1) What it Takes to Shape Economic Growth? or 2) Assessing the Impact of Political Leaders on Economic Growth

<span class='text_page_counter'>(1)</span><div class='page_container' data-page=1>

1) What it Takes to Shape Economic Growth?

or 2)Assessing the Impact of Political Leaders on

Economic Growth

Julieta Peveri

May 2019

Abstract

This paper analyses the context in which national leaders can shape economic

growth and it assesses which individual characteristics are linked to their

perfor-mance. By comparing growth variations across transitions in which the identity

of the ruler is not likely to be dependent on the economic conditions, I found

ev-idence that leaders matter both in autocracies and democracies even though the

level of democracy and development shrink the magnitude of leaders’ effects.

Be-sides, rulers’ transitions have a higher impact on growth whenever one of the leaders

has the possibility of running for a second mandate. In some specifications, I also

found evidence of the positive effect of university education and a mitigate effect

of tenure on growth shifts. Regarding directional results (i.e nominal growth

varia-tions) individual traits play a larger role. Younger leaders and the ones with more

experience in politics are associated with a better economic performance. Some

career backgrounds are also relevant for establishing the leader’s quality as well

as the existence of a term limit. In some of the approaches used, more educated

</div>
<span class='text_page_counter'>(2)</span><div class='page_container' data-page=2>

1

Introduction

At the intersection between Politics and Economics the leader’s identity is becoming a

topic of increasing interest. But can they matter to the point of shaping economic growth?

From a theoretical classic framework the answer would be negative. Indeed, most of the

academic research has highlighted, as determinants of growth, factors such as capital

accumulation, innovation, geography or education. Hence, while those characteristics

have been stable over time, growth rates have not, particularly in developing countries
(Pritchett, 2000). Alternative explanations have tried to understand the growth rates

volatility. For instance, Easterly et al. (1993) show the importance of shocks, particularly

those in terms of trade as a determinant of the long run growth variance. As those

variations are random, they conclude that high growth rates are likely due to good luck.

In that sense, they show a similarity with the classic explanations by delegating the blame

of underdevelopment or the merit of economic miracles to factors beyond our control.

In this paper I focus on the potential role of political leaders to account for growth

instability. The challenge of identifying a causal impact is that growth can also trigger

leaders transitions. In democracies, reelections can be tied to the previous economic
performance and in autocracies, coups are less likely to arise during economic booms

(Londregan and Poole, 1990). Jones and Olken (2005) tackle this endogeneity issue by

focusing on transitions where leaders died in office by natural death, considering those as

unpredictable events. Indeed, they found positive evidence suggesting that rulers matter

for growth but only under autocracies. The first contribution of this paper is to revisit

their work by using more recent data. Thereafter, I find positive evidence for the effect

of leaders also under democracies. Hence, this strategy is mainly limited by the number

of observations and the lack of representatives of the ones who died in office.

In order to go deeper in the analysis, a larger number of exogenous transitions is

needed. For this purpose, I also include the ones where the exiting leader was not able
to run for a reelection due to a term limit constraint. Besides, I select transitions where

the entering ruler either won the elections by a small margin of victory, took power

through royal succession, by constitutional order, was elected by an elite or as an interim

leader (as long as the predecessor exited in a regular manner). Using those four hundred

cases, it is possible to give a brighter insight of the context in which leaders are more

likely to have a stronger impact on growth by including further contextual variables and

leaders’ characteristics. The evidence suggests that leaders have more power in

low-income countries and in more autocratic ones, being less constrained by the institutions.

The possibility of reelection also accentuates the growth volatility. Overall, leaders’ traits
do not play a role to determine the amplitude of their effect in the economy.

The other main question addressed in this paper is whether there are observable

</div>
<span class='text_page_counter'>(3)</span><div class='page_container' data-page=3>

and bad leaders. While at the local level, many studies have given attention to politicians’

characteristics such as gender, education and age, at the national level, the empirical

evidence is limited. For instance, Besley et al. (2011) extended the analysis of Jones and

Olken (2005) and tested if the impact of leaders (also in transitions where the ruler died

by natural death while in office) varied according to his educational attainment. Again,

the analysis was limited to the number of observations. Here, I explore and interact

further individual features and control for contextual variables. While I found that the

robustness of education is not strong (even though the signs are always consistent with the

previous findings), the results suggest that young leaders and the ones with more years

of experience in politics are associated to a better performance. Lawyers and military
leaders also appear to boost economic growth.

The paper is organized as follows. Section 2 reviews the existing related literature;

Section 3 describes the data and sources used in this work; Section 4 explains the

method-ology used in this paper; Section 5.1 revisits Jones and Olken (2005) approach focused on

the rulers who died in office by natural death, Section 5.2 includes new cases of exogenous

transitions; Section 6 provides some robustness about the validity of the methodology and

alternative specifications for growth comparison and Section 7 sets out a conclusion.

2

Conceptual Background

2.1

The role of the leader’s identity

There is no consensus in theory about the significance of political leadership. Even in

the historical literature the opinion is divided. Defenders of the Great Man theory such

as Thomas Carlyle or Sidney Hook argue that history can be largely explained by

deci-sions and actions impulsed by certain leaders. On the other extreme, many philosophers

consider historical events and leaders as determined by the society it self. Spencer (1892)

states: “You must admit that the genesis of a great man depends on the long series of

complex influences which has produced the race in which he appears, and the social state
into which that race has slowly grown. ... Before he can remake his society, his society

must make him.”

When it comes to the Political Economy literature, the ambiguity about rulers’

iden-tity is also there. In the branch of political competition, models that rely on the median

voter theorem predict a convergence of all candidates on one policy platform, while others

that account for parties and politicians ideologies lead to an equilibrium that is linked to

the identity of the political leader.

Another concept of great interest for this work is the role of institutions that has been

</div>
<span class='text_page_counter'>(4)</span><div class='page_container' data-page=4>

(2004). Even by agreeing on the importance of institutions, the role of the political

rulers is not clear. It can be argued that when institutions are strong they would act

as a constraint on the incumbent leader. In this sense, rulers shouldn’t have an impact

on consolidated democracies but they might so in autocracies. For instance, Acemoglu

et al. (2003) show that once controlling for institutions, macroeconomic policies have a

small impact on growth volatility. Nevertheless, being the nation’s leader the main agent

among political institutions, they could have through this power an indirect impact on

economic and social outputs in any regime.

Even though we reach a consensus that politicians can matter for some economic and

social outputs, it is still a strong statement to argue that they can shape the growth
pattern. To the best of my knowledge, the only study to assess national leaders as

an alternative explanation for the fluctuating path of growth is the one by Jones and

Olken (2005). They explore whether the changes of national leaders could be related

to the variability associated with growth rates. Their empirical strategy was to exploit

transitions in which the leader died by natural death while in office, considering those

as “random” ones. The authors tested whether the growth pattern before the leader’s

death was statistically different after his replacement. After finding a positive answer

in an autocratic context, they reproduced the analysis about the effects of leaders on

particular types of policies outcomes and they found a significant one on monetary policy
but ambiguous evidence for changes in physical and trade policy. There is also evidence

provided by Fatas and Mihov (2013) that policy volatility can have a strong and negative

impact on growth.

2.2

Individual characteristics, choices and performance

If identity appears to matter, it is natural to ask which of the individuals

character-istics lead the magnitude and the direction of the effect. Following the work of Jones

and Olken (2005) cited before, Besley et al. (2011) extended the analysis and tested if

the impact of a leader varied according to his educational attainment. They also found a

positive answer, consistent with the findings related to the impact of education on

earn-ings in labor economics. This branch of the literature states that education is seen not
only as the reflect of specific knowledge and skills taught in the academic space but also

as a proxy to unobserved ability. In that sense, firms are willing to pay higher salaries

to more educated workers and more experienced ones because they associate those

indi-viduals with a better performance. By extrapolating the argument for a nation’s head,

one may expect that the more skilled, competent and qualified the leader is for his

po-sition, the better will be the decisions he will lead to, thanks to his expertise, academic

knowledge and the unobserved ability that encompasses motivation, intelligence, facility

</div>
<span class='text_page_counter'>(5)</span><div class='page_container' data-page=5>

At the local level, studies that assess the characteristics of leaders on policy outcomes

are more frequent as it is easier to design an empirical strategy to identify a causal

effect. For instance, through a regression discontinuity design Lahoti and Sahoo (2017)

exploit data from Indian politicians, finding that educated leaders were not necessarily

more competent in improving children’s education. Controversely, Diaz-Serrano and Prez

(2013) study concludes that the educational attainment of population does improve when

a leader with higher education remains in office. Besides education, the impact of the

leader’s gender has also become an active research topic. Chattopadhyay and Duflo (2004)

exploit data from India where some council head positions have been randomly allocated

to a woman. They find that leaders spend more in infrastructure that is directly relevant
to the needs of citizens of their same gender. Brollo and Troiano (2016) use a discontinuity

approach based on close elections and found that in Brazil female mayors are less likely to

engage in corruption, to hire fewer temporary public employees during the electoral year

and they also tend to attract less campaign contributions when running for reelection.

Using the same empirical strategy, Alesina et al. (2015) analyze the effect of leaders’ age

on political governance, reelection rates and policies in Italian municipalities.

Horowitz and Stam (2014) who study how leaders’ characteristics affect military

de-cisions, provide an intuition illustrated by Figure 1 of how leaders’ experience may affect

policies choices.

Figure 1: Theoretical relationship between leader’s experiences and policy outcomes

Life experiences Leader beliefs/

risk attitudes Policy outcomes
Domestic politics

Source: Horowitz and Stam (2014)

The managerial literature can also provide valuable analysis through the importance

they have given to the impact of the CEO on the performance of a firm by analyzing their

leadership style, their risk-taking behavior and some personal traits such as age, gender

or family social class. One specific paper related to the research question stated above is

the one written by Bertrand and Schoar (2003) that analyzes the effect of the CEO on the

performance of a firm through estimating their fixed-effects. Their study concludes that

differences in investment, financing and other organizational strategy variables depend
on the specific characteristics of the firm’s manager.

</div>
<span class='text_page_counter'>(6)</span><div class='page_container' data-page=6>

schol-ars have analyzed the behaviour of leaders since long time ago. The main approach used

is the psychobiographic method which involves an examination of the life history of rulers.

Many of the personality theorists seek to identify traits (defined as personality

character-istics that are stable over time and different situations) motivations and cognitive style

variables to assess how all these features shape styles of decision making, interpersonal

interaction and management in office (Cottam, 2004). For instance, Adorno et al. (1950)

(cited in Cottam (2004)) study the personality associated to authoritarian leaders

show-ing that it was a result of childhood experiences that led to a weak ego. Regardshow-ing

democratic chiefs of state, Dean Keith (1993) analyzes the way we judge who would be

the best leader for a nation among a pool of candidates. As the author states: “Some
character traits go better with certain policy stands or performance expectations. For

instance, our assessments of a candidate’s willingness to solve the problem of the

home-less may depend in part on our perceptions of how compassionate we perceive him or her

to be. Similarly, we may feel that certain personality traits may enhance a candidate’s

prospects for effective performance” (Dean Keith, 1993).

3

Data

To begin with the analysis of the relationship between political transitions and growth,

Figure 2 plot the evolution of the real growth trend for four selected countries where the

vertical lines represent the transitions of chiefs of state. At first sight, it is possible to

detect that a great proportion of those political changes lead to non negligible shifts in

the real GDP. Besides, contrary to what we could expect in many of those transitions

the variations on the economic output are manifested immediately.

One of the main reasons of the lack of empirical research about national leaders in the

economic field may be related to data limitation. Fortunately, in recent years progress

has been made. In 2004, the Archigos dataset (Goemans et al., 2009) with information
about the year and the nature of entry and exit of national leaders in 188 countries from

1875 to 2004 was published. It also contains few other personal variables such as gender,

the year of birth and the year and cause of death. From then on, other contributions

have been made by completing this set with further information. This paper combines

the LEAD base (Elli et al., 2015) with the Cursus Honorum (Baturo, 2016) one to obtain

a large number of information related to the leader’s traits and background.

The Leader Experience and Attribute Descriptions (LEAD) dataset (Elli et al., 2015)

covers information about 2 964 national leaders from 1840 to 2000. It is based on the

Archigos dataset but it includes additional variables related to family background,
profes-sional and personal history, health status, education, military experience, among others.

Their creators are using this dataset for international conflict studies. The Cursus

</div>
<span class='text_page_counter'>(7)</span><div class='page_container' data-page=7>

Figure 2: Real growth path and political transitions in selected countries

(a) Central African Rep (b) Argentina

(c) France (d) China

Source: Bolt et al. (2018)

leaders in office from the period between 1960 and 2010. As the LEAD data base, it

also includes educational and career variables as well as other ones related to the family

background. Unfortunately, many of those variables have a lot of missing values, and as
much I think they may affect the quality of a leader, I do not consider them.

Gender is one of basic traits defining an individual. However, in the remaining sections

I do not analyze the effect of this characteristic do to the few national female leaders.

