You are on page 1of 75

The Journal of Politics

and International Affairs
Volume IX, Issue II
Fall 2015

The Ohio State University

Cormac Bloomfield

David Nield

Editors in Chief
Wesley Swanson
Managing Editor for Content

Caroline Kinnen
Internal Communications

Gabriel Giddens
External Communications

Rachel Williams
Layout Editor
JPIA Editorial Staff

Jeffrey Glazner
Alex Lappert
Cara Schaefer
Kasey Powers
Marcus Andrews
Andrea Stanic
Katlyn Perani

Catey Stanley
Erik Wisniewski
Adrienne Michelson
Megan Simmons
Ryan Moore
Chris Delaney

A special thanks to our faculty advisors, Dr. Paul A. Beck and Ms. Christina Murphy, and The
Ohio State Department of Political Science Chair, Dr. Richard Herrmann, for guiding us and
making this journal possible.

The Journal of Politics


and International Affiars
Volume IX | Issue II | Fall 2015 | Print Edition

The Effect of Party Affiliation on Views Regarding a Woman
Serving as President


The Effects of Water Access On Government Health Care
Spending Worldwide


Christopher L. Pulliam, Syracuse University

Dillon R. Roseen, Daniel Furman & Mala Morjaria, Georgia Institute of Technology

Towards the Contradiction of Legal Supremacy in the
European Union


Cameron Mailhot, University of Minnesota, Twin Cities

French Aid for a French France
Matthew Cullom & Andrea Moneton, Georgetown University


The Journal of Politics and International Affairs at The Ohio State University is published biannually through the Ohio State
Department of Political Science at 2140 Derby Hall, 154 North Oval Mall, Columbus, OH 43210.
The JPIA was founded in the autumn of 2006 and reestablished in Winter 2011. For further information, or to submit questions or
comments, please contact us at
All rights reserved. No part of the publication may be reproduced, stored in a retrival system, or transmitted, in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise, without the written permission of the editors in chief of the
JPIA. The JPIA is copyrighted by the Ohio State Department of Political Science. The content of all papers is copyrighted by the
respective authors.
All assertions of fact and statements of opinion are solely those of the authors. They do not necessarily represent or reflect the views
of the JPIA Editorial Board, the Faculty Advisors, The Ohio State University, nor its faculty and administration.


Editorial Foreword

Cormac Bloomfield
David Nield
Editors in Chief

It is our pleasure to present the Fall 2015 edition of the Journal of Politics & International Affairs. This issue
features four papers chosen from over one hundred submissions from across the nation, making this issue the
most competitive in the history of the journal. Together, these papers showcase the best of undergraduate
political science and economics in a variety of subfields including: international relations, American politics,
comparative politics, political theory, and environmental politics.
This is our second volume as editors in chief of the Journal, and we could not be more proud of the progress
that has been made. From doubling of an editorial staff to multiplying submission rates, the Journal has
grown rapidly in success. Humbled by our success, we are bent on towards even greater achievements and
accomplishments for the Journal going forward, and are excited for the work we have ahead.
This issue would not be in your hands or on your screens if not for the support of the Ohio State University’s
Political Science department. We would like to especially thank Dr. Richard Herrmann for his vision and belief
in the Journal, Dr. Paul Beck for his guidance and advice, and Ms. Christina Murphy, whose assistance and
support is invaluable.
We would also like to thank most importantly you, the reader, for taking the time to read the Journal. We hope
that in reading this journal you will learn about a new topic, a new perspective, or simply enjoy fascinating
articles on a wide-range of topics. Our goal is to provide an appreciation for the quality of social science
research that can be, and is being, done at the undergraduate level nationwide. It is through you that the
research contained within these pages finds meaning.
Cormac Bloomfield and David Nield
Editors in Chief


The Journal of Politics & International Affairs

The Effect of Party Affiliation on Views Regarding a
Woman Serving as President
Christopher L. Pulliam1
The 2016 presidential election could potentially result in a woman serving as the President of the
United States. Women have largely been ignored in presidential politics and much of the public
opinion literature has attempted to answer why. While the contributions of previous scholars
is certainly noteworthy and valuable, past work is outdated and does not directly examine the
relationship between party identification and views on female candidates. Today’s political
climate is uniquely hyper-partisan and it is all but certain that Hillary Clinton, a woman, will
be the presidential nominee for the Democratic Party. This paper addresses the shortcomings
of the previous literature by addressing the relationship between party identification and views
regarding having a woman serve as president within today’s hyper-partisan political climate.
This paper suggests a model where party culture influences the views of its members. I hypothesize that among individuals in the United States those who identify with the Republican Party are
more likely to have negative feelings towards having a woman serve as president of the U.S. than
those who identify with the Democratic Party. The findings of my cross-tabulation confirm my
hypothesis and show that people who identify with the Republican Party are more likely to have
negative views regarding a woman serving as president than those who identify with the Democratic Party. This finding remains, even when controlling for gender.


ith the 2016 Presidential election less than one year away, excitement is growing around
who the potential nominees will be. For the GOP, Donald Trump, Ben Carson, Ted Cruz,

Jeb Bush, and Marco Rubio dominate the conversation. Across the aisle, the Democrat contenders
include Martin O’Malley, Bernie Sanders, and Hillary Clinton, with Quinnipiac University polling
data showing that Clinton is far ahead of the pack (Langer 2015).

Political scientists in the field of public opinion have been fascinated with the idea of hav-

ing a woman in the Oval Office for years. Many scholars have attempted to answer central questions regarding how public opinion influences the possibility of having a female president (e.g.
Welch and Sigelman 1982; Lawless 2004; Falk and Kenski 2006). Previous literature has largely
1 Christopher Pulliam is a 3rd year undergraduate studying political science and economics at Syracuse University. His research interests include public opinion, political behavior, ideology, gender, rightwing politics, and American conservatism. He would like to thank Professor Shana Kushner Gadarian,
Dr. Jonathan Hanson, and Professor Spencer Piston for their invaluable input, advice, and guidance.


