t)
3791
-0.142
(-0.11)
-0.416
(-0.32)
1.182
(0.54)
-0.230
(-0.13)
-0.019
(-0.01)
-0.042
(-0.04)
0.053
(0.05)
0.418
(0.22)
0.062
(1.38)
0.246
(0.50)
-0.012
(-0.80)
A4
Fixed Eff.
b/(t)
-0.135***
(-4.75)
3825
-0.209
(-0.75)
-0.250
(-0.94)
-0.222
(-0.54)
-0.202
(-0.55)
-0.085
(-0.25)
-0.101
(-0.42)
0.123
(0.89)
A5
ECM
b/(t)
-0.205
(-0.76)
-0.236
(-0.91)
-0.196
(-0.52)
-0.202
(-0.57)
-0.057
(-0.18)
-0.082
(-0.36)
0.107
(0.66)
0.138
(0.45)
0.010
(1.38)
0.117
(1.58)
-0.002
(-0.57)
-0.143***
(-4.80)
3702
A6
ECM
b/(t)
0.879***
(11.94)
3825
0.202
(0.77)
-0.029
(-0.11)
-1.066
(-1.20)
0.023
(0.07)
-0.248
(-0.85)
0.260
(0.85)
1.422
(1.54)
A7
System GMM
b/(t)
0.128
(0.48)
-0.161
(-0.61)
-1.186
(-1.20)
-0.066
(-0.21)
-0.326
(-1.11)
0.121
(0.40)
1.321
(1.24)
-0.057
(-0.37)
0.014
(1.10)
-0.212
(-1.33)
-0.013
(-1.60)
0.879***
(11.70)
3702
A8
System GMM
b/(t)
0.891***
(23.61)
3825
-0.077
(-0.21)
0.314
(0.93)
-0.817
(-1.13)
0.282
(0.86)
0.099
(0.34)
0.386
(1.07)
0.976***
(2.70)
A9
System GMM
b/(t)
-0.172
(-0.45)
0.225
(0.63)
-0.741
(-1.03)
0.189
(0.55)
-0.047
(-0.15)
0.279
(0.73)
0.741*
(1.85)
-0.142
(-1.11)
0.021*
(1.92)
-0.148
(-1.58)
-0.005
(-0.86)
0.885***
(22.41)
3702
A10
System GMM
b/(t)
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Regressions with Welfare State Universalism Index (additive universalism score, aggregated over all six major policy programs) as dependent variable. “Pure” dominant-party
systems (excluding the hybrid party regimes) is the reference category. Errors are clustered on country. Year dummies, region dummies, country dummies and constant are omitted from the table. In A9 and A10, the Geddes et al.
regime dummies are treated as endogenous, with four lags used for instrumentation (see Roodman 2009). The maximum time series extend from 1946 to 2004.
Table A.20: Types of autocracies and the universalism of social welfare policy programs.
N
Lag dep. var.
Resource dep.
Size military
Urbanization
Ln population
Ln GDP p.c.
Democracy (GWF)
Other/non-regime
Personalist reg.
Monarchy
Military regime
Hybrid party reg.
Model
A.4
Robustness tests and extensions for Section 5.3
Variable\Statistic
Mean
Standard deviation
Minimum
Maximum
Regime failure
Old-age pension system
Regional pension share
Global pension share
Military regime
Monarchy
Personalist regime
Other autocracy
Ln GDP per capita
Ln population
Ethnic fractionalization
Urbanization
Size of military
Resource dependence
0.05
0.71
0.82
0.85
0.15
0.11
0.26
0.01
7.81
9.10
0.51
33.97
0.66
6.42
0.23
0.41
0.41
0.21
0.36
0.31
0.44
0.11
0.93
1.36
0.25
20.88
0.74
12.73
0
0
0
0.53
0
0
0
0
5.78
5.97
0.00
0
0
0
1
1
1
0.98
1
1
1
1
11.08
14.06
0.93
100
7.68
100
Table A.21: Descriptive statistics for the 3248 observations (and variables) entering Model D3, Table
4 in the paper.