In fact, men represent an extremely high percentage of 98% of the dataset (see Table

1). Even if one might think that this proportion is decreasing nowadays, if we focus

on the sub-sample from 1990 to 2004, it only decreased to 95,5% and even goes up to

96,7% when we consider the period 2000-2004. In his book, Ludwig (2002) argues that

people associates authority with masculinity traits. More interesting, in his analysis

(previous to 2002), he shows how almost half of the women that have been head of state

gained this position for having been the widows or daughters of previous leaders. Among
the remaining women rulers that become leaders for her own merits, most of them just

</div>
<span class='text_page_counter'>(8)</span><div class='page_container' data-page=8>

Table 1: Descriptive Variables

Variable Mean Std. Dev.

Male 0.9836 0.1269

Age 57.6605 10.8997

N of children 3.6941 5.0299

Married 0.9525 0.2128

Married in power 0.9102 0.2860

Divorced 0.1114 0.3147

Physical health 0.1044 0.3058

Mental health 0.0117 0.1077

Outsider 0.0873 0.2824

Political family 0.1518 0.3590

Lower family class 0.2343 0.4238

Middle family class 0.5522 0.4975

Upper family class 0.2135 0.4100

Divorced Parents 0.0532 0.2528

Mother in labor force 0.1299 0.3363

Only child 0.1389 0.3460

First born 0.4170 0.4933

Middle born 0.3619 0.4808

Last born 0.1630 0.3695

Source: Elli et al. (2015); Baturo (2016). All variables
correspond to the characteristic of each leader in the
year he enters in office. Health’s variables are dummies
coded 1 if the leader was sick under the category.
Out-sider is coded 1 if prior to assuming office a leader can
be arguably regarded and seen as an outsider to the
ex-isting political system. Political family is coded 1 if a
member of the leader’s family had occupied the highest
national political posts in the past, whenever possible
to ascertain.

autocratic regime in the history of the world.

Another basic variable to analyze is age. When plotting the mean age when leaders

entry office in Figure 3 we can see almost no variation from the 90’s on, being the mean

age 54.4 years old (in this period) with a standard deviation of 2.14 years old. Thus, we
can argue that young people do not have more access to the head of nation’s position

than before as we could presume.

Concerning some family variables, it appears that the majority (55%) of national

leaders (for whom this data is available) come from middle class families while a 24%

have grown in a low class family and the remaining 21% were raised in families with a high

socio-economic level. Particularly relevant for the forward analysis are the educational

and professional variables. Figure 4 illustrates the percentage of the overall level of

leaders’ education by decade. It is interesting to see how in the last two decades of

</div>
<span class='text_page_counter'>(9)</span><div class='page_container' data-page=9>

Figure 3: Mean age of leaders when entry into office by year

Source: Elli et al. (2015)

Figure 4: Leaders’ level of education by decade

Source: Elli et al. (2015)

than 15% of leaders have not gone to University over the last decade.

Then, Table 2 highlights that despite chiefs of state come from all kinds of

</div>
<span class='text_page_counter'>(10)</span><div class='page_container' data-page=10>

Table 2: Leaders’ Profession

Occupation Percentage

Politician Career 29,45%

Law 28,58%

Military Career 20,65%

Teacher 12,21%

Activist 11,27%

Aristocrat/Landowner 7,09%

Journalism 6,11%

Writer 5,40%

Economics 5,23%

Labor 4,45%

Engineering 3,64%

Agriculture 3,61%

Medicine 3,54%

Religion 1,79%

Police 1,05%

Science 0,74%

Film/Music 0,27%

Interpreter 0,20%

Source: Elli et al. (2015)

percentage of national leaders with a military career is also high. It is also helpful to

analyze the nature of leaders’ exit shown in Table 6. In Section 5.1 I will exploit data

of leaders who died in office which represent around 8% of the dataset. As we could

expect, leaders who died in office are in average older when taking power (63 years old)

and around half of them had run an autocratic government.

Table 3: Leaders’ Exit Type

Exit Type N of Leaders

Leader lost power through regular means 1442

Leader died of natural causes while in power 176

Leader retired due to ill health 52

Leader lost office as a result of suicide 3

Leader lost power through irregular means 436

Leader deposed by another state 52

Leader still in power (2004) 104

Source: Goemans et al. (2009)

The data for economic growth used in this dissertation is drawn from different versions
of the Pen World Table (Heston et al., 2002, 2009; Feenstra et al., 2015) and the Maddison

Project (Bolt et al., 2018). Figure 5 shows that it exists a positive correlation between the

</div>
<span class='text_page_counter'>(11)</span><div class='page_container' data-page=11>

growth rate associated to having a postgraduate education is not significantly different

from the one associated with a university diploma. This paper also uses data from Beck

et al. (2001) and Ginsburg et al. (2010) to identify mandates where the leader have or

not a term limit constraint.

Figure 5: Correlation between growth and leaders’ educational level

Source: Elli et al. (2015) and Heston et al. (2002)

4

Methodology

In the first part of this paper I will replicate the methodology proposed by Jones and
Olken (2005) but using alternative data sets both for growth and leaders. The starting

idea is that the growth process can be described by Equation 11.

git = vi+ ut+ λlit+ it (1)

where git is the growth rate of country i in period t; vi represents a time fixed effect; lit

is an approximation to the leader’s quality and it is the error term which is assumed to

have mean 0 and variance σ_i2. Just like the authors, I assume region-specific

heteroskedas-ticity and a region-specific autocorrelation process of first order. The hypothesis to be

tested is whether λ is different from 0 meaning that political leaders can contribute to
shape economic growth.

Hence, the main problem of interpreting directly the coefficient of leaders fixed-effects

is that leaders transitions are not likely to be exogenous to economic conditions. As

explained by Jones and Olken (2005), this can arise if the probability that a leader is

selected in a country i at a time t depends, among other variables, on the previous growth

</div>
<span class='text_page_counter'>(12)</span><div class='page_container' data-page=12>

lit =







lit−1 P (δ0git+ δ1git−1+ ...)

l0 1 − P (δ0git+ δ1git−1+ ...)

(2)

where l’ is distributed normal, with mean µ variance σ2

l and corr(l, l’) = ρ. This reflects

the main identification problem when it comes to asses the impact of a leader on growth.

For instance, in democracies where the incumbent president can run for reelection, voters

would, among other things, take into account the economic performance during his term

to either replace or reelect him. Also, in autocracies the probability of revolution to

revoke the dictator is likely to depend on the level of inequality and living standards.

To deal with the endogeneity issue, Jones and Olken (2005) suggest to use leaders

who died in office by natural death so the related date of the transition was not predicted
and assumed to be independent to economic conditions. Rather than focusing on leaders

fixed effects, Jones and Olken (2005) choose to compare differences in the average growth

T periods before the death of a leader and T periods after it. More specifically, being

\

P REz the average growth in the T years before the leader l dies and \P OSTz the average

in the T years that follow the death of the leader, if the growth process is described

by Equation 11, then the difference between \P REz and \P OSTz will be distributed as

follows:

\

P REz− \P OSTz ∼ N (0,

σi

T + 2λ

2

σ_l2(1 − p)) (3)

Yet, if the leader’s quality does not matter, then the distribution of differences in
growth rates before and after the death of a leader will not depend on σl or λ. In that

case, Equation 3 could be rewritten as follows:

\

P REz− \P OSTz ∼ N (0, 2

σi

T ) (4)

The idea by Jones and Olken (2005) is to develop a Wald Test to assess the hypothesis

above based on the following statistics:

W = 1

Z

X( \P OST_z− \P RE_z)2
2σei/T

(5)

I then consider further cases where transitions are also not likely to be dependent on

</div>
<span class='text_page_counter'>(13)</span><div class='page_container' data-page=13>

Θj = 1

(a) if lij−1 died by natural death

(b) if lij−1 could not run for elections due to a constitutional term limit and

lij−1 exited in a regular manner

(c) if lij won the elections by less than 55% of the votes and lij−1 exited in a

regular manner

(d) if lij assumed through royal succession, constitutional succession, elected

by an elite or as an interim leader and lij−1 exited in a regular manner

where lij represents the jth leader of country i.

The first case was analyzed above. In the second case, as the incumbent leader is

unable to run for elections, inevitably a new leader has to take power independently

of the previous growth pattern. Nevertheless, some endogeneity can still be present if

candidates are judged based on the performance of leaders from their same party. In the

third case I assume that if a leader takes power following a close election, his victory can

be considered as random. Finally, I add transitions in which the incoming leader was an

interim one; he assumed power by a royal succession; a constitutional one or if he was

selected by an elite. I assume that in those cases the motivations of changing the leader

were not based on the previous growth path. I eliminate cases in which, even though one

of the previous conditions held, the leader who exited power did it through an irregular
manner (such as coup, revolution or murder).

Just like before, I first tested whether growth rates, N years before lt−1 leaves the

office, are in average significantly different N years after he left with the same Wald

tests previously used. Hence, in addition to confirming that rulers do have an impact on

growth, I also wish to study in which context they are more likely to shape growth, I

privilege another approach by estimating the following equation:

|∆growth\j−1,j| = α + β∆Xj,j−1+ φ1Yj + φ2∆Yj−1+ λRegion+  if Θ(lj) = 1 (6)

where j∆growth\j−1,jj is the absolute value of the average growth difference between

the growth rates in the first three years of leader j (or less if the leader stayed one or two

years) and the previous three (or less) years of his predecessor;1 ∆Xj,j−1is a vector of the

differences of individual leaders or country’s continuous characteristics; Yj and Yj−1 are

vectors of leader or country dummy’s variable corresponding respectively to j’s and j-1’s
term and λRegion are region fixed-effects. Finally, I also use the nominal values of those

differences in growth rates to assess the quality of the leader rather than the magnitude

of his impact in order to assess if there are some observable characteristics that may allow

1_{By not considering the whole term, we mitigate the problem about the correlation between the end}

</div>
<span class='text_page_counter'>(14)</span><div class='page_container' data-page=14>

us to differentiate the rulers that manage to boost economic growth.

In addition to the endogeneity of transitions, there is also the concern that the growth

rate of the first year of each term is likely to be constrained by decisions made by the

previous leader and the last years may not be representative of his average performance

if the motivations are different (specially in transitions with term limits). I suggest to

smooth growth rates as follows:

GDPt= 0.8GDPt+ 0.2GDPt−1 if t = out year (7)

GDPt= 0.8GDPt+ 0.2GDPt+1 if t = in year (8)

This approach slightly moderates the first and last year. Indeed, this seems
appropri-ate as I already assigned the transitional year to the ruler who stayed longer in power.

Thus, during the in and out year, each leader stayed more than six months. Nevertheless,

in Section 6 I explore other weighting coefficients.

5

Results

5.1

Natural deaths’ transitions

I begin by using the same methodology and the same growth rate data as Jones and

Olken (2005) (i.e the real GDP per capita growth rate at constant prices using Laspeyres
index obtained from the Pen World Table 6.1) obtaining similar results and conclusions.2

From column (1) of Table 4 we conclude that the mean growth rate that corresponds to

the five years before a leader dies is statistically different after he is replaced. This effect

is even more important when a leader stays more than two years in power. As Jones

and Olken (2005), from this column we conclude that leaders matter in average under

autocracies but not in democratic regimes. 3 Nevertheless, this measure for the real per

capita growth rate was subject to many criticisms as pointed out by Johnson et al. (2013).

These criticisms were based on the divergence of growth rates between PWT releases.

One of the reasons was that the real GDP using Laspeyres index in older versions
(including PWT 6.1) was calculated as a weighted sum of consumption, investment,

public expenditure and exports, where the weights were the share of those components

on the domestic absorption of the benchmark year. If those shares change overtime,

this technique will lead to biased results. The problem due to the change in weights is

2_{I excluded leaders Burnham and Jagan Cheddi, both from Guyana for comparison reasons as this}

country is not available in version 9.0 of the Penn World Table nor in the Maddison Project dataset.

3_{The J statistic that I obtained is more significant than the one by Jones and Olken (2005). This can}

</div>
<span class='text_page_counter'>(15)</span><div class='page_container' data-page=15>

corrected from version 6.3 on, were they rely on the the total domestic absorption instead

of on the sum of the components (Feenstra et al., 2015). Using this more appropriated

measure and keeping the same set of leaders than before, we can therefore see in columns

(2) of Table 4 that the main conclusions remain stable.

Table 4: The impact of 61 random leaders transitions on growth

(1) (2) (3) (4)

PWT 6.1 PWT 6.3 PWT 9.0 Maddison

Timings J p-val J p-val J p-val J p-val

All (N Leaders : 61)

t 1.433** 0.0155 1.282* 0.0694 1.392** 0.0240 1.247* 0.0941

t + 1 1.464** 0.0111 1.297* 0.0610 1.493*** 0.0080 1.293* 0.0632
t + 2 1.311* 0.0520 1.298* 0.0589 1.377** 0.0272 1.321** 0.0475

Tenure = 2 Years (N Leaders : 51)

t 1.557*** 0.0066 1.433** 0.0230 1.570*** 0.0058 1.403** 0.0302
t + 1 1.564*** 0.0066 1.430** 0.0246 1.686*** 0.0017 1.475** 0.0162
t + 2 1.335* 0.0553 1.389** 0.0345 1.509** 0.0109 1.440** 0.0215

Autocrats (N Leaders : 28)

t 1.519** 0.0341 1.423 0.0651* 1.311 0.1253 1.097 0.3280

t + 1 1.592** 0.0209 1.688 0.0117** 1.428* 0.0665 1.462* 0.0517
t + 2 1.460** 0.0498 1.669 0.0133** 1.394* 0.0802 1.536** 0.0326

Democrats (N Leaders : 23)

t 1.372 0.1141 1.359 0.1084 1.747** 0.0164 1.445* 0.0735

t + 1 1.267 0.1758 1.074 0.3634 1.733** 0.0160 1.137 0.2913

t + 2 1.040 0.4081 1.140 0.2827 1.544** 0.0463 1.110 0.3190

Notes: In treatment timing t the PRE period correspond to the five-year before leader’s death and the POST one the 5 years
afterwords (the leader’s death year is not computed in any period). Leaders’ selected are the ones from whom data from PWT 6.1 is
available. Treatment timings t + 1 and t + 2 shift the POST period forward one and two years, respectively.Under the null hypothesis,
the outcome is similar before and after a leader randomly leaves office. P-values indicate the probability that the hypothesis is true.
The Wald Statistic is the test statistic described in Equation 5. The Chi-squared tests allowed for region-specific heteroskedasticity
and a region-specific AR(1) process.

