An econometrical analysis of the interdependencies between the demographic transition and democracy

Table of contents

ABSTRACT

LIST OF FIGURES

LIST OF TABLES

LIST OF ABBREVIATIONS

1. INTRODUCTION

2. MOTIVATING THEORY
2.1. THE DEMOGRAPHIC TRANSITION
2.2. DEMOCRACY
2.3. INTERDEPENDENCIES BETWEEN THE DEMOGRAPHIC TRANSITION AND DEMOCRACY
2.3.1. EDUCATION
2.3.2. URBANIZATION
2.3.3. WOMEN’S EMPOWERMENT
2.3.4. ECONOMIC GROWTH

3. DATA AND DESCRIPTIVE STATISTICS
3.1. DATA
3.1.1. DEMOGRAPHIC DATA
3.1.2. DATA ON DEMOCRACY
3.1.3. OTHER DATA
3.2. DESCRIPTIVE STATISTICS

4. EMPIRICAL ANALYSIS
4.1. OLS MODEL
4.1.1. METHODOLOGY
4.1.2. ANALYSIS
4.2. COUNTRY FIXED EFFECTS
4.2.1. METHODOLOGY
4.2.2. ANALYSIS
4.3. ROBUSTNESS CHECK
4.3.1. YEAR FIXED EFFECTS
4.3.2. INTERACTIONS WITH INCOME
4.3.3. TIME OF DEMOGRAPHIC TRANSITION
4.4. HOW THE DEMOGRAPHIC TRANSITION AFFECTS DEMOCRACY
4.4.1. LIFE EXPECTANCY AND BIRTH RATES
4.4.2. CHANNEL ANALYSIS

5. LIMITATIONS

6. CONCLUSION

APPENDICES
A. DEFINITIONS
B. DESCRIPTION OF DATA
1. THE DEMOGRAPHIC TRANSITION
C. DATA SUMMARIES
D. REGRESSION RESULTS
1. OLS
2. COUNTRY FIXED EFFECTS
3. COUNTRY AND TIME FIXED EFFECTS
4. INTERACTIONS WITH INCOME
5. TIME OF DEMOGRAPHIC TRANSITION
6. EFFECT OF LIFE EXPECTANCIES AND CRUDE BIRTH RATES
7. CHANNEL ANALYSIS

BIBLIOGRAPHY

Abstract

This paper investigates the effect of an exogenous demographic transition on democracy. As possible channels through which this influence makes an impact, education and per capita income and, to a lesser degree, urbanization were identified. These interdependencies were tested using pooled ordinary least squares as well as fixed effects models on the basis of panel data. In conclusion, the demographic transition affects democracy through the aforementioned channels primarily in poor countries that have experienced their demographic transition and subsequent democratization in the second half of the 20th century or later.

List of figures

Figure 1 The demographic transition

Figure 2 Life expectancies and birth rates

Figure 3 Consequences of the demographic transition

Figure 4 Determinants of democracy

Figure 5 Development of democracy and demography

Figure 6 5-year average values of democracy scores and the demographic transition XVII

List of tables

Table 1 OLS model

Table 2 Country fixed effects

Table 3 Country and year fixed effects

Table 4 Interactions with income

Table 5 Time of demographic transition

Table 6 Summary of channel analysis

List of Abbreviations

illustration not visible in this excerpt

1. Introduction

“[…] in many ways demographic change can and should be seen as an essential factor of change that has profound implications for society“ (Reher, 2011, p.12).

The demographic transition¹ has been part of a global transition process that involved changes in politics, economics and society in almost every country. These different processes all have partially overlapping causes and consequences. Therefore, it seems obvious that these movements of change are all interrelated, which is supported by an abundance of research on causal relationships between economic growth and democracy as well as between economic growth and the demographic transition. However, only few works focus on interdependencies between the demographic transition and democratization, even though these two processes happened roughly at the same time and are therefore likely to influence each other. Thus, the goal of this paper is to analyze the relationship between the demographic transition and democracy and the possible channels through which their influences take effect².