xxviii
xxix
3248
98
-0.005
(-0.65)
-0.050
(-0.22)
1.566***
(8.52)
-0.075
(-0.20)
0.895***
(4.81)
1.158*
(1.76)
-0.216
(-1.28)
-0.108*
(-1.78)
-0.295
(-0.85)
0.007
(1.01)
-0.286
(-1.59)
-0.006
(-0.56)
Logit
b/(t)
3248
98
-0.062
(-0.25)
1.687***
(8.42)
-0.140
(-0.37)
0.949***
(4.73)
1.048
(1.43)
-0.243
(-1.37)
-0.123*
(-1.92)
-0.274
(-0.75)
0.008
(1.03)
-0.300
(-1.60)
-0.006
(-0.53)
0.121
(1.44)
-0.007
(-0.91)
Logit
b/(t)
Y
3248
98
0.007
(0.77)
-0.038
(-0.16)
1.374***
(6.44)
-0.096
(-0.20)
0.765***
(3.62)
1.513***
(2.82)
-0.428**
(-2.07)
-0.126
(-1.42)
0.125
(0.29)
0.002
(0.22)
-0.205
(-1.04)
-0.001
(-0.15)
Logit
b/(t)
Y
3248
98
-0.040
(-0.16)
1.524***
(6.62)
-0.178
(-0.36)
0.832***
(3.58)
1.468**
(2.46)
-0.463**
(-2.08)
-0.145
(-1.52)
0.154
(0.33)
0.002
(0.25)
-0.225
(-1.09)
-0.001
(-0.14)
0.130
(1.63)
0.004
(0.45)
Logit
b/(t)
Y
Y
3120
98
0.076
(0.30)
1.444***
(6.51)
-0.210
(-0.45)
0.812***
(3.74)
1.275**
(2.55)
-0.355*
(-1.71)
-0.131
(-1.46)
0.177
(0.40)
0.002
(0.26)
-0.228
(-1.10)
-0.002
(-0.18)
Logit
b/(t)
Y
Y
3120
98
0.080
(0.30)
1.570***
(6.56)
-0.284
(-0.58)
0.868***
(3.68)
1.253**
(2.27)
-0.381*
(-1.75)
-0.147
(-1.53)
0.206
(0.43)
0.002
(0.26)
-0.246
(-1.14)
-0.002
(-0.16)
0.107
(1.37)
Logit
b/(t)
3120
98
0.045
(0.37)
0.709***
(6.53)
-0.093
(-0.45)
0.391***
(3.93)
0.582**
(2.24)
-0.171*
(-1.84)
-0.061
(-1.47)
0.068
(0.34)
0.000
(0.05)
-0.089
(-1.00)
-0.001
(-0.25)
Probit
b/(t)
3120
98
0.048
(0.39)
0.781***
(6.56)
-0.140
(-0.63)
0.427***
(3.86)
0.561**
(1.97)
-0.182*
(-1.89)
-0.070
(-1.57)
0.077
(0.36)
0.000
(0.05)
-0.097
(-1.06)
-0.001
(-0.23)
0.059
(1.51)
Probit
b/(t)
Y
3248
98
-0.001
(-0.43)
-0.026
(-0.24)
0.751***
(8.54)
-0.032
(-0.20)
0.408***
(4.83)
0.520*
(1.69)
-0.099
(-1.33)
-0.048*
(-1.66)
-0.147
(-0.91)
0.003
(0.85)
-0.122
(-1.59)
-0.002
(-0.45)
Probit
b/(t)
Y
3248
98
-0.034
(-0.28)
0.819***
(8.42)
-0.069
(-0.43)
0.443***
(4.75)
0.472
(1.39)
-0.113
(-1.44)
-0.056*
(-1.83)
-0.139
(-0.82)
0.003
(0.86)
-0.128
(-1.61)
-0.002
(-0.43)
0.065*
(1.66)
-0.003
(-0.76)
Probit
b/(t)
0.004
(1.03)
Y
Y
3248
98
-0.013
(-0.12)
0.662***
(6.61)
-0.034
(-0.16)
0.356***
(3.79)
0.695***
(2.67)
-0.191**
(-2.09)
-0.059
(-1.42)
0.054
(0.28)
0.000
(0.01)
-0.084
(-0.97)
-0.000
(-0.03)
Probit
b/(t)
-0.012
(-0.10)
0.744***
(6.70)
-0.086
(-0.40)
0.399***
(3.75)
0.664**
(2.30)
-0.206**
(-2.13)
-0.070
(-1.56)
0.064
(0.30)
0.000
(0.04)
-0.095
(-1.05)
-0.000
(-0.05)
0.069*
(1.76)
0.003
(0.61)
Y
Y
3248
98
Probit
b/(t)
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. Errors are clustered on country. The maximum time series extend from the 1946 to 2004.
Table A.22: “Naive” Logit and Probit regressions, assuming existence of pension programs is exogenously given, with autocratic regime breakdown as
dependent variable.
Year dummies
Region dummies
N
Countries
Time trend
Ln regime duration
Resource dependence
Size military
Urbanization
Ethnic fraction.
Ln population
Ln GDP p.c.