However, the 6.3 version still did not solve entirely the problem of inconsistency

between different versions. The biggest dilemma comes from the extrapolation of the

parity purchasing power from the reference year to the other ones using relative countries’

inflation, as this ignores the difference of countries’ bundle of goods in the computation

of inflation. This bias is likely to be higher if the year under consideration is far away

from the reference one. Yet, according to Bolt et al. (2018) “shifts in the bundles of

products cannot fully account for these differences, leaving measurement error of some

sort as the main (though not very informative) explanation”. Thus, when using a single
year benchmark, the same error is carried through all the data set, and more biased data

make more difficult to detect the significance of a certain effect. In more recent releases

of the Penn World Table (beginning in version 8.0) and of the Maddison Project dataset

they introduced a multiple ICP benchmark approach in order to mitigate this issue. The

methodology as explained by Feenstra et al. (2015) is as follow: “For each country, we

</div>
<span class='text_page_counter'>(16)</span><div class='page_container' data-page=16>

prices for final goods interpolated using the corresponding price trends from countries

national accounts data; and for years before the first or after the last benchmark for each

country the prices of final goods are extrapolated using national account data.” Hence,

when a new benchmark is available previous growth rates will not change unless the

nominal GDP from the national account is corrected.

Using this methodology and keeping the same leaders sample for a proper comparison

we see in columns (3) and (4) of Table 4 that results change: the impact of autocratic

leaders is less significant, but more important: national leaders also matter under

democ-racies. Even more surprising, with PWT 9.0 we conclude that the impact is overall

stronger in democracies. When redoing the analysis with the Maddison Project data, I
confirm the existence of an effect under democracies though it is not lasting. However,

as presidential terms are likely to be shorter than five years, the treatment timing t + 1

and t + 2 will probably contain another transition which would affect the results.

Once detecting the bias induced by the growth data used, I explore new leaders’

natural death while in power available in more recent data sets.4 _{Results are presented in}

Table 5. When using all the leaders included in PWT 6.3 treatment timing t is no longer

significant, which could mean that the effect of a leader takes time to materialize into

economic growth rates. Unlike previous results, with this dataset we can also detect the

effect of political leaders under democracies. Using the Maddison data set (as in PWT
9.0) I found that leaders in democracies matter more than in autocracies for economic

growth. Yet, when restricting the Maddison Project to the period 1950 - 2000 in order

to be comparable to the results using PWT 6.3 and with the ones of the previous table,

results change drastically and the J statistic is no longer significant in any case. The

difference could be either due to the different countries included or simply to some leaders’

idiosyncratic characteristics. It is also necessary to recall that not having a significant J

statistic does not necessarily imply that, in average, leaders do not matter. It can be due

to the fact that the countries of leaders we included had a very volatile growth rate and

we therefore did not detect variations on growth pattern or that the qualities of leaders
from those countries are highly correlated.

What we can learn from this analysis is that a leader’s transition can, in fact, have an

effect on growth, but the context in which it may occur is still undefined as changes in

the sample lead to different conclusions. Contrary to Jones and Olken (2005), I find that

it is not only a matter of political constraints, as leaders can also matter in democratic

regimes. The only robust conclusion is that leaders who stayed longer than two years in

power appear to have a stronger impact on growth (either positive or negative).

One of the concerns that may arise with this methodology is that leaders natural

deaths while in power might not be representative of the overall rulers’ population as
it will be discussed in Section 6. Furthermore, by using those transitions we are not

</div>
<span class='text_page_counter'>(17)</span><div class='page_container' data-page=17>

Table 5: The impact of greater random leaders transitions on growth

(1) (2) (3)

PWT 6.3 Maddison Maddison (1950-2000)

Timings J p-val J p-val J p-val

All (N Leaders : 77) (N Leaders : 131) (N Leaders: 81)

t 1.038 0.3871 1.310*** 0.0099 0.763 0.9443

t + 1 1.180 0.1332 1.126 0.1524 0.714 0.9761

t + 2 1.386 0.0136 1.287** 0.0143 0.798 0.9103

Tenure 2 Years (N Leaders : 67) (N Leaders : 110) (N Leaders : 81)

t 1.150 0.1882 1.497*** 0.0006 0.853 0.8057

t + 1 1.301* 0.0506 1.152 0.1330 0.800 0.8857

t + 2 1.516*** 0.0041 1.446*** 0.0016 0.850 0.8106

Autocrats (N Leaders : 42) (N Leaders : 67) (N Leaders : 50)

t 1.005 0.4612 1.053 0.3585 0.628 0.9816

t + 1 1.256 0.1238 0.745 0.9444 0.668 0.9659

t + 2 1.667*** 0.0043 1.286* 0.0561 0.866 0.7366

Democrats (N Leaders : 28) (N Leaders : 45) (N Leaders : 24)

t 1.403* 0.0831 1.834*** 0.0006 1.192 0.2389

t + 1 1.340 0.1114 1.535** 0.0121 0.895 0.6096

t + 2 1.309 0.1306 2.108*** 0.0000 0.753 0.7996

Notes: In treatment timing t the PRE period correspond to the five-year before leader’s death and the POST one
the 5 years afterwords (the leader’s death year is not computed in any period). Treatment timings t + 1 and t + 2
shift the POST period forward one and two years, respectively.Under the null hypothesis, the outcome is similar

before and after a leader randomly leaves office. P-values indicate the probability that the hypothesis is true.
The Wald Statistic is the test statistic described in Equation 5. The Chi-squared tests allowed for region-specific
heteroskedasticity and a region-specific AR(1) process.

distinguishing between the effect of the disturbance on institutions caused by the leader’s

death and the effect of the leader’s identity (even though this issue is partially mitigated

by excluding the transition year on the Wald Test). In addition, due to the limited

number of cases it is difficult to analyze how different variables affect the impact that

rulers have on growth and on its direction. Indeed, in order to exploit larger cases another

strategy is needed.

5.2

Exploring additional exogenous transitions

By combining leaders’ data with the national constitutions (Ginsburg et al., 2010),

the Cursus Honorum dataset (Baturo, 2016) and the Database of Political Institutions
(Beck et al., 2001), we can detect further transitions in which the leader’s identity was

not likely to be determined by previous growth rates. More precisely, I include in this

section those ones in which, by Constitution, the incumbent leader could not run for the

</div>
<span class='text_page_counter'>(18)</span><div class='page_container' data-page=18>

take power by a legislative constraint. One of the limits of this approach is that electors

preferences may be partially based on the political party. So, the candidate that belongs

to the same party of the incumbent leader may be judged by his predecessor’s economic

performance.

I also include, when data is available, transitions where the leader enters in office after

winning elections by a small margin of victory (less than 55% of the votes) considering

that his victory was random in the sense that his opponent had almost the same chances

to be elected. Finally, I add transitions in which the incoming leader was an interim one,

he assumed power by a royal succession, a constitutional one or he was selected by an

elite. The main assumption is that in those cases the motivations of changing the leader
were not based on economic conditions. I eliminate cases in which, even though one of the

previous conditions held, the leader who exited power did it through an irregular manner

(such as coup, revolution or murder). All those cases are represented in the following

selection equation.

Θj = 1

(a) if lij−1 died by natural death

(b) if lij−1 could not run for elections due to a constitutional term limit and

lij−1 exited in a regular manner

(c) if lij won the elections by less than 55% of the votes and lij−1 exited in a

regular manner

(d) if lij assumed through royal succession, constitutional succession, elected

by an elite or as an interim leader and lij−1 exited in a regular manner

Considering the cases for which growth data is available, 63% of the selected

transi-tions correspond to term limits, 28% are the one treated in the previous section and the

rest correspond to special entry types (cf. Table 6).

Table 6: Leaders’ Selected Transitions

Type of transition N of Transitions

Leader died of natural causes while in power 144

Constitutional term limit 369

Close elections 21

Selected entries types 436

Source: Baturo (2016), Ginsburg et al. (2010), Goemans et al. (2009) and Beck et al. (2001)

I first reproduce the same Wald test as before and confirm that leaders matter for
growth both in autocracies and democracies (cf. Table 7). Yet, it is not longer true

that tenure reinforces the leader’s impact. Moreover, the volatility of growth under those

transitions are more prominent in democratic regimes, as found in the last tests of the

</div>
<span class='text_page_counter'>(19)</span><div class='page_container' data-page=19>

Table 7: Extended exogenous leaders’ transitions and economic growth

J p-val

Timings All (N Leaders : 438)

t 1.923*** 0.0000

t + 1 1.967*** 0.0000

t + 2 1.749*** 0.0000

Tenure >= 2 Years (N Leaders : 329)

t 1.586*** 0.0000

t + 1 1.581*** 0.0000

t + 2 1.775*** 0.0000

Autocrats (N Leaders : 134)

t 1.321** 0.0101

t + 1 1.490*** 0.0002

t + 2 1.357*** 0.0041

Democrats (N Leaders : 207)

t 2.173*** 0.0000

t + 1 2.402*** 0.0000

t + 2 2.158*** 0.0000

Notes: In treatment timing t the PRE period correspond to
the three-year last years of leader’s j-1 term and the POST
one to the three first years of leader j if Θ = 1. If one of
the leaders stayed only one or two years, only those years
were took into account. The transitional year is assigned
to the leader who stayed longer in power. Treatment
tim-ings t + 1 and t + 2 shift the POST period forward one and
two years, respectively. Under the null hypothesis, growth is
similar before and after a leader leaves office. P-values
indi-cate the probability that the hypothesis is true. The Wald
Statistic is the test statistic described in Equation 5. The
Chi-squared tests allowed for region-specific
heteroskedastic-ity and a region-specific AR(1) process.

more likely to have different characteristics and ideologies than in autocracies where the
pool of potential leaders tends to be reduced to a single political party.

Having a higher number of cases allows for controlling for further variables. Hence,

in the previous approach even if we can run Wald tests for different categories of a

certain variable (as done for autocracies or long tenure), it is difficult when it comes

to continuous variables and also when it is necessary to interact several characteristics

simultaneously. For this reason, I focus now on the differences of growth rates between

subsequent leaders in those specific transitions highlighted before. To continue with the

analysis of the magnitude of leaders effects, I first regress the absolute value of this

</div>
<span class='text_page_counter'>(20)</span><div class='page_container' data-page=20>

as follows:

|∆growth\j−1,j| = α + β∆Xj,j−1+ φ1Yj+ φ2∆Yj−1+ λZj+  if Θj = 1 (9)

where ∆growth\j−1,j = is the absolute value of the average growth difference between

the growth rates in the first three years of leader j (or less if the leader stayed one or two

years) and the previous three (or less) years of his predecessor; ∆Xj,j−1is a vector of the

differences of individual leaders or country’s continuous characteristics; Yj and Yj−1 are

vectors of leader or country dummy’s variable corresponding respectively to j’s and j-1’s

term and Zj are control variables such as the initial level of GDP or the country’s level of

development.5 I also run an alternative regression where I smooth growth rates following
Equations 7 and 8.

Results presented in Table 8 show that economies from lower income countries tend

to be more sensitive to political transitions. In fact, in those countries the growth rates

tend to vary around 10% percent more than in high income countries after a new leader

takes power. Potentially, it can be explained by the correlation with the weakness of

political and economic institutions, the regime instability, the level of corruption and

so on. In fact, there is also a positive effect when the level of democracy decreases by

one point according to the Polity IV score (re-scaled from 0 to 10). This would mean

that when controlling for other variables, the statement that growth variations were more
pronounced under democracies does not hold anymore. Hence, the relationship between

the level of democracy and the effect of the leader is not likely to be linear as the coefficient

for the autocratic dummy is negative.

In the selected transitions where the exiting ruler had a possibility to run for reelection,

the variation of the growth rate between consecutive leaders tends to be around 3 to 5

% higher. Yet, I still do not analyze in which direction. This supports the idea that the

possibility of reelection creates incentives to manipulate economic cycles.

The relationship between tenure and the impact of a leader was not clear on the

pre-vious Wald tests. From the first column of Table 8 we infer that under transitions where
a leader who stayed more than two years in power leaves the office the fluctuation of

growth tends to be lower, contradicting Jones and Olken (2005) conclusions. However,

when controlling for interaction variables the sign change although it is not longer

signif-icant. Regarding other individual characteristics, they do not seem to play a role on the

magnitude of a leader’s impact. In fact, only when a politician leaves office, the growth

shift tends to be higher, but this it is only true under the third regression and with an α

of 0.1.