First of all, the direction of causality has to be identified in this context. Dyson (2010) and Reher (2011) classify the onset of mortality decline, which constitutes the starting point of the demographic transition as an exogenous event. One might argue that the demographic transition is linked to a country’s health level, which in turn depends on a country’s political system because only in democratic countries medicine is distributed equally and investments in health are undertaken. While this holds true in a not globalized world, in modern times basic health care is provided through the international community. Hence, differences in the standard of health care caused by the respective political systems are not large enough to cause a demographic transition. Therefore, the direction of causation from democracy to demography can be neglected.

Democratization very much depends on factors such as education, income, urbanization and women’s empowerment, which can be identified as consequences of the demographic transition. Hence, the dominant direction of causation seems to go from demography to democracy. The demographic transition brings about the effect on these factors in the following way:

Higher life expectancies and fewer children per family increase investments in education and thereby raise a country’s educational level. Due to changing age structures and better health conditions, more people are part of a country’s labor force increasing its productivity and eventually resulting in economic growth. Fewer children also support women’s empowerment since women have more personal free time besides caring for the household and can participate in the labor force as well. Finally, changing age structures and higher life expectancies also lead to urbanization: more people move to cities due to better employment opportunities and improvements in the health level, which affect mortality rates in cities more than in rural areas because in cities the danger of infectious diseases is much higher.

Furthermore, education, income, women’s empowerment and urbanization affect democracy, which is established in most major works on democracy, as for example by Lipset (1959), Huntington (1991) and Barro (1999).

Hence, it is likely that an exogenous demographic transition causes better education, higher income, women’s empowerment and urbanization, which in turn leads to democracy.

An empirical analysis therefore tests the hypotheses whether the demographic transition has a statistically significant influence on democracy and, if that is the case, through which channels this influence takes effect. For this purpose, panel data from more than 150 countries since 1800 are examined using different linear pooled ordinary least squares and linear fixed effect models. In addition, interactions with income are tested and a channel analysis is executed. In order to ensure the exogeneity of the demographic transition suitable control variables, most importantly per capita income, are included.

Quantifying these effects is important because an advanced understanding of the determinants of democracy facilitates more profound forecasts for democratization movements all over the world and provides guidelines to influence these processes as policy maker. This is highly relevant since many political instable African and Asian countries, such as Congo, Somalia, Afghanistan and Yemen, are currently undergoing the demographic transition or have just completed this process.

The paper is organized as follows: the topic is introduced by first explaining the motivating theory by giving an overview of the literature concerning the demographic transition and determinants of democracy and by demonstrating the connection between these two subjects. Based on the theoretical part the econometrical analysis subsequently follows and finally leads to a conclusion about the stated hypotheses.

2. Motivating Theory

Many authors argue that the demographic transition is part of a global transition process not only involving demographic but also social, economic and political changes³. This temporal parallelism of change processes in economics, politics, demography and on the social level gives rise to the assumption that they are all interrelated.

While the focus of previous research lay mainly on the relationships between economics and democracy⁴ or demography and economics⁵ this paper focuses on the interdependencies between the demographic transition and democratization processes. For this purpose the consequences of the demographic transition and determinants of democracy are analyzed.

2.1. The demographic transition

Thompson already assumed in 1929 that certain historical stages of nations are linked to their regarding mortality and fertility rates. Notestein (1945) then developed a more detailed four-stage model based on observations in the Western World. His findings demonstrate that mortality rates start to decline at one point in the history of a country while fertility rates remain high, which leads to population growth. With a delay, however, fertility rates start to decline as well and once mortality and fertility rates both reach a low steady state the population stops growing. An illustration of this process according to Sunde&Cervelatti⁶ (2011, p. 103) is presented in figure 1. MT denotes the onset of the mortality transition, which is succeeded by population growth. With a time delay fertility rates start to decline as well (FT) and thus population growth slows down. This process is called “demographic transition”, which is defined in this paper by two criterions that have to be fulfilled by a country in order to be classified as post-transitional⁷ : the life expectancy has to be above fifty and crude birthrates⁸ below thirty. These criterions are two out of three used by Sunde&Cervelatti (2011, p.107) in accordance with Chesnais’ (1992, p.19) norms. The third one, a sustained decline in fertility, is hard to quantify and left out because the criterion of crude birth rates below thirty is more restrictive and therefore ensures the satisfaction of the third criterion as well.