Effect of Party Affiliation Regarding a Woman President

operated in a political climate where women are almost non-existent in presidential politics,
outside the role of wife and/or First Lady. All forty-four of the United States’ presidents have
been male. Sarah Palin and Geraldine Ferraro are the only two womenhas largely operated in a
political climate where women are almost non-existent in presidential politics, outside the role
of wife and/or First Lady. All forty-four of the United States’ presidents have been male. Sarah
Palin and Geraldine Ferraro are the only two women who have been Vice Presidential nominees,
and there has never been a female presidential nominee for either major party.

Considering the male-dominant presidential political landscape, it is exciting and note-

worthy to see a female candidate at the forefront of the presidential race. With President Barrack
Obama becoming the first African American president in 2008, and the possibility of Hillary
Clinton becoming the first female president in 2016, it is clear that the political landscape has
shifted away from one traditional requirement of electability: being a white male.

This paper will provide insight on the relationship between party identification and

whether one sees the potential for a female president as positive or sees the potential for a female
president as negative in today’s political landscape.
Women In a Different Political Landscape
The question of what influences public opinion towards having a woman as president has been
pondered by political scientists for decades. Some suggest that opinion changes when going between different demographic groups (Welch and Sigelman 1982; Fox and Smith 1998; Sigelman
and Welch 1984), while others suggest that issue salience may play an important role (Falk and
Kenski 2006). Still others suggest that gender stereotypes play a significant role (Sanbomatsu
and Dolan 2009; Lawless 2004; Alexander and Anderson 1993). Debate exists whether polling
data is inherently unreliable due to respondents not being truthful and the potential flaw of the
woman-president question or if traditional polling actually overestimates sexism.

Many political scientists have suggested that support for women changes between dif-

ferent demographic groups. Welch and Sigelman found that the core support for a woman to
become president was found among people who had more years of education, were younger, and
attended church less. The core opposition for a woman to become president was found among
people who had less education, were older, and attended church more often. These results sug-


The Journal of Politics & International Affairs

gest that conservatives may be more likely to disapprove of a female president than liberals; yet,
the data did not strongly support this conclusion (Welch and Sigelman 1982).

Similarly, a 1998 study suggests that differences in support for a female candidate may be

the result of regional and cultural differences. Fox and Smith polled students at the University of
Wyoming and University California Santa Barbara. They found that despite similar demographics in the cases (young and well-educated people), the group in California was more likely to
approve of a female presidential candidate (Fox and Smith 1998).

Other explanations have involved differences among gender and race. A 1984 study

conducted by Sigelman and Welch found gender may not influence an individual’s preference
for a female candidate. Women were found to not be consistently more supportive of female
candidates than men. However, among African Americans, women were more likely than men
to support a female candidate. Few gender differences were found among Caucasians (Sigelman
and Welch 1984).

Despite results that show demographic groups have different degrees of approval for a

potential female presidential candidate, others suggest that issue salience may be an important
factor. A 2006 study showed that individuals who said that terrorism, homeland security, and/or
U.S. involvement in Iraq was the most important issue facing the United States were more likely
to feel that a male would do a better job handling the problem as president. The conclusion of
Falk and Kenski was that issue saliency affects gender preferences in presidential candidates
much more than demographic and party identification variables (Falk and Kenski 2006).

Several scholars have suggested that holding gender stereotypes play an important role

with varying degrees of effect. A 2009 study claims that the public perceives gender differences
within both parties. They further state that a candidate’s gender plays a role even when the party
of the candidate is known. Additionally, stereotypes involving gender have different implications
for Democratic women and Republican women (Sanbomatsu and Dolan 2009).

Lawless asserts that women have a harder time wooing public opinion during times of

war (i.e. after September 11, 2001). This research challenges the claim that winning elections
have nothing to do with sex based off of the possibility that women fare poorer in elections when
“men’s issues” are dominant (Lawless 2004).

In contrast, a 1993 study that takes a more information-centric approach finds that when


Effect of Party Affiliation Regarding a Woman President

information about candidates is sparse gender role attitudes are important in initial evaluation of
candidates. Specifically, people tend to attribute certain leadership abilities and skills in handling
problems based on sex when considering a hypothetical candidate. Further, voters who have
equal gender role attitudes are more likely to favorably evaluate an actual female candidate. Finally, all incumbents were rated better on “male” and “female” qualities than challengers (Alexander and Andersen 1993).

Outside of possible explanations for why women are almost non-existent in presidential

politics, much of the literature has examined measurement concerns. Some scholars question
how effective polling can be when asking about sensitive issues, especially those regarding race
and gender. Streb, Burrell, Frederick, and Genovese suggest that people may give false answers
when asked if they would vote for a qualified female presidential candidate. A list experiment
was implemented to see if people were lying on polling questions. This experiment involved
giving the control group a list of four statements. They were asked to identify how many of the
statements made them “angry or upset.” The four statements read: “1. The way gasoline prices
keep going up, 2. Professional athletes getting million-dollar plus salaries, 3. Requiring seatbelts
to be used when driving, and 4. Large corporations polluting the environment.” The test group
was given these four statements with the addition of a fifth statement and were asked to do the
same thing. The fifth statement read: “A woman serving as president.” The key to this experiment is that respondents had to identify how many statements made them “angry or upset” not
which ones. Since respondents did not have to identify which statements they picked, the researchers saw this method as a way to evaluate the public’s true feeling about a female president
without the potential for lying. They found that despite the growing trend of support for a female
president over the years, about 25% of the American public would be angry about a woman serving as president. This result was constant among demographic groups (Streb et al. 2008).

Similar to Streb et al., a 1998 study suggests that the woman-president question is in-

herently flawed. The researchers state that recent studies at the time found that the candidate’s
gender did not have an effect on voter discrimination based on sex. The study involved polling
students at two different universities (the details of this study were stated on page two of this
paper). The results suggest that bias against women candidates change significantly among
regions. The researchers conclude that a woman-president question may be flawed and should


The Journal of Politics & International Affairs

not be taken at face value because people may be unaware of their prejudices or are willing to lie
about them (Fox and Smith 1998).

In contrast, Falk and Kenski suggest that traditional polling methods actually overesti-

mate sexism and underestimate the role that individual characteristics play, such as party identification, in deciding if one will vote for a woman for president. They further claim that many
people would in fact vote for a woman if given a realistic matchup (Falk and Kenski 2006).