Other autocracy
Personalist regime
Monarchy
Military regime
Pension system
Model
xxx
3248
98
R
147.3
16.4
148.6
16.4
-0.002
(-0.77)
3248
98
R
-0.002
(-0.70)
-0.110
(-1.41)
-0.200**
(-2.54)
-0.010
(-0.20)
0.024
(0.38)
-0.015
(-0.40)
0.028
(1.48)
0.146
(1.33)
0.005***
(2.73)
-0.049
(-1.62)
0.001
(1.18)
0.015
(0.79)
0.794***
(5.24)
D2
IVProbit
b/(t)
3248
98
R+G
0.30
81.0
19.9
-0.000
(-0.10)
0.821***
(5.23)
-0.203
(-0.57)
-0.124*
(-1.80)
-0.190**
(-2.48)
-0.020
(-0.41)
0.040
(0.63)
-0.012
(-0.34)
0.030
(1.59)
0.149
(1.36)
0.005***
(2.74)
-0.049*
(-1.65)
0.001
(1.19)
D3
IVProbit
b/(t)
3248
98
R+G
0.27
80.1
19.9
-0.000
(-0.17)
0.807***
(5.08)
-0.208
(-0.58)
-0.107
(-1.37)
-0.201**
(-2.55)
-0.012
(-0.23)
0.024
(0.39)
-0.015
(-0.40)
0.028
(1.47)
0.149
(1.37)
0.005***
(2.75)
-0.050*
(-1.66)
0.001
(1.25)
0.016
(0.81)
D4
IVProbit
b/(t)
3247
98
R+G
0.48
81.3
19.9
-0.055
(-0.19)
0.031
(0.23)
-0.000
(-0.08)
0.821***
(5.24)
-0.210
(-0.58)
-0.124*
(-1.80)
-0.190**
(-2.47)
-0.020
(-0.41)
0.040
(0.63)
-0.012
(-0.34)
0.030
(1.59)
0.149
(1.36)
0.005***
(2.73)
-0.049
(-1.63)
0.001
(1.20)
D5
IVprobit
b/(t)
3247
98
R+G
0.45
80.5
19.9
0.807***
(5.09)
-0.217
(-0.60)
-0.108
(-1.37)
-0.201**
(-2.55)
-0.012
(-0.23)
0.024
(0.38)
-0.015
(-0.40)
0.028
(1.47)
0.149
(1.37)
0.005***
(2.74)
-0.050
(-1.64)
0.001
(1.26)
0.016
(0.81)
-0.075
(-0.26)
0.032
(0.24)
-0.000
(-0.14)
D6
IVprobit
b/(t)
3247
98
R+G
0.48
81.3
19.9
Y
-0.034
(-0.13)
0.007
(0.07)
0.001
(0.27)
0.274**
(2.14)
0.274
(0.84)
-0.105*
(-1.74)
-0.147**
(-1.99)
-0.030
(-0.58)
-0.090
(-1.46)
0.019
(0.58)
0.074***
(4.14)
0.053
(0.47)
0.004**
(2.27)
0.002
(0.08)
0.001
(0.53)
D7
IVProbit
b/(t)
3247
98
R+G
0.45
80.5
19.9
Y
0.267**
(2.13)
0.255
(0.78)
-0.088
(-1.28)
-0.158**
(-2.12)
-0.022
(-0.40)
-0.102
(-1.62)
0.016
(0.48)
0.072***
(4.00)
0.054
(0.49)
0.004**
(2.26)
0.000
(0.01)
0.001
(0.56)
0.013
(0.72)
-0.055
(-0.21)
0.011
(0.10)
0.001
(0.22)
D8
IVProbit
b/(t)
53.3
16.4
3120
98
R
Y
Y
-0.115*
(-1.95)
-0.142*
(-1.87)
-0.028
(-0.52)
-0.090
(-1.42)
0.013
(0.38)
0.075***
(4.10)
0.055
(0.49)
0.004**
(2.35)
0.004
(0.15)
0.000
(0.25)
0.128
(0.81)
D9
IVprobit
b/(t)
52.6
16.4
3120
98
R
Y
Y
-0.101
(-1.48)
-0.152**
(-2.00)
-0.021
(-0.37)
-0.100
(-1.54)
0.011
(0.30)
0.073***
(3.98)
0.055
(0.49)
0.004**
(2.34)
0.003
(0.11)
0.000
(0.28)
0.012
(0.63)
0.119
(0.76)
D10
IVprobit
b/(t)
Y
3283
97
R+G
0.42
82.5
19.9
0.043
(0.22)
-0.030
(-0.42)
-0.003**
(-2.53)
0.004***
(4.86)
-0.017*
(-1.67)
0.002***
(3.69)
0.478***
(9.69)
0.398***
(3.52)
-0.108
(-0.51)
0.161***
(5.21)
0.002
(0.10)
-0.140
(-1.40)
-0.044***
(-2.81)
0.195***
(-1.81)
D11
FE 2SLS
b/(t)
Y
3283
97
R+G
0.37
81.7
19.9
0.004***
(4.89)
-0.018*
(-1.74)
0.002***
(3.55)
0.013**
(2.18)
0.034
(0.17)
-0.028
(-0.39)
-0.004***
(-2.86)
0.477***
(9.67)
0.391***
(3.46)
0.001
(0.03)
0.140***
(4.34)
0.007
(0.35)
-0.138
(-1.38)
-0.047***
(-3.03)
0.198***
(-1.64)
D12
FE 2SLS
b/(t)
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. For the instruments, R= share of the other countries in the region having pension prgroam in given year; G= share of other countries globally having pension program in given
year. As recommended in the literature, all diagnostic test statistics for IVProbit models (Cragg-Donal Wald F-Statistic for weak identification test and related critical values for 10% of maxima IV size, and Sargan p-value for
overidentification tests of the exclusion restriction) are run on structurally equivalent (fixed effects) 2SLS models. Errors are clustered on country in the IVProbit models. The maximum time series extend from the 1946 to 2004.
Table A.23: First-stage IV regressions with existence of old-age pension programs as dependent variable (second-stage regressions have autocratic regime
breakdown as dependent, and are reported in the paper).
Year dummies
Region dummies
Country dummies
N
Countries
Instruments
Sargan p-value
CD–Wald F-stat
Weak Id. c.v. (10%)
Time trend
Regime failures region
Regime failures globally
Ln regime duration
Resource dependence
Size of military
Urbanization
Ethnic fractionaliz.
Ln population
Ln GDP p.c.