Another definition of demographic transition is given by Reher (2004, p.21), who identifies the starting point as “the beginning of the first quinquennium after a peak, where fertility declines by at least 8% over two quinquennia and never increases again to levels approximating the original take-off point“ (Reher, 2004, p.21). The dates of the demographic transition computed by employing above mentioned criterions as well as Reher’s specification are presented in appendix B.1.

Thus, a pre-transitional society with high mortality and fertility rates is rather young compared to a post-transitional society that is characterized by higher life expectancies and lower fertility rates. A society’s changing age structure is considered as one of the most important features of the demographic transition (Reher, 2011, p.14).

illustration not visible in this excerpt

Figure 1 The demographic transition

Empirical studies confirm that a general pattern of an antecedent mortality decline and a following fertility decline exists, even though the time lag between these two developments differs across countries⁹. Reher (2004), for example, observed this demographic process in most countries in the past 150 years and categorizes countries according to the date of their demographic transition, whereby he identifies four groups. The first group constitutes the “forerunners”, whose fertility decline started before 1935 and which are mainly European and a few American countries. The so-called “followers” are countries that experienced their fertility decline between 1950 and 1964 and are either somehow influenced by the Europeans, via power regimes or ethnicity, or countries that are distinguished because of their special regional position such as India, Singapore and Japan. The fertility decline of the third group took place between 1965 and 1979 and therefore, this group is labeled “trailers”, which includes the rest of Latin America and some African and Asian countries. The last group is referred to as the “latecomers”, whose demographic transition started only after 1980 and which consists of some Asian and most of the African countries¹⁰. It is important to note that the general pattern of the demographic transition has been the same in all countries - only the time of the onset and the duration of the transition differ.

There is, however, no consensus about possible causes of the mortality decline because it occurred worldwide under different circumstances and, thus, no one has succeeded in establishing a general pattern, which could be employed to explain the occurrence of the demographic transition in all countries yet. Therefore, some authors, such as Dyson (2010) and Reher (2011)¹¹, claim that the demographic transition occurs exogenously. As opposed to this exogeneity assumption, one might argue that health care (and thereby the occurrence of a demographic transition) is linked to a country’s income and its political system. That is, because only rich nations can afford medicine and only democratic nations have an interest in distributing it, which would mean that the onset of the demographic transition depends on a country’s political and economic situation¹². Globalization, however, raised awareness for the health situation worldwide and international organizations as well as foreign governments have started to provide basic health care for people in less developed countries. Furthermore, globalization facilitates the adoption of new medical inventions and progress in research so that even poor countries can benefit from the latest scientific achievements. This means that in a globalized world a country’s income and political system are not strong enough to cause a demographic transition.

Thus, the argument that income or the political system have a significant effect on the demographic transition might be valid for the 18th and 19th century- in modern times, however, national income and democracy levels only play a minor role in determining a country’s health situation¹³.

Besides from income and the political system as the main threats to the exogeneity assumption of the demographic transition, other factors such as education, women’s empowerment and urbanization might still be correlated with the occurrence of the demographic transition. Therefore, the empirical models account for these factors. Hence following analyses consider the demographic transition as a predominantly exogenous event.

2.2. Democracy

There exist a variety of definitions of democracy and the most common ones in recent literature are based on Schumpeter’s (1947) characterization¹⁴, such as, for example, Huntington’s (1991,p.6) description: “The central procedure of democracy is the selection of leaders through competitive elections by the people they govern“. He points out the significance of a free and fair election process in order to show the importance of a high participation rate to achieve a “people’s rule”¹⁵.

Lipset (1959), in accordance with Schumpeter (1947) and Weber (1946), defines three conditions that classify a democracy: Firstly, a “political formula”, which is a value system that legitimizes the system, specifies institutions and is accepted by all citizens, is necessary. Secondly, one set of political leaders in office is as important as, thirdly, several sets of political leaders out of office that form an opposition. Furthermore, “opportunities for changing governing officials” (p.71) have to be allowed and the “largest possible part of the population” (p.71) has to be able to influence political decisions. He thereby underlines not only the necessity of a balanced distribution of power to avoid conflicts and autocratic movements, but also the necessity of equal and free participation opportunities for the citizens.