A problem that is shared with all past literature is that we are now in a political climate

where the possibility of having a woman president is very likely, if not inevitable. Hillary Clinton is the front-runner for the Democratic nomination, and the Republicans have not yet found
a candidate who is guaranteed to challenge her properly. In the past, asking individuals about
having a woman serve as president have been largely hypothetical. Now that the potential for a
woman to sit in the Oval Office seems very possible, the American public’s attitudes about this
question may have changed significantly.

Additionally, much of the present literature interacts with the relationship between party

identification and views on having a woman serve as president; however, none directly explore
this relationship. When the new hyper-partisan political climate is considered, this omission is
even more important to examine. This paper is an attempt to properly examine the relationship
between party identification and views regarding having a woman serve as president in the modern political climate where the potential for having a female presidential candidate is very real
and strong partisanship divides the country.
Party Culture Influences Views on Women
When looking at how the American public views the possibility of having a female president, it
is natural to consider divisions along party lines. Due to the hyper-partisanship present in today’s
political spectrum, differences between Republicans and Democrats are expected with almost everything. In examining why there may be differences between the two major parties in feelings
towards a female president, it is necessary to consider who makes up each party, how this makeup could influence party culture, and who is seen as a presidential front runner for either party.

Republicans tend to be made up of conservatives, including traditionalist conservatives.

An emphasis of traditionalist conservatism is the preservation of traditional home life. This view


Effect of Party Affiliation Regarding a Woman President

includes valuing gender roles that include men being bread winners and women staying home
and taking care of the children. While not all Republicans subscribe to the above view, a significant portion of their voter base may hold this viewpoint. These individuals may find a woman
holding the highest office of the country as threatening to the traditional concept of the family.
In turn, these individuals may be against seeing a woman in the White House. This line of thinking may influence party culture and direction; thus, affecting the views of other party identifiers
in regard to encouraging negative views towards having a woman serve as president.

Additionally, the Republican Party has taken a hard stance against passage of the Pay-

check Fairness Act. This bill is designed to close the gender wage gap.. While the underlying
causes for this gap are debated, the Paycheck Fairness Act is an attempt to eliminate any workplace pay discrimination due to gender. Considering that Republicans have blocked passage of
this bill suggests that there may be complacency in allowing for this wage gap to remain. This
possible complacency may suggest that people who identify with the GOP may be against women rising to more positions of power, especially the presidency. This concrete policy action may
influence members of the Republican Party to hold some anti-women attitudes and be less favorable to the idea of having a woman serve as president.

Looking at the other side of the aisle, the Democratic Party holds many individuals who

identify themselves as liberal, specifically social liberals. Many feminists and radical feminists
have been drawn to the Democratic Party due to their liberal social values. These individuals
would be strongly in favor of having women in positions of power, especially the presidency.
This line of thinking may influence party culture to encourage pro-women attitudes, and thus,
encourage positive views regarding having a woman serve as president.

Additionally, women tend to identify with Democrats more than with Republicans. As-

suming that women would have more favorable views of having a female president than men
would, it is possible to believe that the membership of the Democratic Party would be more
favorable to having a female president than Republicans.

Important consideration needs to be paid to policy positions as well. The Democratic

Party is largely in favor of passing the Paycheck Fairness Act and have made the gender wage
gap one of their rallying cries. Considering this policy position, it is reasonable to assume that
the Democratic Party appears to be pro-women to many, and would thus attract membership that


The Journal of Politics & International Affairs

would have more favorable views in regard to having a woman president. This concrete policy
position may influence the thinking of party members to be pro-women in focus, and thus, more
positive to the idea of having a woman serve as president.

Perhaps most importantly, party identification may influence individuals’ views on having

a woman president due to the fact that the Republican field of potential presidential candidates is
almost all male, while the top Democratic contender is female. Respondents may remember this
fact in the back of their minds, specifically that Hillary Clinton will most likely be the Democratic nominee. Considering this partisan spilt, respondents who identify as Republican may rate
having a female president lower than respondents who identify as a Democrat.

Essentially, my model suggests that individuals are drawn to a political party for a reason

or a set of reasons. Once one identifies with that party, they are influenced by the culture of that
party. This party culture could be the result of the membership of the party, the policy positions
of that party, and who makes up the “big names” of the party. The culture of the party influences
and may change the views of the individual. A concrete example of the above will be given for

Consider Tom. Tom has very favorable views towards free-market economics. Tom sees

the Republican Party as being favorable towards free-markets and decides to become a member
of the GOP. Now that he identifies as Republican, Tom seeks out what other Republicans think.
He reads blogs, online magazines, and news articles that explain and defend the views of various
wings of the GOP. He keeps up to date on various policy issues and which side the Republican
Party takes. He is aware of current polls and who is being talked about for the GOP presidential
nomination. When faced with the question of if he would vote for a female presidential candidate, Tom may recall the piece he read that eloquently defended traditionalist conservatism. He
may also recall how Republicans are against the Paycheck Fairness Act. Finally, he may recall
that potential presidential candidates for the GOP include almost exclusively males and that the
leading Democratic candidate is Hillary Clinton, a female. Tom may want to be like his fellow
Republicans— a legitimate member of the group. So, Tom considers the above and determines
that a true Republican would not vote for a female presidential candidate, or be neutral to the

Therefore, in order to more effectively understand how the U.S. population views the


Effect of Party Affiliation Regarding a Woman President

possibility of having a woman president, party identification needs to be considered. When the
above are put into consideration, it seems that party culture may have a significant influence
on views regarding a woman serving as president. A comparison of the two major parties also
suggests that Republicans will have a more negative view than Democrats. The hypothesis for
this paper is that among individuals in the United States, those who identify with the Republican
Party are more likely to have negative feelings towards having a woman serve as president of the
U.S., than those who identify with the Democratic Party.
Research Design
In order to test this hypothesis, the variables need to be measured properly. The independent
variable is party affiliation. The dependent variable is feelings towards having a woman serve as
president of the United States. The unit of analysis is individuals.
Data from the American National Election Study (ANES) of 2012 will be used to measure the
independent and dependent variables. This study will be used because it is the most recent study
available, and it focuses on investigating national election outcomes. Given that the president is
chosen through a national election, the ANES is ideal.