Other autocracy
Personalist
Monarchy
Military regime
-0.127*
(-1.83)
-0.190**
(-2.47)
-0.019
(-0.37)
0.040
(0.63)
-0.012
(-0.33)
0.030
(1.60)
0.146
(1.32)
0.005***
(2.71)
-0.048
(-1.61)
0.001
(1.12)
0.809***
(5.42)
Pension share region
Pension share globally
D1
IVProbit
b/(t)
Model
xxxi
3247
R
-0.579
(-1.63)
0.665***
(5.26)
-0.333
(-1.57)
0.451***
(4.48)
0.421
(1.12)
-0.061
(-0.94)
0.001
(0.27)
-0.047
(-1.34)
-0.032
(-0.18)
0.004
(0.94)
-0.063
(-0.94)
-0.003
(-0.80)
0.062
(1.42)
D2
IVProbit
b/(t)
3247
R+G
-0.481
(-1.52)
0.616***
(5.34)
-0.265
(-1.35)
0.422***
(4.55)
0.478
(1.38)
-0.050
(-0.83)
0.002
(0.49)
-0.042
(-1.24)
-0.045
(-0.27)
0.003
(0.84)
-0.056
(-0.87)
-0.003
(-0.88)
D3
IVProbit
b/(t)
3247
R+G
-0.549
(-1.56)
0.669***
(5.38)
-0.322
(-1.54)
0.451***
(4.49)
0.425
(1.13)
-0.062
(-0.97)
0.001
(0.23)
-0.048
(-1.38)
-0.034
(-0.20)
0.003
(0.91)
-0.063
(-0.94)
-0.003
(-0.81)
0.061
(1.40)
D4
IVProbit
b/(t)
3246
R+G
-1.605
(-0.88)
1.750***
(2.79)
-0.396
(-1.20)
0.632***
(5.67)
-0.213
(-1.07)
0.425***
(4.60)
0.484
(1.41)
-0.058
(-0.95)
0.002
(0.66)
-0.037
(-1.10)
-0.057
(-0.36)
0.002
(0.60)
-0.024
(-0.38)
-0.003
(-0.89)
D5
IVProbit
b/(t)
3246
R+G
-0.459
(-1.27)
0.683***
(5.71)
-0.266
(-1.26)
0.454***
(4.51)
0.437
(1.17)
-0.069
(-1.07)
0.002
(0.41)
-0.043
(-1.24)
-0.046
(-0.28)
0.003
(0.69)
-0.029
(-0.45)
-0.003
(-0.83)
0.058
(1.33)
-1.706
(-0.95)
1.739***
(2.77)
D6
IVProbit
b/(t)
3246
R+G
Y
-1.145
(-0.84)
1.026
(1.58)
-2.215***
(-2.98)
0.202
(0.77)
-0.382
(-1.38)
0.185
(1.01)
0.212
(0.65)
-0.056
(-0.55)
0.017***
(3.58)
0.131
(1.50)
0.249
(0.94)
0.006
(0.87)
0.051
(0.69)
0.001
(0.28)
D7
IVProbit
b/(t)
3246
R+G
Y
-2.298***
(-3.09)
0.254
(0.90)
-0.444
(-1.56)
0.212
(1.07)
0.149
(0.42)
-0.065
(-0.60)
0.016***
(3.25)
0.127
(1.44)
0.256
(0.93)
0.006
(0.92)
0.043
(0.59)
0.001
(0.30)
0.064
(1.25)
-1.216
(-0.92)
0.992
(1.54)
D8
IVProbit
b/(t)
3172
R
Y
Y
0.221***
(3.20)
0.229
(0.66)
0.012*
(1.91)
0.030
(0.33)
0.001
(0.13)
-3.180***
(-15.22)
-0.231
(-0.78)
-0.480**
(-1.97)
-0.019
(-0.08)
-0.110
(-0.32)
0.012
(0.10)
D9
IVProbit
b/(t)
3172
R
Y
Y
0.217***
(3.15)
0.227
(0.65)
0.013*
(1.94)
0.025
(0.28)
0.001
(0.17)
0.046
(0.73)
-3.199***
(-17.17)
-0.194
(-0.61)
-0.516**
(-2.12)
-0.002
(-0.01)
-0.166
(-0.48)
0.008
(0.07)
D10
IVProbit
b/(t)
Y
3282
R+G
-0.069
(-0.33)
0.148**
(2.02)
0.001*
(1.72)
-0.006
(-0.57)
0.000
(0.76)
-0.150*
(-1.87)
0.047**
(2.17)
0.066**
(2.05)
0.014
(0.74)
0.161
(1.55)
-0.037**
(-2.34)
0.004***
(3.69)
-0.087*
(-1.81)
D11
FE2SLS
b/(t)
Y
3282
R+G
0.002*
(1.79)
-0.007
(-0.70)
0.000
(0.57)
0.022***
(3.57)
-0.085
(-0.41)
0.151**
(2.06)
-0.159**
(-1.97)
0.066***
(2.98)
0.031
(0.95)
0.023
(1.17)
0.164
(1.58)
-0.043***
(-2.76)
0.003***
(2.80)
-0.079
(-1.64)
D12
FE2SLS
b/(t)
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. R=share other countries in region having pension program that year; G=share other countries globally having pension program that year. Maximum time series are 1946–2003 (for
independent variables).