The concepts of freedom along with equality among all citizens appear in most work about democracy and are therefore the underlying definition applied in this paper. The criterions to quantify this definition are discussed in the empirical part.

In 2011, 87 countries, representing 43 percent of the global population, are considered democracies, while 48 countries, representing 35 percent of the world’s population, are classified as “not free” according to Freedom House¹⁶. The degree of freedom and democratization in the world is higher than ever before in history. But still, more than half of the world’s population lives in unfree or only partially free countries. Questions regarding reasons for democratization movements are often raised in this context, for example: which factors determine democracy and the different timing of democratization processes?

The temporally different occurrence of democratizations is examined most famously by Samuel Huntington, who identifies three different democratization waves¹⁷, each succeeded by reverse movements. The first wave had its origins in the American and French revolution and took place during the relatively long period between 1828 and 1926 mainly in Europe, but also in North America and in a few Latin American countries. The second one occurred after World War II, when Allied occupation promoted democratization processes, most of Latin America democratized and the end of Western colonial rule produced new states, some of which became democratic as well. The third wave started in 1974 and includes many democratization processes in Latin America and South and Eastern Europe after the collapse of the Soviet Union (Huntington, 1993, p.13 f.)¹⁸.

An abundance of authors addressed the question about causes for democratizations and detected determinants reaching from economic to historical and social factors. Most major works on democratization- except Acemoglu et al. (2008)- agree that economic factors are favorable for democracies. They, however, have not reached a consensus concerning the exact pattern of these effects¹⁹.

Other discussed determinants of democracy include education, health, social institutions, religion, historic factors (such as, for example, a colonial history), urbanization, ethnologic fragmentation and geography²⁰. Analyzing all these factors would go beyond the scope of this paper, and therefore the focus in following sections lies on factors that are at the same time considered as possible consequences of the demographic transition.

2.3. Interdependencies between the demographic transition and democracy

Since the demographic transition and democratization processes often happened roughly at the same time, it is surprising that only a limited amount of literature focuses on the political consequences of demographic transitions while many authors examine the effects on the economy and on society.

As explained in section 2.1, the occurrence of a demographic transition is, in modern times, largely independent from the political system. Thus, the following section analyzes reasons and possible channels through which demography influences democracy. While the reverse direction of causation cannot be ruled out completely, the causation from the demographic transition to democracy is considered as the dominant direction.

As one of the few authors who addresses the political implications of a demographic transition, Dyson (2010) points out the higher confidence people have in their future after the demographic transition and the resulting increased awareness for political matters. “The rise in the number of years that they could reasonably expect to live, surely had an influence in helping people to think more about issues of justice and political equity” (p. 202). Moreover, he emphasizes certain outcomes of the transition such as urbanization and population ageing that eventually lead to democratization.

Another approach was developed by Cincotta (2007), who argues that political systems in pre- transitional, young societies are more likely to be instable than in post-transitional societies. This youth bulge argument, derived from civil conflict research²¹, is connected with the logic of authoritarian bargains²² in order to explain the occurrence of authoritarian regimes in pre- transitional societies. That is, before the demographic transition political instability is more likely to occur due to the young age structure of society and because of this insecure environment people are willing to exchange political rights for security. Hence, democracy is less likely to arise in pre-transitional than in older, post-transitional societies. This argumentation is based on the assumption that if people do not have a reason (in this case higher levels of security) to call for an authoritarian regime the chances for democracy increase. Cincotta’s approach, therefore, rather focuses on the determinants of authoritarian regimes than on the determinants of democracy.

This paper, however, focuses on the direct effects of the demographic transition and its consequences on democracy and consequently follows Dyson’s approach (2010). It is likely that longer time horizons and more life experience promote democracy because people have a broader array of perspectives and are able to evaluate a political system more critically. In addition, the longer people expect to live and work in a society the more they want to participate in the political system that sets the general framework for how society works. Furthermore, more life experience also makes people more familiar with different political systems and the respective advantages and disadvantages. Thus, a movement towards democracy is more probable in societies with higher life expectancies than in others.