As mentioned earlier, there is debate in the literature regarding the use of traditional poll-

ing methods when inquiring about certain issues, in this case gender. Considering that one camp
thinks traditional polling overestimates negative views towards women, while the other camp
thinks traditional polling underestimates negative views towards women, this paper will assume
that the ANES is a good way to measure the attitudes of voters. While using the ANES is certainly not perfect, it has been long held that traditional polling methods are good measures of
respondents’ feelings. Since the goal of this paper is not to resolve the above debate, I will yield
to the wisdom of using traditional polling methods.

The variable “WPRES_GDBD” will be used to measure individuals’ feelings towards

having a woman serve as president. The survey question asks, “Would it be good, bad, or neither
good nor bad if the United States has a woman President in the next 20 years?” The answers
were recorded in three categories: “Good if the United States has a woman President”, “Bad if
the United States has a woman President”, and “Neither good nor bad if the United States has a
woman President.” This question reliably measures the dependent variable because it consistent-


The Journal of Politics & International Affairs

ly measures respondents’ feelings about having a female president. This question is also valid
because it contains no systematic error. When examining the question at face value, asking if
having a woman president in 20 years is good, bad, or neither should be a valid measure of respondents’ feelings towards having a woman serve as president.

The variable “PID_SELF” will be used to measure individuals’ party identification. The

survey question asks, “Generally speaking, do you usually think of yourself as a [Democrat, a
Republican, / a Republican, a Democrat], and independent, or what?” The answers were recorded in three categories: “Democrat”, “Republican”, and “Independent.” The question is a reliable
measure of the dependent variable because it measures individuals’ party identification consistently. The question is also valid because it does not suffer from systematic error. At face value, asking someone how they identify themselves in terms of political parties should be a valid
measure of respondents’ party affiliation.

I will be performing a cross-tabulation between the independent and the dependent vari-

ables. Considering that the independent variable is nominal and the dependent variable is ordinal, a cross-tabulation would be an appropriate tool.

In addition to finding the zero-order relationship between the independent and dependent

variables, I will also be finding a partial-effect relationship by controlling for gender. Controlling for gender is important because women may show a more favorable bias towards having
a female president than men. Also, women tend to identify with Democrats more than men do,
and men tend to identify with Republicans more than women do.

I will be using the variable “GENDER_RESPONDENT” to measure the control variable.

The variable combines the variables for gender face-to-face and web modes of administering the
survey. The question may run into reliability issues because the interviewer observes the individual’s gender for administering the survey face-to-face. Considering that the survey was administered by several different interviewers, their categorization of individuals may differ slightly.
Additionally, this question runs into issues of validity. The question relies on the interviewer
correctly identifying if the respondent is male or female. The interviewer may identify individuals incorrectly. Despite these potential issues, it is still unlikely that the dataset is significantly
inaccurate. Since there are no other gender variables in the ANES, this variable is the best option
by default.


Effect of Party Affiliation Regarding a Woman President

Analysis and Assessment
The findings of the cross-tabulations support my hypothesis by showing that individuals in the
United States who identify with the Republican Party are more likely to have negative feelings
towards having a woman serve as president than those who identify with the Democratic Party.

The charts below display the findings of this study. A majority of Democrats (56.2%) see

having a woman serve as president in the next 20 years as good, while only 23.8% of Republicans feel the same way. Additionally, 9.6% of Republicans feel that having a woman serve as
president is bad, while only 3.6% of Democrats feel the same way. Finally, 66.7% of Republicans say that having a woman serve as president in the next 20 years as neither good nor bad,
which is much more in comparison to only 40.2% of Democrats.
Chart 1
Women President in next 20

Feelings On Having a Woman President













Neither good
nor bad










These three findings are significant at the 0.05 level. Pearson’s chi-square value, for four

degrees of freedom is 391.739. The significance value is 0.000, which passes the 0.05 threshold
of significance. The null hypothesis is rejected with 95% confidence. Significant differences
exist between how Democrats and Republicans feel about having a woman serve as president
in the next 20 years. Democrats tend to have more favorable feelings towards having a woman
president than Republicans. Republicans tend to have more negative or neutral feelings towards


The Journal of Politics & International Affairs

having a woman president than Democrats.

Further, this study finds that 49.7% of male Democrats and 18.6% of male Republicans

think that having a woman president in the next 20 years would be good. Also, 3.6% of male
Democrats and 9.1% of male Republicans say that having a woman president in the next 20
years would be bad. Of male Democrats, 46.7% feel that having a woman president in the next
20 years would be neither good nor bad, while 72.3% of male republicans feel the same way.
Chart 2

Feelings on Having a Woman President
(Controlling for Gender)

President in
next 20

























nor bad
















The above findings are significant at the 0.05 level. Pearson’s chi-square value for males

is 178.755 with four degrees of freedom. The significance value is 0.000, which passes the 0.05
threshold of significance. The null hypothesis is rejected with 95% confidence. Significant
differences exist between how male Democrats and male Republicans feel about having a wom-


Effect of Party Affiliation Regarding a Woman President

an serve as president in the next 20 years. Male Democrats tend to have more positive views
towards having a female president than male Republicans. Male Republicans tend to have more
negative or neutral views towards having a female president than male Democrats.

Among women, 60.9% of Democrats feel that having a female president in the next 20

years would be good, while only 29.4% of Republicans feel the same. Further, 3.7% of female
Democrats and 10% of female Republicans think that having a woman president in the next 20
years would be bad. Finally, 35.4% of female Democrats say that having a woman president in
the next 20 years would be neither good nor bad, which is much less than the 60.6% of Republican women who say the same thing.

The above findings are significant at the 0.05 level. Pearson’s chi-square value for fe-

males, with four degrees of freedom, is 190.637. The significance value is 0.000, which passes
the 0.05 significance threshold. With 95% confidence, the null hypothesis is rejected. Significant differences exist between how female Democrats and female Republicans feel about having
a woman serve as president in the next 20 years. Female Democrats tend to have more positive
feelings towards seeing a woman in the Oval Office than female Republicans. Female Republicans tend to have more negative or neutral feelings towards seeing a woman in the Oval Office
than female Democrats.