Table A.24: Second-stage IV regressions with autocratic regime failure as dependent variable and existence of old-age pension programs as endogenous
independent. Robustness testing when lagging all independent variables 1 additional year.
Year dummies
Region dummies
Country dummies
N
Instruments
Regime failures region
Regime failures globally
Ln regime duration
Resource dependence
Size military
Urbanization
Ethnic fraction.
Ln population
Time trend
Ln GDP p.c.
Other autocracy
Personalist regime
Monarchy
3247
R
-0.510
(-1.58)
0.611***
(5.21)
-0.275
(-1.38)
0.421***
(4.54)
0.475
(1.37)
-0.049
(-0.80)
0.002
(0.53)
-0.041
(-1.20)
-0.043
(-0.26)
0.003
(0.87)
-0.056
(-0.88)
-0.003
(-0.87)
Pension system
Military regime
D1
IVProbit
b/(t)
Model
xxxii
3247
98
R
147.3
16.4
148.6
16.4
-0.002
(-0.77)
3247
98
R
-0.002
(-0.70)
-0.111
(-1.41)
-0.200**
(-2.54)
-0.011
(-0.20)
0.024
(0.38)
-0.015
(-0.40)
0.028
(1.48)
0.147
(1.33)
0.005***
(2.73)
-0.049
(-1.62)
0.001
(1.18)
0.015
(0.79)
0.794***
(5.24)
E2
IVProbit
b/(t)
3247
98
R+G
0.08
80.9
19.9
-0.000
(-0.18)
0.820***
(5.20)
-0.174
(-0.47)
-0.125*
(-1.81)
-0.190**
(-2.48)
-0.020
(-0.40)
0.040
(0.63)
-0.012
(-0.34)
0.030
(1.60)
0.149
(1.35)
0.005***
(2.74)
-0.049
(-1.64)
0.001
(1.17)
E3
IVProbit
b/(t)
3247
98
R+G
0.07
80.1
19.9
-0.001
(-0.26)
0.805***
(5.04)
-0.174
(-0.47)
-0.108
(-1.38)
-0.201**
(-2.55)
-0.012
(-0.23)
0.024
(0.38)
-0.015
(-0.40)
0.028
(1.47)
0.149
(1.36)
0.005***
(2.75)
-0.050*
(-1.65)
0.001
(1.23)
0.016
(0.81)
E4
IVProbit
b/(t)
3246
98
R+G
0.11
81.3
19.9
-0.052
(-0.18)
0.037
(0.28)
-0.000
(-0.17)
0.820***
(5.21)
-0.178
(-0.48)
-0.125*
(-1.81)
-0.190**
(-2.47)
-0.020
(-0.40)
0.040
(0.62)
-0.012
(-0.33)
0.030
(1.60)
0.149
(1.35)
0.005***
(2.74)
-0.049
(-1.61)
0.001
(1.18)
E5
IVprobit
b/(t)
3246
98
R+G
0.11
80.5
19.9
0.806***
(5.05)
-0.180
(-0.48)
-0.108
(-1.38)
-0.200**
(-2.54)
-0.012
(-0.23)
0.024
(0.37)
-0.015
(-0.40)
0.028
(1.47)
0.149
(1.36)
0.005***
(2.75)
-0.049
(-1.63)
0.001
(1.23)
0.016
(0.81)
-0.071
(-0.24)
0.039
(0.29)
-0.001
(-0.24)
E6
IVprobit
b/(t)
3246
98
R+G
0.11
81.3
19.9
Y
-0.020
(-0.08)
0.016
(0.14)
0.000
(0.10)
0.246*
(1.85)
0.381
(1.42)
-0.106*
(-1.76)
-0.146**
(-1.98)
-0.030
(-0.58)
-0.090
(-1.45)
0.019
(0.56)
0.074***
(4.15)
0.053
(0.48)
0.004**
(2.27)
0.003
(0.13)
0.001
(0.51)
E7
IVProbit
b/(t)
3246
98
R+G
0.11
80.5
19.9
Y
0.237*
(1.81)
0.371
(1.38)
-0.089
(-1.29)
-0.157**
(-2.11)
-0.022
(-0.40)
-0.102
(-1.61)
0.016
(0.46)
0.072***
(4.00)
0.055
(0.49)
0.004**
(2.27)
0.002
(0.06)
0.001
(0.53)
0.014
(0.73)
-0.039
(-0.15)
0.020
(0.18)
0.000
(0.04)
E8
IVProbit
b/(t)
53.4
16.4
2091
98
R
Y
Y
-0.077
(-1.14)
-0.134*
(-1.75)
-0.029
(-0.54)
-0.089
(-1.37)
0.007
(0.22)
0.069***
(3.85)
0.126
(1.09)
0.005***
(2.77)
0.003
(0.15)
-0.000
(-0.44)
0.149
(0.89)
E9
IVprobit
b/(t)
52.7
16.4
2091
98
R
Y
Y
-0.030
-0.058
(-0.74)
-0.147**
(-1.97)
-0.019
(-0.32)
-0.101
(-1.49)
0.005
(0.14)
0.066***
(3.71)
0.128
(1.11)
0.005***
(2.75)
0.003
(0.11)
-0.000
(-0.37)
0.015
(0.75)
0.134
(0.81)
E10
IVprobit
b/(t)
Y
3282
97
R+G
0.08
82.5
19.9
0.043
(0.22)
-0.028
(-0.42)
-0.003**
(-2.53)
0.004***
(4.86)
-0.017*
(-1.67)
0.002***
(3.69)
0.478***
(9.69)
0.398***
(3.52)
-0.108
(-0.51)
0.161***
(5.21)
0.002
(0.10)
-0.140
(-1.40)
-0.044***
(-2.81)
0.195***
(4.44)
E11
FE 2SLS
b/(t)
Y
3282
97
R+G
0.08
81.7
19.9
(-0.39)
-0.004***
(-2.86)
0.004***
(4.89)
-0.018*
(-1.74)
0.002***
(3.55)
0.013**
(2.18)
0.034
(0.17)