Besides from the effect of higher life expectancies on democracy, consequences of the demographic transition such as education, urbanization, women’s empowerment and economic growth also play an important role since they are at the same time determinants of democracy. There might also be other factors that connect the demographic transition and democracy. The aforementioned ones, however, seem to be the most significant ones and therefore the focus in the following course of this paper will lay on these four.

In order to deduce these relationships, the subsequent sections explain these factors with respect to the effect the demographic transition has on each one of them and the influence of these factors on democracy.

2.3.1. Education

Lower fertility rates and higher life expectancies do not only change the structure of society as a whole, but also family structures. Fewer children per family increase the share of disposable income for each child’s education and higher life expectancies increase the return on investments in education. Declining mortality rates also increase the incentive for societal investments in education because society benefits from an educated labor force for a longer time period. Bar and Leukhina (2011) argue that lower mortality rates lead to better knowledge transmission because individuals are carriers of ideas, which also contributes to a higher level of education in posttransitional societies than in others. Empirical studies support the positive relationship between the demographic transition and education as well (e.g. Lorentzen, 2005).

The second part of the argument - the influence of education on democracy - is a rather old field of study. John Dewey argued in favor of education to obtain democracy as early as 1916 because in educated societies people understand political contexts better, are encouraged to think about different forms of governance and question critically the current political situation. In the past

fifteen years, however, a passionate debate about the empirical evidence for this relationship has arisen. Barro (1999), for example, states in his study about the origins of democracy that the propensity for democracy rises with years of schooling. Even though Acemoglu (2005) criticizes Barro’s and other’s studies for using wrong models²³ and, in his analysis, finds no evidence for a link between education and democracy, many other authors succeeding him, who use revised and more sophisticated models such as Glaeser et. al (2006), Papaioanou et al. (2008),Wacziarg (2011) and Hegre el. al (2012), support the idea of a causal relationship between education and democracy theoretically as well as empirically.

Hence, democracy is very likely to occur in more educated societies. After a demographic transition people invest more in education, a better knowledge transfer takes place and more people have more time to study so that better education is a direct result of this transition and thereby is one element that connects the demographic transition and democracy.

2.3.2. Urbanization

“Any account of the demographic transition that fails to include urbanization as one of its major components is seriously incomplete“ (Dyson, 2011). Studies on urbanization often refer to economic changes to explain a drift towards cities. Mortality rates and population growth, however, are an important determinant of urbanization as well. Therefore, de Vries (1990) developed the „stylized sector-specific model of the demographic transition“ and its implications for urban growth. In pre-transitional times, death rates in urban areas are usually higher than in rural areas because infectious diseases, communicated by contagion in confined space, were the main cause for high mortality rates. The demographic transition then reduced mortality rates through better health care, with the result that the urban mortality rate fall more rapidly than the rural mortality rate. Thus, the urban population increased faster than the rural population as long as the birthrate was higher than the death rate. Once birth and death rates reached the same level, the process of the demographic transition was completed and the result was an increase of the urban population. Another connection between the demographic transition and urbanization is a temporary strong population growth in general, which does not only contribute to relative faster growth of cities compared to rural areas but also to new city foundings, resulting in absolute urban growth. Empirical evidence for the relationship between the demographic transition and urbanization is given by Woods (2003). In addition, Dyson (2010, 2011) and Canning (2011) establish a similar argumentation and agree that mortality decline increased the urban population.

Urbanization as a determinant of democracy is extensively discussed since “democracy made its first effective appearance in the Greek city states” (Laski, 1937, p. 76). Max Weber (1950, p. 315) characterizes cities as a certain type of political community, which is confirmed by Lipset’s analysis about the requisites of democracy (1960)²⁴. One of the main arguments in this discussion is that people in cities can organize themselves more easily and have a better notion of inequalities and the political situation, which promotes exercising political rights and democratic movements. Moreover, cities are characterized by a growing middle class²⁵, which is often identified as an important component of democracy (Lipset, 1959).

Dyson (2010) emphasizes the different distribution of political power in cities, which is promoted through new administrative and governance structures that become necessary in urban areas. But he also argues that democracy can only arise in moderate growing cities because cities growing too quickly, which have to deal with space problems, involve the danger of civil conflict and political unrest.