When comparing males and females who feel that having a woman serve as president is

good, there is an additive relationship. The change between the percentage of male Democrats
who see it as good and the percentage of male Republicans who see it as good is -31.1%. The
change between the percentage of female Democrats who see it as good and the percentage of
female Republicans who see it as good is -31.5%. These values are not close to zero and yet are
close together, indicating a non-spurious, additive relationship.

Further, when comparing males and females who feel that having a woman serve as pres-

ident is bad, there is also an additive relationship. The change between the percentage of male
Democrats and male Republicans who see it as bad is 5.5%. The change between the percentage
of female Democrats and male Democrats who see it as bad is 6.3%. Again, these values are not
close to zero and yet are close together, indicating a non-spurious, additive relationship.

In this comparison of males and females who feel that having a female serve as president

is neither good nor bad, the relationship is once again additive. The change between the per-


The Journal of Politics & International Affairs

centage of male Democrats and male Republicans who see it as neither good nor bad is 25.6%.
The change between the percentage of female Democrats and female Republicans who see it as
neither good nor bad is 25.2%. The relationship cannot be spurious due to the fact that these values are not close to zero. Further, the relationship must be additive because these values are very
close together.
The research showed that there is a significant relationship between party identification and
feelings towards having a woman serve as president of the United States. While more work still
needs to be done, the results of this study suggest that my model may be correct. Dealing with
causality and reverse causality is always difficult, especially with cross-tabulations. Indeed, I
cannot discount the potential for reverse causality; however, the data does not discount the possibility that the causal arrow points from party identification to views on having a woman serve
as president. Individuals who identify with Republicans are more likely to have negative views
about having a woman president in the next 20 years than individuals who identify with Democrats. When controlling for gender, both male and female Democrats are more likely to have
positive views regarding having a woman president in the next 20 years than their respective
Republican counterparts.

The relationship between the independent, dependent and control variables is additive.

This means that gender has a very minor effect on feelings towards having a woman president.
Further, the independent and dependent variables retain a causal relationship even after applying the control variable. While gender helps to explain differences in feelings towards having
a woman serve as president, party identification still causes the difference in feelings towards
having a woman serve as president among individuals.

Considering the results of this study, it seems that negative gender bias towards a female

president would be minimal. Most Americans seem to have positive feelings regarding having a
woman in the White House, or are neutral on the subject. With 2016 fast approaching, the data
in this study bodes well for Hillary Clinton.

As the leading contender for the Democratic nomination, it seems that she will have no

problem gaining the support of her party. With the majority of Democrats saying that having a


Effect of Party Affiliation Regarding a Woman President

woman serve as president would be good, there should be few issues of gender bias in the primaries. So far, the race seems to reflect the data of this study. Clinton is considered the far and
away favorite. She leads Senator Sanders and Governor O’Malley in the polls and in the endorsement race by large margins (Langer 2015, Bycoffe 2015).

Once the primaries are over, the story only changes slightly. The research shows that

the total number of individuals who have negative feelings towards having a woman in the Oval
Office is 5.4%. Assuming that all respondents have been honest, the level for possible gender
bias among voters is minimal. However, if the 2016 election is similar to recent presidential
elections, even the slightest disadvantage could decide the election. There is a very good chance
that this slight gender bias could be highlighted because she will most likely be running against
a male candidate since the frontrunners in the GOP nomination race are all men (Langer 2015).
Even then, the level of gender bias may be minimal because many individuals who see having a
woman serve as president as a negative would end up voting for the Republican candidate regardless. More work needs to be done after the election in order to properly determine if gender
bias hurt Clinton. Similar work has been done on how President Obama won in 2008 despite
racial prejudice (Piston 2010). This same kind of research should be done in the future in order
to determine how gender bias influenced Americans’ vote choice in 2016.

Assuming that the results of this study are true, it seems that the United States could have

its first female president. This suggestion is not surprising considering that the U.S. elected its
first African American president in 2008, and again in 2012. While issues of racism and sexism
still exist, it seems that they do not significantly hinder the presidential election process. Fortunately, the 2016 election will come down to the quality of the candidates, not whether they are
male or female.


The Journal of Politics & International Affairs

Alexander, Deborah, and Kristi Andersen. 1993. “Gender as a Factor in the Attribution of Leadership
Traits.” Political Research Quarterly 46 (3) (Sep.):527-45.
American National Election Study. 2012. “ANES 2012 Time Series.”
analysis/?dataset=nes2012 (December 1, 2014).
Bycoffe, Aaron. 2015. “The Endorsement Primary.” November 23. http://projects.fivethirtyeight.
com/2016-endorsement-primary/ (November 25, 2015).
Falk, Erika, and Kate Kenski. 2006. “Sexism Versus Partisanship: A New Look at the Question of
Whether America is Ready for a Woman President.” Sex Roles 54 (7-8) (Apr 2006):413-28.
Falk, Erika, and Kate Kenski. 2006. “Issue Saliency and Gender Stereotypes: Support for Women as
Presidents in Times of War and Terrorism.” Social Science Quarterly 87 (1) (MARCH):1-18.
Fox, RL, and Eran Smith. 1998. “The Role of Candidate Sex in Voter Decision-Making.” Political
Psychology 19 (2) (JUN):405-19.
Langer Research Associates of New York, NY. 2015. “ABC News/Washington Post Poll: 2016 Election”
November 22.
(November 23, 2015).
Lawless, Jennifer L. 2004. “Women, War, and Winning Elections: Gender Stereotyping in the PostSeptember 11th Era.” Political Research Quarterly 57 (3) (Sep 2004):479-90.
Piston, Spencer. 2010. “How Explicit Racial Prejudice Hurt Obama in the 2008 Election.” Political
Behavior 32 (4):431-51.
Sanbonmatsu, Kira, and Kathleen Dolan. 2009. “Do Gender Stereotypes Transcend Party?” Political
Research Quarterly 62 (3) (Sep 2009):485-94.
Sigelman, Lee, and Susan Welch. 1984. “Race, Gender, and Opinion Toward Black and Female
Presidential Candidates.” The Public Opinion Quarterly 48 (2) (Summer):467-75.
Streb, Matthew J., Barbara Burrell, Brian Frederick, and Michael A. Genovese. 2008. “Social Desirability
Effects and Support for a Female American President.” The Public Opinion Quarterly 72 (1)
Welch, Susan, and Lee Sigelman. 1982. “Changes in Public Attitudes Toward Women in Politics.” Social
Science Quarterly 63 (2) (June):312-22.