0.477***
(9.67)
0.391***
(3.46)
0.001
(0.03)
0.140***
(4.34)
0.007
(0.35)
-0.138
(-1.38)
-0.047***
(-3.03)
0.198***
(4.51)
E12
FE 2SLS
b/(t)
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. For the instruments, R= share of the other countries in the region having pension program in given year; G= share of other countries globally having pension program in given
year. As recommended in the literature, all diagnostic test statistics for IVProbit models (Cragg-Donal Wald F-Statistic for weak identification test and related critical values for 10% of maxima IV size, and Sargan p-value for
overidentification tests of the exclusion restriction) are run on structurally equivalent (fixed effects) 2SLS models. Errors are clustered on country in the IVProbit models. The maximum time series extend from the 1946 to 2004.
Table A.25: First-stage IV regressions with existence of old-age pension programs as dependent variable (second-stage regressions have democratization
as dependent, and are reported below).
Year dummies
Region dummies
Country dummies
N
Countries
Instruments
Sargan p-value
CD–Wald F-stat
Weak Id. c.v. (10%)
Time trend
Regional regime failures
Global regime failures
Ln regime duration
Resource dependence
Size military
Urbanization
Ethnic fractionaliz.
Ln population
Ln GDP p.c.
Other autocracy
Personalist regime
Monarchy
Military regime
-0.127*
(-1.84)
-0.190**
(-2.47)
-0.019
(-0.37)
0.039
(0.63)
-0.012
(-0.33)
0.030
(1.60)
0.146
(1.32)
0.005***
(2.72)
-0.048
(-1.61)
0.001
(1.12)
0.809***
(5.42)
Pension share region
Pension share globally
E1
IVProbit
b/(t)
Model
xxxiii
3247
98
R
147.3
16.4
3247
98
R
148.6
16.4
0.009*
(1.71)
-1.193**
(-2.51)
0.801***
(4.66)
-0.762**
(-2.42)
0.309**
(2.04)
0.596*
(1.84)
0.005
(0.04)
0.016
(0.41)
-0.420*
(-1.84)
0.011**
(2.12)
-0.186
(-1.28)
-0.000
(-0.05)
0.120*
(1.68)
E2
IVProbit
b/(t)
3247
98
R+G
0.08
80.9
19.9
0.011**
(2.10)
-1.008**
(-2.05)
0.710***
(5.00)
-0.643**
(-2.03)
0.231
(1.60)
0.725**
(2.54)
0.020
(0.20)
0.027
(0.72)
-0.441**
(-2.04)
0.010*
(1.96)
-0.165
(-1.15)
-0.001
(-0.13)
E3
IVProbit
b/(t)
3247
98
R+G
0.07
80.1
19.9
0.009*
(1.69)
-1.122**
(-2.10)
0.816***
(4.65)
-0.738**
(-2.25)
0.308**
(2.04)
0.610*
(1.87)
0.003
(0.03)
0.014
(0.38)
-0.429*
(-1.91)
0.010**
(1.99)
-0.185
(-1.26)
-0.001
(-0.08)
0.118
(1.64)
E4
IVProbit
b/(t)
3246
98
R+G
0.11
81.3
19.9
3.559
(1.31)
1.035
(1.21)
0.012**
(2.27)
-0.989**
(-1.99)
0.726***
(5.11)
-0.623**
(-2.02)
0.255*
(1.76)
0.725**
(2.48)
0.018
(0.18)
0.036
(0.95)
-0.454**
(-2.14)
0.010*
(1.90)
-0.140
(-0.96)
-0.001
(-0.21)
E5
IVprobit
b/(t)
3246
98
R+G
0.11
80.5
19.9
-1.099**
(-2.05)
0.824***
(4.74)
-0.714**
(-2.24)
0.327**
(2.15)
0.618*
(1.87)
0.003
(0.03)
0.025
(0.64)
-0.440**
(-2.01)
0.010*
(1.95)
-0.160
(-1.07)
-0.001
(-0.16)
0.111
(1.58)
3.439
(1.27)
0.978
(1.16)
0.010*
(1.89)
E6
IVprobit
b/(t)
3246
98
R+G
0.11
81.3
19.9
Y
0.970
(0.64)
0.425
(0.75)
0.021***
(3.72)
-2.984***
(-14.23)
0.026
(0.11)
-0.586**
(-2.22)
-0.008
(-0.05)
0.044
(0.17)
0.041
(0.37)
0.236***
(4.27)
-0.013
(-0.04)
0.013**
(2.28)
0.022
(0.25)
0.002
(0.44)
E7
IVProbit
b/(t)
3246
98
R+G
0.11
80.5
19.9
Y
-3.002***
(-14.60)
0.106
(0.39)
-0.644**
(-2.42)
0.039
(0.20)
-0.018
(-0.07)
0.025
(0.22)
0.224***
(4.00)
0.004
(0.01)
0.013**
(2.28)
0.010
(0.11)
0.002
(0.45)
0.078
(1.24)
0.857
(0.58)
0.391
(0.71)
0.019***
(3.28)
E8
IVProbit
b/(t)
53.4
16.4
2091
98
R
Y
Y
-3.200***
(-14.68)
-0.003