In order to analyze the performance of the leaders, I focus on the nominal values

of growth variations between consecutive chiefs of state of the selected transitions (cf.

</div>
<span class='text_page_counter'>(21)</span><div class='page_container' data-page=21>

Table 8: When do leaders matter more?

(1) (2) (3) (4)

VARIABLES |∆ growth | |∆ growth smooth | |∆ growth | |∆ growth smooth |

Initial GDP 0.000108 2.43e-05 -0.000378 -0.000481

(0.000582) (0.000485) (0.000610) (0.000504)
Low-income 0.105*** 0.0921*** 0.130*** 0.114***
(0.0294) (0.0245) (0.0336) (0.0278)
Lower-middle income 0.0121 0.00518 0.0167 0.00936
(0.0145) (0.0120) (0.0149) (0.0123)
Upper-middle income 0.0154 0.0101 0.0152 0.0115

(0.0120) (0.00998) (0.0125) (0.0103)
Polity IV -0.0134*** -0.0109*** -0.0131*** -0.0106***

(0.0234) (0.0195) (0.0238) (0.0197)
Reelectionj 0.00153 -0.00207 0.0125 0.00547

(0.0127) (0.0107) (0.0168) (0.0142)
Reelectionj−1 0.0338*** 0.0309*** 0.0549** 0.0442**

(0.0128) (0.0108) (0.0218) (0.0180)
Reelectionj*

Reelectionj−1

-0.0351 -0.0257
(0.0268) (0.0224)
Autocraticj -0.0438*** -0.0374*** -0.0391* -0.0340**

(0.0167) (0.0139) (0.0207) (0.0171)
Autocraticj−1 -0.0244** -0.0170* -0.0190 -0.0111

(0.0115) (0.00954) (0.0161) (0.0133)
Autocraticj*

Autocraticj−1

-0.00719 -0.00604
(0.0229) (0.0189)
∆ Years of exp. 0.000221 0.000275 0.000224 0.000289
(0.000274) (0.000229) (0.000274) (0.000227)
∆ Entry age -9.27e-05 -0.000106 -0.000116 -0.000127
(0.000303) (0.000255) (0.000303) (0.000253)
Tenure >= 2j -0.0139 -0.00797 -0.00987 0.0113

(0.00894) (0.00749) (0.0181) (0.0150)
Tenure >= 2j−1 -0.0169* -0.00564 -0.0111 0.0152

(0.00939) (0.00789) (0.0180) (0.0149)
Tenure >= 2j*

Tenure >= 2j−1

-0.000571 -0.0216
(0.0211) (0.0175)
Univ degreej 0.00741 0.00553 0.0235 0.0113

(0.0116) (0.00976) (0.0214) (0.0177)
Univ degreej−1 0.00691 0.00893 0.0256 0.0160

(0.0108) (0.00899) (0.0216) (0.0179)
Univ degreej*

Univ degreej−1

-0.0205 -0.00517
(0.0244) (0.0202)
Politicianj 0.00961 0.00674 0.0116 0.00850

(0.0138) (0.0114) (0.0138) (0.0114)
Politicianj−1 0.0183 0.0102 0.0230* 0.0148

(0.0118) (0.00987) (0.0118) (0.00977)

Lawj 0.00378 0.00267 0.00228 0.000266

(0.00764) (0.00636) (0.00773) (0.00639)
Lawj−1 -0.00105 0.000505 -0.00176 -0.000493

(0.00798) (0.00664) (0.00800) (0.00661)
Military careerj -0.00284 -0.00335 -0.00775 -0.00796

(0.0111) (0.00923) (0.0111) (0.00921)
Military careerj−1 -0.00950 -0.0115 -0.00927 -0.0120

(0.00989) (0.00828) (0.00987) (0.00819)
Constant 0.114*** 0.0913*** 0.0849** 0.0603*
(0.0349) (0.0291) (0.0404) (0.0335)

Region Fixed-Effects No No Yes Yes

Observations 400 398 400 398

R-squared 0.185 0.177 0.224 0.229

Adj R-squared 0.139 0.131 0.157 0.161

Standard errors in parentheses

*** p<0.01, ** p<0.05, * p<0.1

Equation 10). Results are presented in Table 9. The independent variables are the same

as before, except in this case I include the variation of the level of democracy rather than

its level. The intuition is that while the current level of constraints on the executive is

important to understand the magnitude of leaders’ effects; whether a leader performs

</div>
<span class='text_page_counter'>(22)</span><div class='page_container' data-page=22>

Table 9: Who perform better?

(1) (2) (3) (4)

VARIABLES ∆ growth ∆ growth smooth ∆ growth ∆ growth smooth

Initial GDP -0.00116* -0.00127** -0.00163** -0.00166***

(0.000697) (0.000580) (0.000732) (0.000610)

Low-income 0.148*** 0.117*** 0.149*** 0.122***

(0.0351) (0.0292) (0.0402) (0.0335)

Lower-middle income -0.00814 -0.0111 -0.00209 -0.00601

(0.0171) (0.0142) (0.0176) (0.0147)

Upper-middle income 0.00171 -0.00103 0.00174 0.000155

(0.0142) (0.0118) (0.0148) (0.0124)

∆ Polity IV -0.0134 -0.0101 -0.0162 -0.0121

(0.0406) (0.0338) (0.0409) (0.0342)

Reelectionj -0.00862 -0.0116 0.0162 0.00744

(0.0152) (0.0128) (0.0202) (0.0173)

Reelectionj−1 0.0239 0.0253* 0.0593** 0.0511**

(0.0154) (0.0130) (0.0262) (0.0219)

Reelectionj*
Reelectionj−1

-0.0716** -0.0564**

(0.0323) (0.0272)

Autocraticj 0.0311 0.0240 0.0414 0.0315

(0.0233) (0.0194) (0.0266) (0.0222)

Autocraticj−1 -0.0251 -0.0194 -0.0223 -0.0174

(0.0230) (0.0192) (0.0267) (0.0223)

Autocraticj*
Autocraticj−1

-0.0155 -0.0110

(0.0275) (0.0229)

∆ Years of exp. 0.000910*** 0.000819*** 0.000929*** 0.000835***

(0.000329) (0.000274) (0.000330) (0.000276)

∆ Entry age -0.000804** -0.000711** -0.000834** -0.000752**

(0.000364) (0.000306) (0.000364) (0.000307)

Tenure >= 2j 0.00641 0.00293 -0.0176 -0.00722

(0.0107) (0.00899) (0.0217) (0.0182)

Tenure >= 2j−1 -0.00719 0.00358 -0.0267 -0.00258

(0.0113) (0.00950) (0.0216) (0.0181)

Tenure >= 2j*
Tenure >= 2j−1

0.0340 0.0148

(0.0254) (0.0213)

Univ degreej 0.0106 0.00711 0.0154 0.00828

(0.0139) (0.0117) (0.0258) (0.0215)

Univ degreej−1 -0.00853 -0.00508 -0.000556 -0.00131

(0.0129) (0.0108) (0.0261) (0.0218)

Univ degreej*
Univ degreej−1

-0.00586 -0.000255

(0.0294) (0.0246)

Politicianj 0.00438 0.00770 0.00669 0.00950

(0.0165) (0.0137) (0.0166) (0.0138)

Politicianj−1 0.0190 0.00718 0.0247* 0.0120

(0.0143) (0.0119) (0.0143) (0.0119)

Lawj 0.00159 -0.000785 0.00347 0.000246

(0.00917) (0.00763) (0.00931) (0.00777)

Lawj−1 -0.0192** -0.0153* -0.0191** -0.0156*

(0.00955) (0.00794) (0.00959) (0.00800)

Military careerj -0.00699 -0.00668 -0.0127 -0.0112

(0.0133) (0.0111) (0.0134) (0.0112)

Military careerj−1 -0.0189 -0.0186* -0.0162 -0.0166*

(0.0119) (0.00992) (0.0119) (0.00994)

Constant -0.00714 9.48e-05 -0.00757 -0.00555

(0.0341) (0.0284) (0.0408) (0.0340)

Region Fixed-Effects No No Yes Yes

Observations 400 398 400 398

R-squared 0.132 0.137 0.168 0.169

Adj R-squared 0.083 0.090 0.096 0.096

</div>
<span class='text_page_counter'>(23)</span><div class='page_container' data-page=23>

∆growth\j−1,j = α + β∆Xj,j−1+ φ1Yj + φ2∆Yj−1+ λZj +  if Θ(lj) = 1 (10)

Among contextual variables, it turns out that leaders from low-income countries, apart

from having a stronger impact, they have a positive one. Thus, it does not mean that

they are of a better quality than the ones from developed countries. It can simply imply

that those nations growth faster. In parallel, the initial level of GDP has a negative sign

as it becomes more difficult to boost the rhythm of growth in more developed countries.

When it comes to the impact of term limits, it appears that when we switch from a regime

with a possibility to run for the immediate elections to one with a term limit, growth

rates tend to increase. However, when both leaders can rerun for the following elections
there is not such an effect on economic growth. This could mean that when a leader

has the chance to become a candidate in the following elections he sacrifices some of the

economic performance to put more emphasis in the electoral campaign, vote buying or

other social preferences.

The key feature of this analysis is the relevance of individuals traits. As in the private

sector, the previous experience in politics translate into a better relative performance

being associated with a higher growth rate. What seems contradictory is that age has the

opposite sign. Nevertheless, similar results have already been found in the managerial

literature. Walter and Scheibe (2013) review the existing literature and highlight the
robust evidence across studies suggesting that younger leaders are more willing to innovate

and take risks and also that older leaders are more likely to exhibit an inactive leadership

(meaning that they take little action within their leadership role). Besides, Chevalier and

Ellison (1999) show that younger managers earn higher rates of returns and Bertrand

and Schoar (2003) argue that “CEOs from older generations appear to be less aggressive

on average, choosing a lower level of capital expenditures, lower financial leverage, and

higher cash holdings.” In the political sphere, this positive effect can also be due to the

fact that younger leaders have a longer expected career in politics, and thus they have

more motivations to “behave well”.

Having an university degree does not seem to influence the quality of a leader, as

opposed to Besley et al. (2011) results, even though the signs of those coefficients are

consistent with the economic intuition. As the authors used the same methodology

than Jones and Olken (2005) but testing whether the Wald Tests varied according to

rulers’ educational attainment, this difference could be either explained by the new sort

of transitions included in the analysis or by the correlation between education and all

the other characteristics introduced in Equation 10. I show in Table 10 that the second

argument leads the reasoning, as when only regressinggrowth\j−1,jwith respect to leaders’

university degree I do find that when a ruler without degree is followed by a more educated

</div>
<span class='text_page_counter'>(24)</span><div class='page_container' data-page=24>

et al. (2011).

Table 10: Correlation between leaders’ university degree and increases in growth rates

(1)
∆ growth

Univ degreej 0.0195*

(0.0104)

Univ degreej−1 -0.00704

(0.00965)

Constant -0.00792

(0.0109)

Observations 530

R-squared 0.007

Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1

Regarding the three most popular careers between heads of nations, there is positive

evidence that lawyers and military rulers are associated with a positive impact on growth

rates. In fact, when a leader from those background leaves the office, the following

one tends to perform worse. In this case I did not include interaction dummies, as

it would require to interact each profession with all the other ones to have a proper

comparison. Surprisingly, having a political academic background does not lead to a

better performance. On the contrary, in the only specification where this variable is

significant it suggests that when a politician leaves the office growth rates tend to increase

(column (3)).

While the contextual variables have a strong explanatory power on a leader’s room

for manoeuvre and the leader’s background was not relevant, the opposite holds when we

analyze the nominal performance between subsequent leaders. Another final comment to

be done is with respect to the goodness of fit of the previous regressions. When comparing

Table 8 and 9 it comes to the light that the proportion explained in the first one is around

the double of the cases of nominal variations. It can suggest that is relatively easier to

assess the context in which a leader can have a stronger effect on growth. Yet, the quality

of a leader relies more on unobserved skills and thus it is more difficult to identify what

</div>
<span class='text_page_counter'>(25)</span><div class='page_container' data-page=25>

6

Robustness checks

6.1

Endogeneity concerns and external validity

The main inconvenient when analyzing the impact of politicians on growth is that

leaders’ transitions are likely to be dependent on economic conditions. This is confirmed

in column (1) of Table 11 where it is shown that growth rate is significant for predicting

a transition the following year. Even though it is still true for the selected transitions

in Section 5.2, when doing a decomposition of the different cases, growth is only a

sig-nificant predictor of term limits transitions. Nevertheless, in those cases, no matter the
performance of the incumbent leader, a new ruler has to take power by a constitutional

constraint, solving (at least partially) this endogeneity issue.

Table 11: Growth as a predictor of the selected transitions

(1) (2) (3) (4) (5) (6)

Marginal Effects All transitions Transitions Θ = 1 Term Limits Died in office Close elections Selected entries types

growtht−1 -0.380*** -0.310* -0.402* -0.249 0.366 -0.382

(0.1004) (0.1639) (0.2073) (0..2924) (0.7844) (0.5425)

Observations 9,536 7,574 2,906 6,454 657 1,241

seEform in parentheses
*** p<0.01, ** p<0.05, * p<0.1

Yet, the fact that growth decreases the year before those transitions could raise other

concerns. For instance, as it is more likely that a leader without possibility of reelection

performs worse during his last year, we could underestimate his overall performance by

only considering his final period. In the same line, it could harm the economic

perfor-mance at the beginning of the following leader’s term. It becomes then important to

“smooth” growth rates by different specifications which will be done in Section 6.2.