Effect of Party Affiliation Regarding a Woman President


The Journal of Politics & International Affairs

The Effects of Water Access On Government Health Care
Spending Worldwide
Dillon R. Roseen1, Daniel Furman2 & Mala Morjaria3
Water access worldwide affects numerous areas of personal health. This paper attempts to analyze the effects that water access has on total government health care expenditure in different
countries. By utilizing a cross-sectional analysis from over 80 countries, we provide a unique
view of how lack of water access may burden health care. This paper grounds its hypothesis in
Pareto economic welfare policy, which helps explain the relationship between these variables.
In addition to water access, we utilized additional independent variables such as private health
care spending and tobacco use and a dummy variable adjusting for differences in developed
and non-developed countries. These variables enabled us to gain a holistic view of the effects of
water access on health care spending.


ccording to a report from the World Health Organization and UNICEF released in 2014,
around 748 million people worldwide lack access to clean, drinkable water (UNICEF

2014). 3.4 million people die every year from preventable water-related diseases, such as diarrhea
(Prüss-Üstün 2008), of which 90% can be attributed to unsafe drinking water. Diarrhea was the
second leading cause of death in the world in 2012 among children under five, causing at least one
child death every minute (World Health Organization 2014). Evidence for the causal relationship
between unsafe drinking water and diarrheal deaths is overwhelming (Fewtrell 2004). Water borne
illnesses and deaths disproportionately affect people in poor regions of the world who lack access
to proper health care facilities. In developing countries, access to health care services is especially
scarce given the low number of health workers in proportion to the total population (Peters 2008).
This gap in service raises health care costs, which greatly affects people of lower economic status

Dillon Roseen earned his B.S. in Economics & International Affairs from Georgia Tech in December 2015. His research interests include international distributive justice, development studies, and
law. After a few more adventures, he intends to start work in the national security sector in Washington,
Daniel Furman is a fourth-year Business Administration student at Georgia Tech. His research interests include financial econometrics, revenue management, and social welfare. Daniel is set to graduate
in May 2016. Upon graduation, he will be joining The Boston Consulting Group in their Atlanta office.
Mala Morjaria graduated from Georgia Tech in December 2015 with her B.S. in Industrial and
Systems Engineering. Her research interests include sports economics and econometrics.
Faculty guidance was provided by Dr. Shatakshee Dhongde.


Effects of Water Access on Government Health Care Spending

who must pay a larger share of their total income to receive proper care. The WHO reports that
approximately 100 million people every year fall into poverty as a result of the costs associated with illness (Etienne 2010). A large share of the health care burden in developing countries
comes as a result of completely preventable diseases, such as those caused by unsanitary water.
If health care facilities in developing countries are freed from the burden of treating preventable
water-related diseases, resources could be redirected to provide better quality and more effective
care. Therefore, our analysis seeks to identify the relationship between water access and governmental health care expenditure.
Our main hypothesis states that if a country has better access to improved potable drinking
sources4, then that country will spend less money on health-related costs5 as a percentage of total
government expenditures. Our hypothesis is supported by the rationale that as fewer individuals
become sick from water-borne illnesses, there will be less of a financial burden on the health-related spending of a specific country. This hypothesis corresponds to the foundation of the social
welfare economic theory. The social welfare economic theory analyzes the relationship between
national policies on the overall well being of a community.
Following Adam Smith’s teachings of the Invisible Hand, social welfare economics
states that the allocation of sums tends to benefit the whole community. Our hypothesis also
makes use of Pareto’s concept of social welfare economics. Pareto found that there is a point
(the Pareto Optimal/Efficient point) where afterward, any gains in social welfare efficiencies are
contrasted by subsequent losses by another party (University of Toronto 2014). This can be seen
by the large differences in health care systems and water infrastructure between developed and
developing countries, which corresponds with the points before and after the Pareto Efficiency
Point. Our hypothesis concludes that successfully improving water access in both developed and
developing countries will result in a reduction of health care spending.
4 Drinking water is water used for domestic purposes, drinking, cooking and personal hygiene
(WHO). Safe drinking water comes from a household connection; public standpipe; borehole; protected
dug well; protected spring; rainwater and is free from any microbial, chemical and physical characteristics
that meet WHO guidelines or national standards on drinking water quality (WHO).
Total expenditure on health (THE) is measured as the sum of all financing agents managing funds
to purchase health goods and services.


The Journal of Politics & International Affairs

Literature Review
The existing literature suggests that there is, in fact, a global water crisis, particularly in developing nations (Kirpich 2014, Frederiksen 2005, Walter 2014). One millennium development
goal (MDG) is to halve the proportion of people in the world without access to safe drinking
water. While this goal was met in 2010 (World Health Organization 2014), a crisis remains in
some countries. In many developing nations, a vast number of people still live without access to
improved water sources.
The seminal work in the area of water infrastructure and health comes from Annette
Prüss-Üstün, Robert Bos, Fiona Gore, and Jamie Bartram of the World Health Organization. In
their work titled “Safer Water, Better Health” published in 2008, they concluded that one tenth
of the global disease burden could be “prevented by improving water supply, sanitation, hygiene
and management of water resources” (Prüss-Üstün 2008). The study was initiated as a progress
report on the MDG to improve environmental sustainability, specifically Target 10, which is
described above.
A different study performed by Yongsi (2010) finds that in Cameroon a statistically high
number of cases of diarrhea were caused by poor water quality. The overall burden of these
diseases led to an increase in the burden on health care providers in Cameroon. The results of
this study can be applied to many other developing nations. The causes of variability in diarrheal
levels between neighborhoods in Cameroon are particularly interesting. While Yongsi’s report
illustrated the correlation between water quality and the health care burden, it failed to explain
regional variations in diarrheal cases throughout Cameroon. Once these differences are identified, policy makers can learn how to intervene in communities to lessen the total burden on
health care.
Furthermore, as countries spend less on health care they can redirect governmental resources to improve infrastructure to help prevent water-borne illnesses. We can better see the effects of proper water interventions by looking at Eder, Schooley, Fullerton, and Murguia’s (2012)
research on post-project impacts in Bolivia. Their study looked at the water intervention process
by analyzing the $26 million Development Assistance Program (DAP) that supported water
access projects. After six years, DAP communities had better infrastructure for community water
systems and household water systems than control communities. Also, DAP sponsored commu-