(-0.01)
-0.521**
(-2.12)
-0.037
(-0.19)
-0.083
(-0.24)
0.015
(0.13)
0.230***
(4.13)
0.309
(0.78)
0.016***
(2.90)
0.009
(0.12)
-0.001
(-0.27)
E9
IVprobit
b/(t)
52.7
16.4
2091
98
R
Y
Y
-3.229***
(-16.42)
0.053
(0.13)
-0.573**
(-2.33)
0.005
(0.02)
-0.161
(-0.47)
0.003
(0.03)
0.219***
(3.89)
0.335
(0.86)
0.016***
(2.88)
0.003
(0.03)
-0.001
(-0.22)
0.073
(0.97)
E10
IVprobit
b/(t)
Y
3282
97
R+G
0.08
82.5
19.9
0.162
(1.25)
0.052
(1.13)
0.002***
(3.42)
0.001**
(2.00)
-0.013**
(-1.97)
0.001**
(2.13)
-0.200***
(-3.98)
0.058***
(4.31)
0.042**
(2.10)
-0.000
(-0.04)
0.138**
(2.11)
-0.011
(-1.13)
0.008
(0.28)
E11
FE 2SLS
b/(t)
Y
3282
97
R+G
0.08
81.7
19.9
0.001**
(2.02)
-0.013**
(-2.02)
0.001**
(2.06)
0.006
(1.43)
0.158
(1.22)
0.053
(1.15)
0.002***
(3.01)
-0.202***
(-3.99)
0.063***
(4.53)
0.033
(1.61)
0.002
(0.14)
0.139**
(2.12)
-0.013
(-1.29)
0.010
(0.34)
E12
FE 2SLS
b/(t)
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. For the instruments, R= share of the other countries in the region having pension prgroam in given year; G= share of other countries globally having pension program in given
year. As recommended in the literature, all diagnostic test statistics for IVProbit models (Cragg-Donal Wald F-Statistic for weak identification test and related critical values for 10% of maxima IV size, and Sargan p-value for
overidentification tests of the exclusion restriction) are run on structurally equivalent (fixed effects) 2SLS models. The first-stage regressions for these models are reported in Table A.25. Errors are clustered on country in the
IVProbit models. The maximum time series extend from the 1946 to 2004.
Table A.26: Second-stage IV regressions with democratization as dependent variable and existence of old-age pension programs as endogenous independent variable.
Year dummies
Region dummies
Country dummies
N
Countries
Instruments
Sargan p-value
CD–Wald F-stat
Weak Id. c.v. (10%)
Time trend
Region regime failures
Global regime failures
Log regime duration
Resource dependence
Size military
Urbanization
Ethnic fractionaliz.
Ln population
Ln GDP p.c.
Other autocracies
Personalist regime
Monarchy
0.011**
(2.15)
-1.077**
(-2.49)
0.695***
(5.02)
-0.664**
(-2.19)
0.230
(1.59)
0.715**
(2.53)
0.022
(0.22)
0.028
(0.74)
-0.433**
(-1.97)
0.010**
(2.08)
-0.165
(-1.16)
-0.001
(-0.10)
Pension program
Military regime
E1
IVProbit
b/(t)
Model
Second-stage model
Pension program
Ln GDP p.c.
Ln population
Ethnic fractionaliz.
Urbanization
Size military
Resource dependence
Time trend
Region dummies
Year dummies
First-stage model
Regional share pensions
1
b/(t)
2
b/(t)
3
b/(t)
4
b/(t)
5
b/(t)
0.270
(0.80)
0.133
(1.64)
0.061*
(1.65)
-0.189
(-0.88)
0.002
(0.48)
-0.125
(-1.22)
-0.016
(-1.46)
0.002
(0.86)
0.288
(0.85)
0.133
(1.64)
0.060
(1.64)
-0.188
(-0.88)
0.002
(0.46)
-0.125
(-1.22)
-0.016
(-1.46)
0.002
(0.82)
-0.971
(-1.61)
-0.010
(-0.12)
0.104**
(2.01)
-0.020
(-0.08)
0.007
(1.47)
-0.063
(-0.52)
-0.006
(-0.67)
0.019***
(4.04)
Y
-1.038*
(-1.70)
-0.009
(-0.11)
0.107**
(2.08)
-0.022
(-0.08)
0.007
(1.54)
-0.060
(-0.50)
-0.006
(-0.66)
0.020***
(4.11)
Y
-2.127**
(-2.30)
0.033
(0.39)
0.157***
(2.67)
0.037
(0.13)
0.011*
(1.84)
-0.080
(-0.72)
-0.007
(-0.87)
0.579***
(6.14)
0.507***
(4.20)
0.323
(1.44)
-0.001
(-0.06)
0.043***
(3.64)
-0.021
(-0.24)
0.004***
(3.64)
-0.005
(-0.22)
0.001
(0.57)
-0.000
(-0.08)
Y
0.255*
(1.93)
0.815***
(8.43)
Global share pensions
Ln GDP p.c.