Barro (1999), on the other hand, claims that according to his empirical studies democracy tends to fall in line with urbanization, but does not give a profound theoretical explanation for this negative relationship. Therefore, these results could have either occurred due to theoretical explanations, which have not been covered so far²⁶ or due to estimation errors. Since the majority of authors, however, argue in favor of a positive relationship between urbanization and democratization this paper follows this approach.

Urban governance structures, a broader middle class and better communication facilities than in rural areas all have a positive effect on democratization. Thus, urbanization can be seen as another outcome of the demographic transition that causes democratization.

2.3.3. Women’s empowerment

Declining fertility rates have a huge impact on women’s lives because caring for fewer children gives them more time to spend, for example, at school or at work. Being more educated and pursuing a profession also goes along with becoming independent, self-confident and accepting social and political responsibilities. There are many empirical studies that find a large negative effect of a high fertility rate on female labor force participation (Bloom, Canning, Fink, Finlay, 2009) and on female economic and political empowerment (Canning, 2010). Dyson (2001, 2010) also claims that fertility and mortality declines lead to a reduction in gender differentiation and higher female autonomy.

Women’s empowerment contributes to democratization because equality and broad participation are its prerequisites. Barro (1999) agrees with this position by stating that “democracy is also negatively and significantly related to the gap between male and female primary attainment”, meaning that higher levels of democracy can be reached in more equal societies. Waylan (1996) emphasizes the substantial role of women in bringing about the collapse of authoritarianism in Latin America and Eastern Europe.

Consequently, the demographic transition, particularly fertility decline, promotes women’s empowerment, which in light of higher equality among the citizens determines democracy.

2.3.4. Economic growth

There are several arguments in favor of a positive influence of the demographic transition on economic growth. Higher life expectancies and more people in an employable age increase the labor force, and it becomes more productive through higher health levels (Bloom, Canning, Sevilla, 2001). Lorentzen et al. (2005) argue that high pre-transitional mortality rates reduce economic growth because they shorten time horizons and lead to risky behavior and lower investments in physical and human capital. They also see mortality rates as a source of the poverty trap because poor health conditions and a weak economy reinforce each other. Hence, post-transitional societies are characterized by better prerequisites for economic growth. Not only Galor and his Unified Growth Theory (2011), but also Canning (2011), in accordance with Dyson (2010) and Dahan & Tsiddon (1998) suggests that the demographic transition played a major role in the emergence of economic growth, primarily because of mortality decline and age- structural changes. This relationship was questioned by Acemoglu & Johnson (2006), who do not obtain evidence that large increases in life expectancy lead to increases in per capita income. Sunde&Cervelatti (2011), however, find in post-transitional times a positive relationship of life expectancy on economic growth, even though the effect before the demographic transition is ambiguous. Hence, according to the overwhelming majority of literature, economic growth is at least partially determined by the demographic transition.

The tendency that richer countries are often more democratic can be observed worldwide. If there is a direct connection between income and democracy is, however, controversial. While Acemoglu et al. (2008) did not find evidence for a causal relationship in empirical studies, Barro (1999), for example, did. Most major works on democracy (Lipset,1959, Huntington 1991,

Rueschmeyer/Stephens/Stephens, 1992, Tavres/Wacziarg, 2000, Murtin/Wacziarg, 2011) consider economic performance as a determinant - perhaps even the main one. They argue that wealth is a prerequisite for intelligent participation in political processes since people have the ability to think about broader questions, such as governance systems, as they are usually better educated. It also changes the social position of workers and strengthens the middle class, which has a positive impact on civic participation. Economic growth also weakens the power of a small economically and therefore often politically leading class and raises the influence of the working class.

Hence, the demographic transition promotes economic growth, which is a determinant of democracy so that income is, besides education, urbanization and women’s empowerment, a fourth channel through which demographic transition and democracy are connected.

Accordingly, this analysis yields two hypotheses, which will be examined in the empirical part of this paper

(1.) The demographic transition has a positive influence on the rise of democracy.
(2.) The effect of the demographic transition on democracy works primarily through the channels of education, women’s empowerment, urbanization and economic growth.