Another potential limit of this paper is the representativeness of the leaders involved
in the selected transitions, specially for Section 5.1 where the number of observations

does not allow to introduce many controls. Table 12 shows that leaders who died in office

are likely to differ in their individual characteristics. As it is intuitive, leaders who died

in office are likely to be 8 years older in their last year of their mandate than the rest

(65 years old in average vs. 57) and to have an average longer tenure. Furthermore,

even if in the sample used by Jones and Olken (2005) those leaders were not more likely

to be autocratic (cf. column (1)), it is no longer true when accounting for all leaders

available in the sample (cf. column (3)). As those characteristics are important both for

the magnitude and the direction of the ruler’s effect, using only those transitions may
bias the results. Regarding the decades, the same analysis has been done and it is also

true that leaders are less likely to die in power in recent years (from 1970 on).

When it comes to the more general approach done in Section 5.2, the problem is

different. In the transitions selected, leaders have in average the same characteristics than

</div>
<span class='text_page_counter'>(26)</span><div class='page_container' data-page=26>

Table 12: How representative are leaders from exogenous transitions?

All leaders
-Leaders who died in office

All leaders
-Leaders for which Θ = 1

(1) (2) (3)

Sample PWT 6.1 Sample Maddison

Age -8.57*** -8.28*** 0.034

(1.395) (0.970) (0.511)
Tenure -4.42*** -4.40*** 0.317

(0.883) (0.695) (0.320)

Autocrat -0.04 -0.072* -0.038

(0.063) (0.0425) (0.0234)
Country’s characteristics

Log Real GDP per capita -0.032 0.008 0.218***
(0.128) (0.096) (0.052)
Eastern Europe -0.005 -0.035 0.021*
(0.025) (0.025) (0.012)
Latin America & the Caribbean 0.189 0.020 -0.4645***

(0.053) (0.036) (0.022)
North Africa & the Middle East -0.049 -0.026 -0.073

(0.036) (0.027) (0.015)
Sub-Saharan Africa 0.086 0.055 0.065***

(0.057) (0.035) (0.015)
Western Europe and North America 0.038 0.034 0.394***

(0.051) (0.038) (0.024)

Asia -0.090** -0.048 0.056***

(0.042) (0.029) (0.015)

Source: Author’s calculation based on Heston et al. (2002); Bolt et al. (2018)

Leaders from Latin America and the Caribbean are largely over-represented and the

opposite occurs for leaders from Western Europe and North America. Hence, the obtained

results with the transitions where Θj = 1 for the magnitude of leaders’ effect might be

overestimated for developed countries. However, if we assume that in a given context the

relative quality of a leader only depends on his individual characteristics, the directional
results would remain stable. Regarding the decades, no difference in those proportions

appears to be significant.

6.2

Alternative growth specification

As stated before, when accounting for the effects of political cycles on leaders’ identity,

it is important to consider the delay between political decisions and growth realizations.

Besides, the first or the last years of each political term may reflect different motivations

and not be representative of the mean leaders’ performance. In the previous section, I

smooth growth rates according to Equations 7 and 8. Here I test what happens with

alternative specifications. First, I assume a more extreme case, where the last and first

</div>
<span class='text_page_counter'>(27)</span><div class='page_container' data-page=27>

when I calculate the growth rate that I will name smooth 2, I replace:

GDPt= 0.2GDP gt+ 0.8GDP gt−1 if t = out year

GDPt= 0.2GDPt+ 0.8GDPt+1if t = in year

On the other side, I also consider an intermediate approach (that I will label smooth

3 ) where the GDP of transitional years account for a half of this year.

GDPt= 0.5GDP gt+ 0.5GDP gt−1 if t = out year

GDPt= 0.5GDPt+ 0.5GDPt+1if t = in year

Another alternative is to compare the differences between the first three years of each
leader’s term (or the last three ones) to check if there are differences in motivation between

the beginning and the end of each political cycle. Hence, in those cases transitions that

involve terms with longer tenures might bias results as the period of comparison between

two subsequent leaders will be more distant and time effects may play a role. Furthermore,

when comparing only the last years, results have to be interpreted carefully as the last

period of the second leader can be related to atypical realizations of growth.

Results are displayed in Table 13 and Table 14 for the absolute values of differences

in growth rates for subsequent leaders of the selected transitions. In the first Table,

without introducing the interaction dummies and the regional fixed-effects, results from
previous section (re-posted in the first two rows) appear to be overall robust in term of

sign and magnitude. Still, new variables become significant. For instance, when giving

less weight to the transitional years in columns (3) and (4), it appears that when a leader

with a military career leaves the office the shift in growth rate is weaker than under

other exogenous transitions. In those cases, and also in the one where we only compare

the last years of a leader, the same is valid when the exiting ruler holds an university

degree. Yet, we have to be careful as in those regressions the interaction variables become

more important than before. When controlling for those in Table 14 some signs change.

Within transitions where one of the leaders had a longer tenure and the other did not, the
variation in growth rates tend to be higher (confirming that those have a higher impact

on growth pattern). But when both leaders remained in power more than two years, the

effect is compensated and it tends to zero. With respect to the role of the university

degree, in column (3) we can see that when both leaders have a university education

the variability of growth tends to be higher. The same is true when we only compare

between leaders’ last term as the sum of both having a degree and the interaction term is

</div>
<span class='text_page_counter'>(28)</span><div class='page_container' data-page=28>

Table 13: When do leaders matter more? - Alternative specifications

(1) (2) (3) (4) (5) (6)

VARIABLES |∆ growth | |∆ growth smooth | |∆ growth | |∆ growth smooth 3| |∆ growth first 3 years | |∆ growth last 3 years|

Initial GDP 0.000108 2.43e-05 2.93e-05 -5.34e-05 0.000699 0.000980*
(0.000582) (0.000485) (0.000485) (0.000425) (0.000918) (0.000501)
Low-income 0.105*** 0.0921*** 0.128*** 0.0920*** 0.112*** 0.121***
(0.0294) (0.0245) (0.0245) (0.0215) (0.0331) (0.0253)
Lower-middle income 0.0121 0.00518 0.000582 0.000602 0.0152 0.0249**

(0.0145) (0.0120) (0.0120) (0.0105) (0.0178) (0.0124)
Upper-middle income 0.0154 0.0101 0.00356 0.00488 0.0136 0.0204**

(0.0120) (0.00998) (0.00999) (0.00875) (0.0142) (0.0103)
Polity IV -0.0134*** -0.0109*** -0.0089*** -0.0088*** -0.0112*** -0.0116***

(0.0234) (0.0195) (0.0195) (0.0171) (0.0252) (0.0201)
Reelectionj 0.00153 -0.00207 -0.00753 -0.00549 -0.000198 0.0200*

(0.0127) (0.0107) (0.0107) (0.00939) (0.0137) (0.0109)
Reelectionj−1 0.0338*** 0.0309*** 0.0303*** 0.0276*** 0.0383*** 0.00699

(0.0128) (0.0108) (0.0108) (0.00949) (0.0138) (0.0110)
Autocraticj -0.0438*** -0.0374*** -0.0230* -0.0285** -0.0328* -0.0460***

(0.0167) (0.0139) (0.0139) (0.0122) (0.0182) (0.0143)

Autocraticj−1 -0.0244** -0.0170* -0.0199** -0.0142* -0.0208* -0.0154

(0.0115) (0.00954) (0.00955) (0.00837) (0.0125) (0.00985)
∆ Years of exp. 0.000221 0.000275 4.65e-05 0.000274 -8.30e-05 -3.34e-05
(0.000274) (0.000229) (0.000229) (0.000200) (0.000292) (0.000236)
∆ Entry age -9.27e-05 -0.000106 3.18e-06 -0.000136 0.000195 0.000250

(0.000303) (0.000255) (0.000256) (0.000224) (0.000326) (0.000261)
Tenure >= 2j -0.0139 -0.00797 -0.0152** -0.00898 -0.0178* -0.0147*

(0.00894) (0.00749) (0.00750) (0.00657) (0.00960) (0.00768)
Tenure >= 2j−1 -0.0169* -0.00564 -0.0185** -0.00145 -0.0167* -0.0148*

(0.00939) (0.00789) (0.00790) (0.00692) (0.00997) (0.00807)
Univ degreej 0.00741 0.00553 -0.000325 0.00221 0.00541 -0.00268

(0.0116) (0.00976) (0.00977) (0.00856) (0.0126) (0.00996)
Univ degreej−1 0.00691 0.00893 0.0245*** 0.0133* 0.0159 0.0154*

(0.0108) (0.00899) (0.00900) (0.00788) (0.0118) (0.00927)

Politicianj 0.00961 0.00674 0.00529 0.00520 0.00973 0.00812

(0.0138) (0.0114) (0.0115) (0.0100) (0.0148) (0.0118)

Politicianj−1 0.0183 0.0102 0.000878 0.00211 0.0111 0.0147

(0.0118) (0.00987) (0.00988) (0.00865) (0.0127) (0.0102)

Lawj 0.00378 0.00267 0.00730 0.00418 -0.00143 -0.000636

(0.00764) (0.00636) (0.00637) (0.00558) (0.00819) (0.00657)

Lawj−1 -0.00105 0.000505 -0.00157 0.00111 -0.00229 0.00146

(0.00798) (0.00664) (0.00664) (0.00582) (0.00851) (0.00685)
Military careerj -0.00284 -0.00335 -0.00579 -0.00207 0.00148 -0.00248

(0.0111) (0.00923) (0.00924) (0.00809) (0.0119) (0.00952)
Military careerj−1 -0.00950 -0.0115 -0.0153* -0.0151** -0.000892 -0.00781

(0.00989) (0.00828) (0.00829) (0.00726) (0.0107) (0.00850)
Constant 0.114*** 0.0913*** 0.103*** 0.0836*** 0.0949** 0.0940***

(0.0349) (0.0291) (0.0291) (0.0255) (0.0389) (0.0300)

Region Fixed-Effects No No No No No No

Observations 400 398 398 398 389 400

R-squared 0.185 0.177 0.209 0.180 0.158 0.196

Adj R-squared 0.134 0.131 0.165 0.134 0.110 0.152

Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1

on economic outcomes (either in a positive or negative way).

When it comes to the directional results of these specifications, the outcomes are

presented in Table 15. Qualitatively, results previously found are robust. Again, new

mechanisms become significant when testing for different specifications. Although for low

income nations there is a positive growth trend with respect to high income economies,

the opposite is true for lower-middle and upper-middle income countries.

Another new feature is the positive effect of a change of regime from democracy

to autocracy when only comparing the 3 first or last years. As controversial it may

sound, there is mitigated evidence in the literature about the link between democracy
and growth. In the empirical side, my results are in line with the evidence provided by

</div>
<span class='text_page_counter'>(29)</span><div class='page_container' data-page=29>

Table 14: When do leaders matter more? - Alternative specifications II

(1) (2) (3) (4) (5) (6)
VARIABLES |∆ growth | |∆ |growth smooth |∆ growth smooth 2 | |∆ growth smooth 3 | |∆ growth first 3 years | |∆growth last 3 years |
Initial GDP -0.000378 -0.000481 -0.000552 -0.000583 -0.000155 0.000699

(0.000610) (0.000504) (0.000494) (0.000437) (0.000969) (0.000523)
Low-income 0.130*** 0.114*** 0.153*** 0.112*** 0.128*** 0.144***
(0.0336) (0.0278) (0.0273) (0.0241) (0.0371) (0.0288)
Lower-middle income 0.0167 0.00936 0.00301 0.00357 0.0154 0.0295**

(0.0149) (0.0123) (0.0121) (0.0106) (0.0183) (0.0127)
Upper-middle income 0.0152 0.0115 0.00789 0.00753 0.0105 0.0194*
(0.0125) (0.0103) (0.0101) (0.00894) (0.0149) (0.0107)

Polity IV -0.0131*** -0.0106*** -0.0082*** -0.0084*** -0.0101*** -0.0110***

(0.0238) (0.0197) (0.0193) (0.0171) (0.0256) (0.0204)
Reelectionj 0.0125 0.00547 0.00261 0.00194 0.0172 0.0434***

(0.0168) (0.0142) (0.0140) (0.0123) (0.0183) (0.0144)
Reelectionj−1 0.0549** 0.0442** 0.0428** 0.0379** 0.0671*** 0.0444**

(0.0218) (0.0180) (0.0177) (0.0156) (0.0236) (0.0187)
Reelectionj*

Reelectionj−1

-0.0351 -0.0257 -0.0349 -0.0244 -0.0591** -0.0690***
(0.0268) (0.0224) (0.0220) (0.0194) (0.0293) (0.0230)
Autocraticj -0.0391* -0.0340** -0.0221 -0.0282* -0.0198 -0.0373**

(0.0207) (0.0171) (0.0168) (0.0148) (0.0225) (0.0178)
Autocraticj−1 -0.0190 -0.0111 -0.0157 -0.00965 -0.0153 -0.0143