Effects of Water Access on Government Health Care Spending

nities were around “30% more likely to be rated good to very good or satisfactory for status of
water infrastructure” (Eder 2012). The communities were able to improve their health and living
standards by investing in water infrastructure, thus reducing the overall health burden.
Other existing sources provided us with good modeling techniques to frame our research.
Dinar and Saleth’s (2005) study introduced an interesting modeling technique in which countries
were grouped according to their level of development to help identify the most common constraints on water systems. They find that these constraints stem from poor resource management
and inefficient water use rather than a physical limitation on water access. Therefore, they stress
that improving management infrastructure can solve the water crisis. Dinar and Saleth chose to
use the Water Institution Health Index indicators to evaluate and compare the relative health of
water institutions. In their study, 43 selected countries were arranged “into three broad groups
with countries having good, moderate, or poor performing water institutions” (Dinar 2005). This
grouping helped identify the weaker performing water institutions that need the most help to
better allocate resources and perfect inefficiencies. Our analysis will adopt a similarly grouped
modeling technique in order to better identify how countries in various stages of development
respond when there is improved water access.
Our study brings a unique perspective to prior literature and hopefully can serve as a supplement for past work. We focus on a cross sectional analysis of countries around the world to
get a holistic perspective on health care spending. Currently, there is little analysis on the direct
effects of improved water access on government health expenditures in a country. Many studies
analyze the microeconomic impact of poor water access on an individual or household’s health
care costs. Macroeconomic studies on water access, however, tend to be limited to specific regions or countries and fail to link water access to a country’s health care spending. The research
in this paper attempts to provide a broad level, holistic perspective that other research lacks.
This study is primarily interested in the impact a country’s access to improved water sources has
on its general government health care expenditure. Our dependent variable in this model is general government expenditure on health (GGHE). To measure GGHE, we looked at a country’s total government health expenditure as a percentage of its general government expenditure (GGE).


The Journal of Politics & International Affairs

We also incorporated several independent variables in our study. Our primary independent variable is the proportion of the population using improved water sources as a percentage
of the total population. Access to an improved water source is the primary independent variable.
If our hypothesis proves to be true, a country with better access to improved water sources will
have a significantly lower GGHE.
Table 1.1 Defining independent and dependent variable



Proportion of the population using Improved Water
Sources (% of total pop)
General Government Expenditure on Health (% of
Total Government Expenditure)
Private Share of Total Health Expenditure (%)


Smoking Adults (% of pop over age 15)


Infant Mortality Rate (per 1,000 births)
Dummy variable signifying if a country is developing (1) or developed (0)


Data Source
United Nations Statistics
The World Health Organization
The World Health Organization
The World Health Organization

Variables Included in Simple Regression:
publichealth= β₀ + β₁water
1. Improved Water Sources - The percentage of the total population who use any of the
following types of water supply for drinking: piped water into dwelling, plot or yard;
public tap/standpipe; borehole/tube well; protected dug well; protected spring; rainwater collection and bottled water (if a secondary available source is also improved).
It does not include unprotected well, unprotected spring, water provided by carts with
small tanks/drums, tanker truck-provided water and bottled water (if secondary source
is not an improved source) or surface water taken directly from rivers, ponds, streams,
lakes, dams, or irrigation channels. This is the standard definition provided by the
United Nations site for MDG indicators.
2. General Government Health Expenditure - GGHE as % of GGE, provided by the
World Health Organization Global Health Expenditure Database.

Effects of Water Access on Government Health Care Spending

Additional Variables Included in Multivariate Regression:
publichealth= β₀ + β₁water + β2privatehealth + β3mortality + β₄tobacco
1. Private Share of Total Health Expenditure - Private share of health spending as given
as a percentage of total health expenditure (THE)6. The WHO Global Health Expenditure Database provided this data set. We included this variable because countries with
a larger share of THE funded by private entities (such as individuals, NGOs, etc.)
would have a lower percentage of THE come from a public source.
2. Infant Mortality Rate – As defined by UNICEF, the probability that a child born in a
specific year will die before reaching the age of one, expressed as a rate per 1,000 live
births. UNICEF also provided the data for this variable. Infant mortality could affect
THE, which is why we included it in our regression. In a country with a high infant
mortality rate, public health spending would likely be low because there is inadequate
access to proper medical services.
3. Smoking Adults - Prevalence of current tobacco use of both sexes aged 15 and over,
provided by the World Health Organization Statistical Information System. We felt it
was important to include this variable in our research because of the large effect that
tobacco use has on health. For instance, the Centers for Disease Control and Prevention estimate that cigarette smoking causes 1 in every 5 deaths in the United States
(CDC 2014).
Taken as a collective, these independent variables provide the foundation upon which we
constructed our statistical model. Ideally, these variables will give us a clear picture regarding the
factors that influence total public health spending. For the purpose of the research, we often utilized data compiled by Gapminder but originally available from sources like the WHO, UNICEF,
and the UN Statistics Division.
The addition of a dummy variable, developing, allowed us to control for different types
of countries: both developed and developing nations. Appendix 1.1 shows a breakdown of the
developed nations and developing nations used in this study. Splitting these countries up allowed
us to control for the large differences in health care systems and water infrastructure between

Total Health Expenditures includes both public and private sources


The Journal of Politics & International Affairs

developed and non-developed countries. Developed nations tend to be richer than non-developed
nations and therefore spend more money on health care systems. Additionally, water infrastructure in developed nations far surpasses that in the developing world. The dummy variable gave
us deeper insights into the realities for both types of countries.
Table 2.1 Descriptive Statistics: Breakdown of variables based on 81 surveyed countries’ data.
Data sources are noted above. Includes data from both developed and developing nations.