3. Data and descriptive statistics

3.1. Data

In order to support these hypotheses the following section provides an empirical analysis based on different linear panel data models for more than 150 countries over the past 200 years. Panal data are advantageous compared to cross-sectional or time-series data because they „usually give the resarcher a large number of data points, increasing the degrees of freedom and reducing the collinearity among explanatory variables-hence improving the efficiency of econometric estimates“ (Hsiao, 2003, p.3). Moreover, panel data are particularly useful to analyze the dynamics of change regarding demography and democracy across different countries. There are, however, also limitations of using panel data, such as sample selection bias because data availability might not be equal across countries and years.

Demographic data, such as life expectancy and fertility rates, were taken from the World Bank’s world development indicator database (WDI) and are available since 1960. This database also contains statistics on education, urbanization, female labor force participation and GDP. GDP and other economic variables were also drawn from Acemoglu&Johnson (2007) and Banks’ cross national time series database²⁷. As another measure of education, the Barro-Lee dataset of educational attainment in the world is used, which provides data about average years of schooling in 5-year intervals from 1950 until 2010. As measure for democratization, Polity IV data²⁸ as “the most widely used data resource for studying regime change and the effects of regime authority“²⁹ are employed, which cover the time span from 1800 to 2010 for more than 150 countries. A detailed description of all variables and sources is given in appendix A.

3.1.1.Demographic data

The most important demographic data are mortality and fertility rates in order to estimate the date of the demographic transition. As one measure, a binary variable labeled demographic transition (“demotran”) was created, which turns 1 if life expectancy is higher than fifty and crude birthrates fall below 30/1000. If one of these criterions is not met the variable stays 0. Life expectancy as taken from the World Bank dataset is defined as “number of years a newborn infant would live if prevailing patterns of mortality at the time of its birth were to stay the same throughout its life“³⁰ and crude birth reflect the number of live births occurring during the year, per 1,000 population and are estimated at midyear. As the World Bank data are only available since 1960, all analyses including the binary variable “demotran” only take into account countries that experienced their transition after 1960.

Data about the onset of fertility decline from Reher’s analysis (2004)³¹ are taken as another measure for the demographic transition.

The difference between these two measures amounts to 15 years on average with a standard deviation of 10.3, which is due to the fact that Reher’s data underlies a different definition for the demographic transition than the one applied for the World Bank data. Also, the data of the World Bank only start in 1960 while Reher’s data go back to 1900. Thus, only data for countries that experienced their demographic transition after 1960 can be compared, which are mainly African, Asian and some Latin American countries (whose data are less accurate). Taking this into consideration, the two measures from completely different datasets are quite close in terms of the time horizon they reflect.

In addition, the time difference between the present and the time of the demographic transition (yearssincedt)³² were calculated on the basis of Reher’s data. For summary statistics of the demographic data see appendix C.

In the following course of the analysis WDI data underlie the binary variable “demotran” and Reher’s data are the basis for the continuous variable “years passed since (pending to) the demographic transition”. It is important to keep in mind that the variables reflect different time horizons³³ and different meanings. While the binary variable “demotran” measures the effect of the occurrence of the demographic transition (in the past) on democracy today, the continuous variable represents the effect that years passed since or pending to the demographic transition have on democracy today. Thus, the results from these different strategies to examine the stated hypotheses have to be interpreted with respect to their respective time horizon and meaning.

[...]

---

¹ Characterized as transition from high mortality and fertility rates to low mortality and fertility rates. Declining mortality rates (and thereby rising life expectancies) followed by falling fertility rates, eventually, lead to a higher proportion of old people in society.

² holding income constant

³ See, for example, Dyson (2010) or Reher (2011)

⁴ Barro (1996), Glaeser et al. (2004), Acemoglu et al. (2008),

⁵ Bloom,Canning&Sevilla (2001), Reher (2011), Lorentzen,McMillan&Wacziarg (2005), Acemoglu&Johnson (2006), Cervelatti&Sunde(2011)

⁶ The figure reproduces Fig. 1.1 from Chesnais (1992) and Fig. 4.2 from Livi-Bacci (1992).