(0.0161) (0.0133) (0.0131) (0.0115) (0.0172) (0.0138)
Autocraticj*

Autocraticj−1

-0.00719 -0.00604 0.000850 -0.000934 -0.0139 -0.00720
(0.0229) (0.0189) (0.0186) (0.0164) (0.0247) (0.0196)
∆ Years of exp. 0.000224 0.000289 7.04e-05 0.000297 -0.000101 -5.14e-05
(0.000274) (0.000227) (0.000223) (0.000197) (0.000292) (0.000235)
∆ Entry age -0.000116 -0.000127 -3.35e-05 -0.000151 0.000154 0.000227

(0.000303) (0.000253) (0.000248) (0.000219) (0.000325) (0.000259)
Tenure >= 2j -0.00987 0.0113 0.0441*** 0.0284** -0.00767 -0.0116

(0.0181) (0.0150) (0.0147) (0.0130) (0.0191) (0.0155)
Tenure >= 2j−1 -0.0111 0.0152 0.0424*** 0.0373*** -0.00170 -0.0107

(0.0180) (0.0149) (0.0146) (0.0129) (0.0191) (0.0154)
Tenure >= 2j*

Tenure >= 2j−1

-0.000571 -0.0216 -0.0784*** -0.0475*** -0.0125 -0.000926
(0.0211) (0.0175) (0.0172) (0.0152) (0.0225) (0.0181)
Univ degreej 0.0235 0.0113 -0.0259 -0.00841 0.00488 0.0224

(0.0214) (0.0177) (0.0174) (0.0153) (0.0239) (0.0184)
Univ degreej−1 0.0256 0.0160 -0.00159 0.00227 0.0214 0.0461**

(0.0216) (0.0179) (0.0176) (0.0155) (0.0240) (0.0185)
Univ degreej*

Univ degreej−1

-0.0205 -0.00517 0.0403** 0.0187 -0.00195 -0.0356*
(0.0244) (0.0202) (0.0198) (0.0175) (0.0269) (0.0209)
Politicianj 0.0116 0.00850 0.00696 0.00674 0.0118 0.0119

(0.0138) (0.0114) (0.0112) (0.00986) (0.0148) (0.0118)
Politicianj−1 0.0230* 0.0148 0.00569 0.00650 0.0137 0.0181*

(0.0118) (0.00977) (0.00959) (0.00847) (0.0126) (0.0101)
Lawj 0.00228 0.000266 0.00363 0.00107 -0.000978 -0.000543

(0.00773) (0.00639) (0.00627) (0.00554) (0.00822) (0.00663)
Lawj−1 -0.00176 -0.000493 -0.00326 -0.000213 -0.000510 0.00153

(0.00800) (0.00661) (0.00648) (0.00573) (0.00850) (0.00686)
Military careerj -0.00775 -0.00796 -0.0103 -0.00646 -0.00236 -0.00540

(0.0111) (0.00921) (0.00904) (0.00798) (0.0119) (0.00955)
Military careerj−1 -0.00927 -0.0120 -0.0177** -0.0167** -0.00102 -0.00773

(0.00987) (0.00819) (0.00803) (0.00709) (0.0107) (0.00845)
Constant 0.0849** 0.0603* 0.0665** 0.0535* 0.0751* 0.0574*
(0.0404) (0.0335) (0.0328) (0.0290) (0.0450) (0.0347)
Region Fixed-Effects Yes Yes Yes Yes Yes Yes
Observations 400 398 398 398 389 400
R-squared 0.224 0.229 0.289 0.249 0.207 0.240
Adj R-squared 0.157 0.161 0.226 0.184 0.136 0.173

Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1

happen the longer a political regime has been in place. Hence, it is not the regime per
se that leads to a better economic performance, but the regime shift. As they state, it

is coherent with the idea that a single leadership (or regime) change may be a necessary

condition for a nation to boost economic growth (Jones and Olken, 2005; Jong-A-Pin

and Yu, 2010). Finally, when only comparing the beginning of each leader’s term in

column (5) it is also shown that there is a negative effect when a leader with long tenure

</div>
<span class='text_page_counter'>(30)</span><div class='page_container' data-page=30>

Table 15: Who perform better? - Alternative specifications

(1) (2) (3) (4) (5) (6)

VARIABLES ∆ growth ∆ growth smooth ∆ growth ∆ growth smooth 3 ∆ growth first 3 years ∆ growth last 3 years

Initial GDP -0.00163** -0.00166*** -0.00209*** -0.00179*** -0.00372*** -0.000453
(0.000732) (0.000610) (0.000638) (0.000543) (0.00111) (0.000661)
Low-income 0.149*** 0.122*** 0.108*** 0.104*** 0.112*** 0.133***
(0.0402) (0.0335) (0.0351) (0.0298) (0.0422) (0.0363)
Lower-middle income -0.00209 -0.00601 -0.0259* -0.0143 -0.0554*** -0.00280
(0.0176) (0.0147) (0.0154) (0.0131) (0.0207) (0.0159)
Upper-middle income 0.00174 0.000155 -0.0130 -0.00530 -0.0448*** -0.000728

(0.0148) (0.0124) (0.0130) (0.0110) (0.0169) (0.0134)

∆ Polity IV -0.0162 -0.0121 0.00183 -0.00509 -0.0195 0.0321

(0.0409) (0.0342) (0.0358) (0.0304) (0.0423) (0.0369)

Reelectionj 0.0162 0.00744 -0.00755 -0.00144 0.0102 -0.0148

(0.0202) (0.0173) (0.0181) (0.0154) (0.0210) (0.0182)
Reelectionj−1 0.0593** 0.0511** 0.0600*** 0.0471** 0.108*** 0.0344

(0.0262) (0.0219) (0.0229) (0.0195) (0.0271) (0.0237)
Reelectionj*

Reelectionj−1

-0.0716** -0.0564** -0.0527* -0.0450* -0.118*** -0.0187
(0.0323) (0.0272) (0.0285) (0.0242) (0.0336) (0.0291)

Autocraticj 0.0414 0.0315 0.0328 0.0261 0.0488* 0.0554**

(0.0266) (0.0222) (0.0232) (0.0197) (0.0278) (0.0240)
Autocraticj−1 -0.0223 -0.0174 -0.0239 -0.0158 -0.0198 -0.0489**

(0.0267) (0.0223) (0.0233) (0.0199) (0.0274) (0.0241)
Autocraticj*

Autocraticj−1

-0.0155 -0.0110 -0.00817 -0.00789 -0.0253 -0.00244

(0.0275) (0.0229) (0.0240) (0.0204) (0.0283) (0.0248)
∆ Years of exp. 0.000929*** 0.000835*** 0.000651** 0.000721*** 0.000917*** 0.000855***

(0.000330) (0.000276) (0.000289) (0.000246) (0.000335) (0.000298)
∆ Entry age -0.000834** -0.000752** -0.000438 -0.000617** -0.000718* -0.000474
(0.000364) (0.000307) (0.000321) (0.000273) (0.000372) (0.000329)
Tenure >= 2j -0.0176 -0.00722 -0.0280 -0.00756 -0.0235 -0.0155

(0.0217) (0.0182) (0.0190) (0.0162) (0.0219) (0.0196)
Tenure >= 2j−1 -0.0267 -0.00258 0.0141 0.0164 -0.0502** -0.0256

(0.0216) (0.0181) (0.0189) (0.0161) (0.0219) (0.0195)
Tenure >= 2j*

Tenure >= 2j−1

0.0340 0.0148 0.0222 0.00611 0.0520** 0.0229

(0.0254) (0.0213) (0.0222) (0.0189) (0.0257) (0.0229)

Univ degreej 0.0154 0.00828 -0.0132 -0.000688 -0.0105 0.00370

(0.0258) (0.0215) (0.0225) (0.0191) (0.0274) (0.0233)
Univ degreej−1 -0.000556 -0.00131 0.00144 -0.00131 0.00295 -0.0227

(0.0261) (0.0218) (0.0228) (0.0194) (0.0276) (0.0235)
Univ degreej*

Univ degreej−1

-0.00586 -0.000255 0.0153 0.00717 0.00785 0.0205

(0.0294) (0.0246) (0.0257) (0.0219) (0.0310) (0.0265)

Politicianj 0.00669 0.00950 0.0175 0.0139 0.0134 0.00667

(0.0166) (0.0138) (0.0145) (0.0123) (0.0170) (0.0150)

Politicianj−1 0.0247* 0.0120 -1.57e-05 0.000917 0.0234 0.0187

(0.0143) (0.0119) (0.0125) (0.0106) (0.0146) (0.0129)

Lawj 0.00347 0.000246 -0.00551 -0.00318 -0.000213 -0.00273

(0.00931) (0.00777) (0.00812) (0.00691) (0.00944) (0.00840)
Lawj−1 -0.0191** -0.0156* -0.0181** -0.0141** -0.0240** -0.0231***

(0.00959) (0.00800) (0.00836) (0.00711) (0.00972) (0.00866)
Military careerj -0.0127 -0.0112 -0.00739 -0.00930 -0.0170 -0.00388

(0.0134) (0.0112) (0.0117) (0.00997) (0.0136) (0.0121)
Military careerj−1 -0.0162 -0.0166* -0.0180* -0.0172* -0.0121 -0.0178*

(0.0119) (0.00994) (0.0104) (0.00884) (0.0122) (0.0107)

Constant -0.00757 -0.00555 0.0254 0.00393 0.0592 0.00747

(0.0408) (0.0340) (0.0356) (0.0303) (0.0436) (0.0368)

Observations 400 398 398 398 389 400

R-squared 0.168 0.169 0.166 0.175 0.221 0.162

Adj R-squared 0.096 0.096 0.092 0.102 0.151 0.089

Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1

6.3

Leaders’ fixed effects

In this Section I will use an alternative approach. Recall the starting Equation:

git = vi+ ut+ λlit+ it (11)

</div>
<span class='text_page_counter'>(31)</span><div class='page_container' data-page=31>

per-formance even across exogenous transitions, is that the end dates of the leader who takes

power can be related to atypical realizations of growth if he is not himself in case (a) or

(b) of the selection Equation or if the following one is not in (c) or (d). Nevertheless,

having a larger number of observations allows to have more than 200 cases in which two

consecutive transitions are such that Θj = 1. To exploit those cases, I first estimate

Equation 11 with all the leaders in the sample. Then, I retrieve the estimated λ and

calculate the differences between rulers’ fixed-effects when both of the leaders’ identities

can be considered exogenous with respect to economic conditions (i.e when Θj = 1 and

Θj+1 = 1) as shown in Equation 12.

∆\λj−1,j = α + β∆Xj,j−1+ φ1Yj + φ2∆Yj−1+ λZj +  if Θj = 1 and Θj+ 1 = 1 (12)

One of the advantages of this approach is that we previously control in Equation 11

by the countries fixed-effects on growth path and years fixed-effects. Another difference

is that we consider the leaders’ whole term. The main inconvenient is that as I use in
Equation 12 the estimated coefficients of λ from Equation 11, we should interpret those

results with precaution due to the bias induced.

In Table 16, I analyze the context in which leaders matter through the fixed-effect

approach. The results of interest are those of column (4) where I consider the transitions

in which both of the subsequent leaders’ identities are exogenous. Yet, most of the

coefficients’ signs and significance are robust to regressions with less controls and/or to

the ones where I consider transitions in which only one of the leaders identity is exogenous

(columns (1) to (3)). As before, I do find that the magnitude of the impact of the leader

is constrained by the level of democracy. While in low income nations the volatility of
growth between political terms seems to be higher with respect to high income ones, it is

no longer true for middle-income economies. As in the previous alternative specifications,

I found that when at least one of the leaders has a university degree, the fluctuations on

growth of those political transitions tend to be more pronounced. What goes at odds

of previous results is the sign of a transition where the exiting leader is a military one.

Here, it appears that in those cases the variations on growth are likely to be higher.

Then, I consider the nominal variations in fixed-effects in Table 17. As stated in

Section 6.2, I confirm that regimes changes are associated to a better (temporary)

per-formance in terms of economic growth. Nevertheless, now moves through democracies

appear to have a strong and positive effect on the performance of a leader. This can state,
that within a specific regime, leaders who impulse reforms through inclusive institutions

manage to boost economic growth.