Ln population
Ethnic fractionaliz.
Urbanization
Size military
Resource dependence
Time trend
Region dummies
Year dummies
N
Countries
Instruments
Sargan p-value
CD–Wald F-stat
Weak Id. c.v. (10%)
-0.004
(-0.15)
0.017*
(1.71)
0.062
(0.79)
0.004***
(2.91)
-0.024
(-0.99)
0.001
(0.65)
0.000
(0.14)
5535
116
R
894.9
16.4
0.826***
(6.80)
-0.078
(-0.31)
-0.004
(-0.17)
0.017*
(1.72)
0.065
(0.81)
0.004***
(2.92)
-0.024
(-0.99)
0.001
(0.67)
0.001
(0.54)
5535
116
R+G
0.08
472.1
19.9
0.001
(0.05)
0.044***
(3.67)
-0.022
(-0.25)
0.004***
(3.61)
-0.004
(-0.18)
0.001
(0.65)
0.002**
(2.33)
Y
5535
116
R
894.9
16.4
5535
116
R+G
0.08
472.1
19.9
Y
Y
0.014
(0.65)
0.049***
(3.67)
0.004
(0.04)
0.004***
(2.99)
-0.003
(-0.12)
0.000
(0.15)
Y
Y
3061
116
R
320.0
16.4
Table A.27: IV probit regressions with existence of old-age pension programs as dependent variable
in first-stage and democratization as dependent variable in second-stage regressions. Robustness tests
run on long-time series sample using democratization variable based on BMR
Notes: *** p < 0.01, ** p < 0.05, * p < 0.10. The maximum time series extend from the late 1880s to 2004. Constants and year dummies are
omitted from the table. The instrumental variables are the shares of other countries in the region (R)/globally (G) that have in place an old-age pension
system. The model including region and year dummies and both instruments did not converge.
xxxiv
References
Boix, Carles, Michael Miller and Sebastian Rosato. 2013. “A Complete Data Set of Political Regimes,
1800-2007.” Comparative Political Studies 46(12): 1523–1554.
Cheibub, José Antonio, Jennifer Gandhi and James Vreeland. 2010. “Democracy and dictatorship
revisited.” Public Choice 143:67–101.
Collier, David and Richard E. Messick. 1975. “Prerequisites versus diffusion: Testing alternative
explanations of social security adoption.” American Political Science Review 69(04):1299â1315.
Cutright, Philips. 1965. “Political Structure, Economic Development, and National Social Security
Programs.” American Journal of Sociology 70(5):537–550.
De Mesa, Alberto A. and Carmelo Mesa-Lago. 2006. “The structural pension reform in Chile: Effects,
comparisons with other Latin American reforms, and lessons.” Oxford Review of Economic Policy
22(1):149â167.
Flora, Peter, Jens Alber and others. 1983. State, economy, and society in Western Europe 1815-1975:
a data handbook in two volumes. Vol. 1 Campus Frankfurt aM/London/Chicago.
Flora, Peter, Jens Alber and others. 1983. State, economy, and society in Western Europe 1815-1975:
a data handbook in two volumes. Vol. 2 Campus Frankfurt aM/London/Chicago.
Geddes, Barbara, Joseph Wright and Erica Frantz. 2014. “Autocratic Breakdown and Regime Transitions: A New Data Set.” Perspectives on Politics 12(2):313–331.
Haber, Stephen and Victor Menaldo. 2011. “Do natural resources fuel authoritarianism? A reappraisal
of the resource curse.” American Political Science Review 105(1):1–26.
Hicks, A., Misra J. & Ng T. N. 1995. “The programmatic emergence of the social security state.”
American Sociological Review p. 329â349.
Mares, Isabela. 2005. “Social Protection Around the World: External Insecurity, State Capacity, and
Domestic Political Cleavages.” Comparative Political Studies 38(6):623–651.
Mares, Isabela and Matthew E. Carnes. 2009. “Social Policy in Developing Countries.” Annual Review of Political Science 12:93–113.
Mesa-Lago, Carmelo. 1978. Social Security in Latin America: pressure groups, stratification, and
inequality. Pittsburgh: University of Pittsburgh Press.
Mesa-Lago, Carmelo. 2007. “Social security in Latin America: pension and health care reforms in
the last quarter century.” Latin American Research Review 42(2):181–201.
xxxv
Mesa-Lago, Carmelo. 2008. Reassembling Social Security: A Survey of Pensions and Health Care
Reforms in Latin America: A Survey of Pensions and Health Care Reforms in Latin America.
Oxford University Press.
Miller, Michael K. 2013. “Democratic Pieces: Autocratic Elections and Democratic Development
since 1815.” British Journal of Political Science FirstView:1–30.
xxxvi