⁷ Post-transitional: Countries that experienced the onset of the demographic transition

⁸ Crude birth rates denote the number of births per 1000 people per year

⁹ See for example Coale (1975)

¹⁰ See appendix B.1 for a summary of the Reher-groups

¹¹ Canning (2011,p.3): “tackling the central thesis of the book [Dyson’s Population and development, 2010] which is that the demographic transition is a largely self-contained process that proceeds independently of social and economic development. The argument is that the mortality transition occurs exogenously and is the cause of a subsequent decline in fertility “ Reher (2004,p. 12): The demographic transition “is considered as a largely autonomous process“.

¹² Franco et al. (2004) and Cutler et al. (2006), for example, argue that political institutions significantly affect health and Pritchett &Summers (1996) state that income has a significant effect on health.

¹³ Smith (2003) and Adams et al. (2003) find that changes in income and lagged values of income do not predict changes in health. Christopher Martyn questions the strength of the evidence for politics being a determinant of health in Franco’s paper (2004) (p.1423)

¹⁴ "The democratic method is that institutional arrangement for arriving at political decisions in which individuals acquire the power to decide by means of a competitive struggle for the people's vote" (as cited in Barro, 1999, p. 160).

¹⁵ “Democracy” comes from the Greek words δῆμος ( d ē mos ) meaning "people" and κράτος ( kratos ) meaning "power“

¹⁶ The number of countries qualifying as partly free stood at 60 and they were home to 22 percent of the world’s total population (http://www.freedomhouse.org/report/freedom-world-2012/full-report-essay). Freedom House is a US-based non-governmental organization that supports democratic change, monitors freedom, and advocates democracy and human rights around the world.

¹⁷ “A wave of democratization is a group of transitions from nondemocratic to democratic regimes that occur within a specified period of time and that significantly outnumber transitions in the opposite direction during that period of time”(Huntington, 1993, 15)

¹⁸ See appendix B.2 for a summary of countries belonging to the different democratization-waves

¹⁹ See Lipset (1959), Huntington (1991), Rusechemeyer,Stephens&Stephens (1992) or Barro (1999)

²⁰ Lipset (1959), Huntington (1991), Barro (1999), Papaioannou and Siourounis (2006), Glaeser et al. (2007), Hegre et al. (2012)

²¹ See Goldstone (1991, 2001) and Collier (2000) for theoretical explenantion and Urdal (2006) for empirical evidence

²² “The social contract in dictatorships is commonly explained as an “authoritarian baragain” between rulers and citizens by which citizens relinquish political rights for economic security“ (Desai, Olofsgard, Yusuf, 2009, Abstract)

²³ According to Acemoglu country and time fixed effects are crucial for regressions that estimate the effect of education on democracy.

²⁴ See also Mumford (1995, p.21) and Amin & Thrift (2004, p. 231) and their argumentation in favor of a positive correlation between cities and democracy

²⁵ The middle class grows in cities extraordinary fast due to the economic structure of cities (new type of occupations are available)

²⁶ For example that autocratic regimes can control cities easier than other areas

²⁷ This database was launched by Arthur Banks in 1968 at the State University of New York at Binghampton and provides data for more than 200 countries from 1815 until 2007 (Arthur S. Banks, "Cross-National Time-Series Data Archive (CNTS) 1815-2007", http://hdl.handle.net/1902.1/11448 Databanks International [Distributor] V1 [Version])

²⁸ The Polity IV Project is is under the direction of Dr. Monty G. Marshall and supported by the Political Instability Task Force, Societal-Systems Research, and Center for Systemic Peace.

²⁹ http://www.systemicpeace.org/polity/polity4.htm

³⁰ http://data.worldbank.org/indicator/SP.DYN.LE00.IN

³¹ Using data from UN Demographic Yearbook Reher defined onset of fertility decline as follows: It “has been set at the beginning of the first quinquennium after a peak, where fertility declines by at least 8% over two quinquennia and never increases again to levels approximating the original take-off point“ (Reher, 2004, p.21)

³² This measures contains negative values if the demographic transition lies in the future an spositive values if a country has already undergone the demographic transition.

³³ WDI 1960-2011 and Reher 1865-2000