I retrieve again the negative link between the possibility of being in power two

</div>
<span class='text_page_counter'>(32)</span><div class='page_container' data-page=32>

Table 16: Determinants of the magnitude of leaders’ fixed-effect

Θj= 1 Θj= 1 ˆ Θj+1 = 1

(1) (2) (3) (4)

VARIABLES |∆ λ | |∆ λ | |∆ λ | |∆ λ |

Initial GDP -0.000372 -0.00111 -0.000349 -0.00116

(0.000729) (0.000743) (0.000787) (0.000776)

Polity IV -0.0137*** -0.0128*** -0.0148*** -0.0144***

(0.0293) (0.0290) (0.0353) (0.0336)

Autocraticj -0.0660*** -0.0364 -0.0315 -0.0160

(0.0208) (0.0252) (0.0288) (0.0371)

Autocraticj−1 0.00527 0.0100 -0.0375* -0.0256

(0.0142) (0.0192) (0.0220) (0.0276)

Autocraticj*
Autocraticj−1

-0.0292 -0.0211

(0.0276) (0.0433)

∆ Years of exp. -0.000515 -0.000605* -0.000704* -0.000614

(0.000344) (0.000334) (0.000404) (0.000379)

Low-income 0.0415 0.0726* 0.0109 0.0719

(0.0368) (0.0409) (0.0548) (0.0567)

Lower-middle income -0.0494*** -0.0334* -0.0616*** -0.0388*

(0.0180) (0.0180) (0.0218) (0.0212)

Upper-middle income -0.0458*** -0.0375** -0.0554*** -0.0424**

(0.0149) (0.0151) (0.0177) (0.0173)

Reelectionj 0.0351** 0.0451** -0.0250 -0.0125

(0.0159) (0.0204) (0.0239) (0.0292)

Reelectionj−1 -0.0371** -0.0265 0.00562 0.0423

(0.0161) (0.0265) (0.0243) (0.0375)

Reelectionj*
Reelectionj−1

-0.0727** -0.115**

(0.0326) (0.0493)

∆ Entry age 0.000653* 0.000524 0.000586 0.000531

(0.000376) (0.000365) (0.000468) (0.000445)

Tenure >= 2j 0.0154 0.0115 0.0242* 0.00350

(0.0112) (0.0216) (0.0143) (0.0263)

Tenure >= 2j−1 -0.00495 -0.00389 -0.00777 -0.0234

(0.0117) (0.0215) (0.0136) (0.0273)

Tenure >= 2j*
Tenure >= 2j−1

0.0127 0.0414

(0.0254) (0.0313)

Univ degreej 0.00814 0.0279 -0.00491 0.0307

(0.0143) (0.0260) (0.0177) (0.0344)

Univ degreej−1 0.0369*** 0.0630** 0.0236 0.0542*

(0.0135) (0.0263) (0.0181) (0.0321)

Univ degreej*
Univ degreej−1

-0.0256 -0.0347

(0.0296) (0.0372)

Politicianj 0.00693 0.00593 0.00738 0.00598

(0.0169) (0.0165) (0.0222) (0.0211)

Politicianj−1 0.00410 0.0106 0.0108 0.00425

(0.0148) (0.0143) (0.0192) (0.0180)

Lawj 0.0107 0.0148 0.00614 0.00950

(0.00958) (0.00941) (0.0113) (0.0108)

Lawj−1 -0.00874 -0.00770 0.00368 0.00426

(0.00999) (0.00973) (0.0114) (0.0108)

Military careerj 0.0137 0.00860 0.000774 -0.00393

(0.0138) (0.0135) (0.0163) (0.0152)

Military careerj−1 0.00512 0.00667 0.0302* 0.0266*

(0.0123) (0.0119) (0.0155) (0.0147)

Constant 0.145*** 0.111** 0.172*** 0.142**

(0.0430) (0.0485) (0.0553) (0.0632)

Region Fixed-Effects No Yes No Yes

Observations 401 401 271 271

R-squared 0.135 0.223 0.155 0.305

Adj R-squared 0.087 0.156 0.083 0.214

</div>
<span class='text_page_counter'>(33)</span><div class='page_container' data-page=33>

Table 17: Leaders’ performance through a fixed-effect approach

Θj= 1 Θj= 1 ˆ Θj+1 = 1

(1) (2) (3) (4)

VARIABLES ∆ λ ∆ λ ∆ λ ∆ λ

Initial GDP -0.000674 -0.000858 -0.000424 -0.000560

(0.000755) (0.000782) (0.000851) (0.000882)

∆ Polity IV 0.0153*** 0.0152*** 0.0285*** 0.0281***

(0.0462) (0.0458) (0.0732) (0.0737)

Autocraticj 0.0999*** 0.0817*** 0.142*** 0.135***

(0.0263) (0.0302) (0.0408) (0.0506)

Autocraticj−1 -0.105*** -0.118*** -0.140*** -0.139***

(0.0259) (0.0296) (0.0408) (0.0450)

Autocraticj*
Autocraticj−1

0.0315 0.00550

(0.0311) (0.0507)

∆ Years of exp. 0.000689* 0.000647* 0.000376 0.000384

(0.000362) (0.000360) (0.000455) (0.000452)

Low-income -0.0386 -0.0418 -0.0419 -0.0438

(0.0541) (0.0538) (0.0727) (0.0726)

Lower-middle income -0.00881 -0.00739 0.0133 0.0118

(0.0190) (0.0189) (0.0239) (0.0239)

Upper-middle income -0.0123 -0.00987 0.00799 0.00922

(0.0157) (0.0157) (0.0197) (0.0197)

Reelectionj -0.0381** -0.0560*** -0.0441 -0.0911***

(0.0183) (0.0216) (0.0284) (0.0330)

Reelectionj−1 0.0405** -0.0208 0.0706** -0.0425

(0.0189) (0.0353) (0.0298) (0.0530)

Reelectionj*
Reelectionj−1

0.0871** 0.174***

(0.0426) (0.0642)

∆ Entry age -0.000879** -0.000861** -0.000898* -0.000959*

(0.000414) (0.000413) (0.000526) (0.000522)

Tenure >= 2j 0.00265 -0.0287 0.000352 -0.0136

(0.0118) (0.0233) (0.0160) (0.0299)

Tenure >= 2j−1 0.00375 -0.0256 0.0126 -0.00144

(0.0122) (0.0233) (0.0149) (0.0316)

Tenure >= 2j*
Tenure >= 2j−1

0.0404 0.0211

(0.0273) (0.0364)

Univ degreej 0.0268* -0.0123 0.0328 -0.00573

(0.0160) (0.0311) (0.0201) (0.0409)

Univ degreej−1 -0.0425*** -0.0849*** -0.0456** -0.0793**

(0.0151) (0.0312) (0.0201) (0.0398)

Univ degreej*
Univ degreej−1

0.0514 0.0445

(0.0352) (0.0456)

Politicianj -0.0158 -0.0258 -0.0164 -0.0308

(0.0187) (0.0188) (0.0249) (0.0252)

Politicianj−1 0.00691 0.00956 -0.00164 0.000575

(0.0159) (0.0159) (0.0214) (0.0214)

Lawj 0.0109 0.0142 0.0137 0.0155

(0.01000) (0.0100) (0.0125) (0.0125)

Lawj−1 -0.00463 -0.00564 -0.0110 -0.0107

(0.0105) (0.0105) (0.0126) (0.0125)

Military careerj 0.00242 0.00397 0.00410 0.00932

(0.0146) (0.0147) (0.0179) (0.0180)

Military careerj−1 -0.0171 -0.0135 -0.0289* -0.0211

(0.0131) (0.0131) (0.0172) (0.0175)

Constant 0.0291 0.0880* 0.0111 0.0582

(0.0384) (0.0464) (0.0490) (0.0616)

Observations 359 359 257 257

R-squared 0.116 0.140 0.154 0.183

Adj R-squared 0.060 0.076 0.078 0.094

</div>
<span class='text_page_counter'>(34)</span><div class='page_container' data-page=34>

limit to one without it, are associated to a decrease in 9% on growth rates.

Regarding leaders characteristics, younger leaders are still associated with a better

performance and so do the ones with higher experience in politics (although it is only

significant in the first two regressions). What is interesting about those cases, is the

relevance that gains education when using this approach. The signs are aligned with the

economic intuition, as having a university degree is associated to a higher leader

fixed-effect. When it comes to the rulers’ profession, the signs are robust according to what

was previously affirmed although they are not longer significant.

To finish, I consider an in-between approach between leaders’ fixed-effect and the one

developed in Section 5.2. I reuse a two step procedure where the first estimated equation
is the same as the one used for Jones and Olken (2005) for empirical estimating the

previous Wald Test (Equation 13).

git = γ1P REj+ γ2P OSTj + ut+ vi+ it (13)

where git is the growth rate of country i at time t, ut and vi are respectively time

and country fixed-effects and P REj and P OSTj are dummies that are equal to 1 in the

three years before and after (respectively) leader j enters office if Θj = 1. I then estimate

Equation 14 considering the dependent variable both in absolute and nominal values.

∆γ_bj = α + β∆Xj,j−1+ φ1Yj+ φ2∆Yj−1+ λZj +  if Θj = 1 and Θj+1 = 1 (14)

where ∆γj = γb2 - γb1 of Equation 13. As for the leaders’ fixed-effect, results take

into account the country’s growth trend as well as global economic shocks. Yet, the

difference here is that instead of considering the whole leader’s term, I only focus on the
transitional years just like before and thus I do not need both Θj and Θj+1 to be equal

to one. But due to multicolinearity issues, I only have 226 transitions in which both

the PRE and POST coefficient were estimated. Results for the absolute difference are

presented in Table 19 where Columns (1) and (3) retrieve the estimations of Section 5.2

for comparison reasons. Once more, signs and coefficients are consistent with previous

findings, even though when controlling for countries and time dummies, some variables

have a stronger effect, in particular tenure and the possibility of reelection. Here, when

the exiting leader has a long tenure, the effect on growth variability is negative even

when controlling for the interaction dummies which go at odds with what was previously

found. Columns (2) to (4) suggest that when a politician leaves the office, the shift in

growth tends to be higher and the opposite occurs when it is a military leader. This could

suggest some different managing styles from leaders of different careers. Some of them

</div>
<span class='text_page_counter'>(35)</span><div class='page_container' data-page=35>

Table 18: When do leaders matter more? - Alternative approach

(1) (2) (3) (4)

VARIABLES |∆ growth | |∆γ | |∆ growth | |∆γ |

Initial GDP 0.000108 -0.00216 -0.000378 -0.00432**
(0.000582) (0.00167) (0.000610) (0.00172)
Low-income 0.105*** 0.155** 0.130*** 0.135

(0.0294) (0.0624) (0.0336) (0.0842)
Lower-middle income 0.0121 -0.0206 0.0167 -0.0183

(0.0145) (0.0271) (0.0149) (0.0273)
Upper-middle income 0.0154 -0.00940 0.0152 -0.00938
(0.0120) (0.0216) (0.0125) (0.0221)
Polity IV -0.0134*** -0.0178*** -0.0131*** -0.0119***

(0.0234) (0.0359) (0.0238) (0.0368)
Reelectionj 0.00153 -0.0212 0.0125 0.00374

(0.0127) (0.0206) (0.0168) (0.0249)
Reelectionj−1 0.0338*** 0.0731*** 0.0549** 0.172***

(0.0128) (0.0223) (0.0218) (0.0390)
Reelectionj*

Reelectionj−1

-0.0351 -0.150***
(0.0268) (0.0461)
Autocraticj -0.0438*** -0.0689*** -0.0391* -0.0367

(0.0167) (0.0239) (0.0207) (0.0285)
Autocraticj−1 -0.0244** -0.0160 -0.0190 -0.00472

(0.0115) (0.0141) (0.0161) (0.0170)
Autocraticj*

Autocraticj−1

-0.00719 -0.0179
(0.0229) (0.0275)
∆ Years of exp. 0.000221 0.000609 0.000224 0.000504
(0.000274) (0.000417) (0.000274) (0.000403)
∆ Entry age -9.27e-05 -0.000513 -0.000116 -0.000513
(0.000303) (0.000475) (0.000303) (0.000475)
Tenure >= 2j -0.0139 -0.00875 -0.00987 -0.0527

(0.00894) (0.0194) (0.0181) (0.0397)
Tenure >= 2j−1 -0.0169* -0.0376** -0.0111 -0.0819**

(0.00939) (0.0163) (0.0180) (0.0410)
Tenure >= 2j*

Tenure >= 2j−1

-0.000571 0.0595
(0.0211) (0.0443)
Univ degreej 0.00741 0.0203 0.0235 0.0391

(0.0116) (0.0200) (0.0214) (0.0358)
Univ degreej−1 0.00691 0.00549 0.0256 0.0225

(0.0108) (0.0191) (0.0216) (0.0376)
Univ degreej*

Univ degreej−1

-0.0205 -0.00312
(0.0244) (0.0422)
Politicianj 0.00961 -0.00304 0.0116 0.000746

(0.0138) (0.0223) (0.0138) (0.0216)
Politicianj−1 0.0183 0.0356* 0.0230* 0.0468**

(0.0118) (0.0202) (0.0118) (0.0197)
Lawj 0.00378 0.00211 0.00228 0.00847

(0.00764) (0.0113) (0.00773) (0.0112)
Lawj−1 -0.00105 -0.00706 -0.00176 -0.0108

(0.00798) (0.0111) (0.00800) (0.0108)
Military careerj -0.00284 -0.0236 -0.00775 -0.0256

(0.0111) (0.0174) (0.0111) (0.0169)
Military careerj−1 -0.00950 -0.0361** -0.00927 -0.0224

(0.00989) (0.0167) (0.00987) (0.0165)
Constant 0.114*** 0.199*** 0.0849** 0.152**
(0.0349) (0.0585) (0.0404) (0.0677)
Region Fixed-Effects No No Yes

Observations 400 226 400 226

R-squared 0.185 0.303 0.224 0.392

Adj R-squared 0.139 0.231 0.157 0.294
Standard errors in parentheses

*** p<0.01, ** p<0.05, * p<0.1

or passive way leading to growth fluctuations.

Finally, Table 15 presents the results of the subsequent’s leaders’ performance with

this approach. Results are overall robust in terms of sign and significance with some

exceptions. The entry age seems no longer to play a role, even though the sign is still

negative and in Column (4) the p-value is 0.12. Tenure has also a contradictory sign

</div>
<span class='text_page_counter'>(36)</span><div class='page_container' data-page=36>

Table 19: Who perform better? - Alternative approach

(1) (2) (3) (4)

VARIABLES ∆ growth ∆γ ∆ growth ∆γ

Initial GDP -0.00116* -0.00458** -0.00163** -0.00783***
(0.000697) (0.00205) (0.000732) (0.00195)
Low-income 0.148*** 0.199*** 0.149*** 0.131

(0.0351) (0.0764) (0.0402) (0.0950)
Lower-middle income -0.00814 -0.0301 -0.00209 -0.0432
(0.0171) (0.0330) (0.0176) (0.0304)
Upper-middle income 0.00171 -0.00878 0.00174 -0.0227
(0.0142) (0.0265) (0.0148) (0.0249)
∆ Polity IV -0.0013 0.00028 -0.0016 0.0011

(0.0406) (0.0512) (0.0409) (0.0477)
Reelectionj -0.00862 -0.0170 0.0162 0.0335

(0.0152) (0.0254) (0.0202) (0.0283)
Reelectionj−1 0.0239 0.0798*** 0.0593** 0.250***

(0.0154) (0.0274) (0.0262) (0.0432)
Reelectionj*

Reelectionj−1

-0.0716** -0.259***
(0.0323) (0.0509)
Autocraticj 0.0311 0.0265 0.0414 0.0365

(0.0233) (0.0236) (0.0266) (0.0301)
Autocraticj−1 -0.0251 -0.0160 -0.0223 -0.00158

(0.0230) (0.0205) (0.0267) (0.0203)
Autocraticj*

Autocraticj−1

-0.0155 -0.0336
(0.0275) (0.0323)
∆ Years of exp. 0.000910*** 0.000824* 0.000929*** 0.000569
(0.000329) (0.000510) (0.000330) (0.000453)
∆ Entry age -0.000804** -0.000566 -0.000834** -0.000818

(0.000364) (0.000584) (0.000364) (0.000522)
Tenure >= 2j 0.00641 0.0146 -0.0176 -0.118***

(0.0107) (0.0237) (0.0217) (0.0444)
Tenure >= 2j−1 -0.00719 -0.0281 -0.0267 -0.167***

(0.0113) (0.0208) (0.0216) (0.0463)
Tenure >= 2j*

Tenure >= 2j−1

0.0340 0.178***
(0.0254) (0.0495)
Univ degreej 0.0106 -0.00759 0.0154 0.0123

(0.0139) (0.0246) (0.0258) (0.0403)

Univ degreej−1 -0.00853 0.00527 -0.000556 0.0203

(0.0129) (0.0233) (0.0261) (0.0423)
Univ degreej*

Univ degreej−1

-0.00586 0.00852
(0.0294) (0.0475)
Politicianj 0.00438 0.00915 0.00669 0.0210

(0.0165) (0.0274) (0.0166) (0.0245)
Politicianj−1 0.0190 0.0293 0.0247* 0.0429*

(0.0143) (0.0246) (0.0143) (0.0219)

Lawj 0.00159 0.00148 0.00347 0.00937

(0.00917) (0.0139) (0.00931) (0.0127)
Lawj−1 -0.0192** -0.0121 -0.0191** -0.0213*

(0.00955) (0.0136) (0.00959) (0.0122)
Military careerj -0.00699 -0.00463 -0.0127 -0.0193

(0.0133) (0.0214) (0.0134) (0.0194)
Military careerj−1 -0.0189 -0.0276 -0.0162 -0.0111

(0.0119) (0.0203) (0.0119) (0.0184)
Constant -0.00714 0.0231 -0.00757 0.0935

(0.0341) (0.0615) (0.0408) (0.0687)

Region Fixed-Effects No No Yes Yes

Observations 400 226 400 226

R-squared 0.132 0.188 0.168 0.403

Adj R-squared 0.083 0.104 0.096 0.308

Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1

to argue why having a longer tenure should have a specific impact on the quality of a

</div>
<span class='text_page_counter'>(37)</span><div class='page_container' data-page=37>

7

Final Remarks

Explaining growth volatility remains a pending issue of the classic theories. While a

part of this fickleness may be caused by unpredictable shocks, it is reasonable to argue

that politicians are, at least to some extent, responsible for the performance of a nation.

In this paper I found robust evidence to support this statement. The empirical strategy

relied on leaders’ transitions under which the identity of the leader was not likely to

depend on economic conditions. I first revisited Jones and Olken (2005) methodology by
using only leaders who died in power by natural death. Thus, with more recent data I

showed that leaders not only matter under autocracies (as the authors suggested) but also

under democratic regimes. Yet, results where sensitive to the leader’s sample suggesting

that the detection of such an effect is likely to depend on individual, spatial, or time

specificities.

To control for potential determinants of the magnitude of leaders’ effect, a larger

number of exogenous transitions was needed. I then included transitions where the exiting

leader was not able to run for a reelection due to a term limit constraint and transitions

where the entering ruler either won the elections by a small margin of victory, assumed
through royal succession, a constitutional order, was elected by an elite or as an interim

leader (as long as the predecessor exited in a regular manner). I first focused on the

magnitude of the growth shift from one leader to another. As we could expect leaders

matter more in poorer countries and in less democratic ones because they are subject to

less institutional constraints. An interesting mechanism is the one driven by term limits.

I found evidence that whenever a transition involves a shift from a leader who had a

possibility of reelection to one with an immediate term limit the fluctuation in growth

rates will be more pronounced. In general, leaders’ individual characteristics do not play

a role on the context on which leaders matter, even though in alternative specifications

where I smooth growth rates during the transitional years or when I consider leaders’
fixed-effects some of them become relevant. For instance, in some of the specifications I

found that more educated leaders are likely to manage to have a greater impact, while

the evidence for the transitions involving leaders with larger tenure or military chiefs is

mitigated.

I then focus on the nominal variation across those exogenous transitions as a proxy

of a leader’s performance. In those cases, rulers’ characteristics are much more relevant.

Nevertheless, it is more difficult to explain those variations than the ones in absolute

value, suggesting that there is still a large proportion of leaders’ qualities that rely on

unobserved skills. In line with evidence provided by the managerial literature, younger
leaders are associated with a better economic performance which can be explained either

by their risk behavior or by the fact that they have a longer expected time career in

</div>
<span class='text_page_counter'>(38)</span><div class='page_container' data-page=38>

is also favorable to the leader’s quality, even tough having an academic background in

this field does not seem to make the difference. On the contrary, the empirical evidence

suggests that when leaders with legal background or with a military career leave the

office, the growth rates tend to decrease. Finally, another robust evidence is that having

an immediate term limit seems to be beneficial for the economic performance. Indeed, it

can suggest that rulers who have the possibility of running for reelection put more effort

in political campaign, in the social sphere or in other areas that attract more voters rather

than in the economic performance.

Showing that, politicians have the power to shape the growth pattern open doors for

further research. It is then important to go deeper in the analysis of understanding and
proving through which mechanisms each one of those variables affects the magnitude of

the influence that a leader can have and the direction of this one. Besides, overcoming

the limitation of missing data about leaders’ party ideologies, leaders’ childhood variables

and different life experiences, it is likely to enrich the analysis as well as interacting the

leader’s effect with the other potential explanations of growth volatility such as financial

</div>
<span class='text_page_counter'>(39)</span><div class='page_container' data-page=39>

References

Daron Acemoglu and Simon Johnson. Disease and development: The effect of life

ex-pectancy on economic growth. Journal of Political Economy, 2007.

Daron Acemoglu and James A. Robinson. Why Nations Fail: The Origins of Power,

Prosperity and Poverty. Crown Business, 2012.

Daron Acemoglu, Simon Johnson, and James A. Robinson. The colonial origins of
com-parative development: An empirical investigation. The American Economic Review,

91(5), 2001.

Daron Acemoglu, Simon Johnson, James Robinson, and Yunyong Thaicharoen.

Insti-tutional causes, macroeconomic symptoms: volatility, crises and growth. Journal of

monetary economics, 50(1):49–123, 2003.

Alberto Alesina, Traviss Cassidy, and Ugo Troiano. Old and young politicians.

Econom-ica, 2015.

Alexander Baturo. Cursus honorum: Background, careers and experience of political
leaders in democracy and dictatorship new data and analyses. Politics and Governance,

4(2):138157, 2016.

Thorsten Beck, George Clarke, Alberto Groff, Philip Keefer, and Patrick Walsh. New

tools in comparative political economy: The database of political institutions. the

world bank economic review, 15(1):165–176, 2001.

Marianne Bertrand and Antoinette Schoar. Managing with style: The effect of managers

on firm policies. Quarterly Journal of Economics, CXVIII:1169–1208, 2003.

Timothy Besley, Jose G. Montalvo, and Marta ReynalQuerol. Do educated leaders
mat-ter? The Economic Journal, 121:205–227, 2011.

Jutta Bolt, Robert Inklaar, Herman de Jong, and Jan Luiten van Zanden. Rebasing

maddison: new income comparisons and the shape of long-run economic development.

GGDC Research Memorandum, 174, 2018.

Fernanda Brollo and Ugo Troiano. What happens when a woman wins an election?

evidence from close races in brazil. Journal of Development Economics, 122:28–45,

2016.

Raghabendra Chattopadhyay and Esther Duflo. Women as policy makers: Evidence from

</div>
<span class='text_page_counter'>(40)</span><div class='page_container' data-page=40>

Judith Chevalier and Glenn Ellison. Are some mutual fund managers better than others?

cross-sectional patterns in behavior and performance. Journal of Finance, 54(3):875–

899, 1999.

Martha L. Cottam. Introduction to political psychology. Lawrence Erlbaum Associates,

2004.

Simonton Dean Keith. Putting the best leaders in the white house: Personality, policy,

and performance. Political Psychology, 14(3):537548, 1993.

Luis Diaz-Serrano and Jessica Prez. Do more educated leaders raise citizens education?
IZA DP No. 7661, 2013.

William Easterly, Michael Kremer, Lant Pritchett, and Lawrence H. Summers. Good

policy or good luck? country growth performance and temporary shocks. NBER

Working Paper No. 4474, 1993.

Cali M. Elli, Michael C Horowitz, and Allan C. Stam. Replication data for: Introducing

the lead data set. Harvard Dataverse, Doi:10.7910/DVN/SYZZEY0, 2015.

Antonio Fatas and Ilian Mihov. Policy volatility, institutions, and economic growth.

Review of Economics and Statistics, 95(2):362–376, 2013.

Robert C Feenstra, Robert Inklaar, and Marcel P Timmer. The next generation of the

penn world table. American economic review, 105(10):3150–82, 2015.

Edward L. Gaeser, Rafael La Porta a ndFlorencio Lopez-de Silanes, and Andrei Shleifer.

Do institutions cause growth? NBER Working Paper No. 10568, 9(3):271–303, 2004.

Tom Ginsburg, James Melton, and Zachary Elkins. On the evasion of executive term

limits. Wm. & Mary L. Rev., 52:1807, 2010.

Henk E. Goemans, Kristian Skrede Gleditsch, and Giacomo Chiozza. Introducing
archi-gos: A dataset of political leaders. Journal of Peace Research, 46 (2):269–83, 2009.

Alan Heston, Robert Summers, and Bettina Aten. Penn world table version 6.1, 2002.

URL />

Alan Heston, Robert Summers, and Bettina Aten. Penn world table version 6.3 center

for international comparisons of production. Income and Prices at the University of

Pennsylvania, 2009.

</div>
<span class='text_page_counter'>(41)</span><div class='page_container' data-page=41>

Simon Johnson, William Larson, Chris Papageorgiou, and Arvind Subramanian. Is newer

better? penn world table revisions and their impact on growth estimates. Journal of

Monetary Economics, 60(2):255274, 2013.

Benjamin F. Jones and Benjamin A. Olken. Do leaders matter? national leadership and

growth since world war ii. Quarterly Journal of Economics, 120:835–864, 2005.

Richard Jong-A-Pin and Jakob De Haan. Political regime change, economic liberalization

and growth accelerations. Public Choice, 146(1-2):93–115, 2011.

Richard Jong-A-Pin and Shu Yu. Do coup leaders matter? leadership change and
eco-nomic growth in politically unstable countries. 2010.

Rahul Lahoti and Soham Sahoo. Are educated leaders good for education? evidence from

india. 2017.

John B Londregan and Keith T Poole. Poverty, the coup trap, and the seizure of executive

power. World politics, 42(2):151–183, 1990.

Arnold Ludwig. King of the mountain: The nature of political leadership. The University

Press of Kentucky, 2002.

Douglass C. North. Institutions and economic theory. The American Economist, 36(1):

3–6, 1992.

Lant Pritchett. Understanding patterns of economic growth: searching for hills among

plateaus, mountains, and plains. The World Bank Economic Review, 14(2):221–250,

2000.

Dani Rodrik, Arvind Subramanian, and Francesco Trebbi. Institutions rule: The primacy

of institutions over geography and integration in economic development. Journal of

Economic Growth, 9(2), 2004.

Herbert Spencer. The study of sociology, volume 5. Kegan Paul, Trench, Trăubner, 1892.

Frank Walter and Susanne Scheibe. A literature review and emotion-based model of

age and leadership: New directions for the trait approach. The Leadership Quarterly,

</div>

<!--links-->
Slide đề tài THUYẾT TRÌNH môn ANH văn TALK ABOUT THE STATE OF TRADING SHARES ON VIET NAM STOCK EXCHANGE NOWADAYS HOW DOES IT INFLUENCE ON THE ECONOMICS OF VIET NAM (1)

• 18
• 1
• 0