Std. Dev



publichealth (% of GGE)






water (% of total pop)






privatehealth (% of THE)






mortality (per 1,000 births)
tobacco (% of tobacco users)






Table 2.2 Descriptive Statistics: Developed Countries Only



Std. Dev



publichealth (% of GGE)






water (% of total pop)






privatehealth (% of THE)






mortality (per 1,000 births)






tobacco (% of tobacco users)






Table 2.3 Descriptive Statistics: Developing Countries Only



Std. Dev



publichealth (% of GGE)






water (% of total pop)






privatehealth (% of THE)






mortality (per 1,000 births)






tobacco (% of tobacco users)







Effects of Water Access on Government Health Care Spending

Figure 1.1 Water Sources and Government Health Expenditures for 81 countries

Gauss Markov Assumptions:
It is just as important to interpret the results of data as it is to verify the reliability of data. The
data in the study was verified using the Gauss Markov Assumptions for Multivariate Regression.
Assumption 1: Linear Parameters
We checked that our regressions were linear in parameters, meaning our model could be written:
Y= β₀ + β₁X1 + β₂X2 + β₃X3 + β₄X4+ βkXk + u
Assumption 2: Random Sampling
The data we used included every country that had information for our desired variables. The
equation for a random sample of the population can be shown as:
Yi= β₀ + β₁Xi1 + β₂Xi2 + β₃Xi3 + β₄Xi4+ βkXik+ ui
Assumption 3: Zero Conditional Mean
Based on our regression equation, the residuals for all our countries were plotted in the histogram
below. While Figure 2.1 shows a slight rightward skew in our data, it is not enough to violate the
normality assumption.


The Journal of Politics & International Affairs

Figure 2.1 Histogram Verifying Zero Conditional Mean

Assumption 4: No Perfect Collinearity
We carefully selected independent variables to ensure that there was no perfect collinearity and
that the dependent variable, publichealth, was related to the independent variables in a predictable manner. A correlation matrix is produced to demonstrate that there is no perfect collinearity:
Table 3.1 Correlation Matrix Verifying No Perfect Collinearity








Assumption 5: Homoskedasticity
Finally, we had to make the assumption of homoskedasticity which means that variance in the
error term is the same for all combinations of our explanatory variables. Our variety of independent variables should control for unknown errors in homoskedasticity. Although these last three
assumptions are rather all encompassing, they are key to utilizing the multivariate regression
model assuming best linear unbiased estimators.


Effects of Water Access on Government Health Care Spending

Table 4.1 Simple Regression: STATA Results, water regressed on dependent variable publichealth
Independent Variable
Number of Observations

Simple Regression

* denotes significance at the 10% level ** denotes significance at the 5% level *** denotes significance
at the 1% level

In the initial simple regression, the relationship between improved water access and public health
care spending was positively correlated for the 81 surveyed countries. From an economic standpoint, the coefficient of .101 shows that there is an increase in the value of publichealth by.101
when water increases by one unit. Because this coefficient is significant at the 1% level there is
a statistically significant correlation between access to improved water sources and public health
care spending and the correlation is economically viable. However, this relationship violates our
hypothesis. Instead of a negative correlation like we initially predicted, there was a positive correlation between improved access to potable water sources and public health care spending. Put
simply, this means that, as a whole, countries with better access to water spend more money on
health care. Our initial explanation for this result is that countries with better access to water are
typically richer. Presumably, rich countries would also spend more public money on health care,
rather than relying on outside, private sources for support. Since this regression includes both
developed and non-developed nations, future regressions that split developed and non-developed
nations may have different results. We ran an additional regression using dummy variables in
Table 6.1 to test this hypothesis. First, though, we looked at a multivariate regression without the
dummy variables to serve as a baseline for the dummy variable regression.


The Journal of Politics & International Affairs

Table 5.1 Multiple Regression: STATA Results without Dummy Variables
Independent Variables

Unrestricted Multiple Regres-

Restricted Multiple Regression


Number of Observations






When water is included in a multivariate regression with the other independent variables,
it is no longer statistically significant. Instead, tobacco becomes significant at the 5% level and
privatehealth becomes significant at the 1% level. This result does not fit our initial hypothesis. However, the resulting significance of privatehealth would logically have a strong effect on
public expenditures when looking at the data as a random sample of 81 countries. Our findings
suggested that water access is not the most integral component of GGHE. However, this makes
sense logically as other factors in our analysis overwhelmed the effects of water. Certain variables, like privatehealth and tobacco, were significant in both the unrestricted and restricted regressions. The significance of these variables could have easily disguised the subtle yet important
effects that water has on GGHE.
Furthermore, our variable mortality is not individually significant. An F-test was conducted to determine if mortality and water were jointly significant. The F-statistic of .44 was significantly lower than the critical value at a 5% significance level. Thus, we failed to reject the null
hypothesis, which indicates the two variables are not jointly significant.
When thinking logically about water and GGHE, it is important to look at the countries
being analyzed. For this reason, we attempted to split the group of countries into developed and
non-developed countries as defined by the World Bank.7 We split our sample of 81 countries
7 Developing countries are classified as countries with a Gross National Income per capita per year of


Effects of Water Access on Government Health Care Spending

from the multivariate regression into smaller samples of developed and non-developed countries
as indicated in Appendix 1.1. To perform a better regression using a dummy variable, we added
the development dummy variable and labeled it developing. Developing countries were marked
(1) and developed countries were marked (0). We also added the slope interaction dummy variable, d1, to analyze how developing interacts with the variable water. In this new regression, we
once again found mortality to be insignificant and therefore dropped it from our multivariate
regression model.
Table 6.1 Multiple Regression: with Slope Changing Dummy Variables
Independent Variables
d1 (water*developing)
Number of Observations


The regression with a slope changing dummy variable on water has quite a different
result than the one with no distinction between the development statuses of countries. For this
model, we found water to be statistically significant. The effect the variable has on GGHE can
be described by saying there is a (.262 - .263 * developing) unit change in publichealth, when
water increases by 1 unit. That means as a developed countries’ improved water access increases by 1%, GGHE (as a proportion of GGE) also increases by .262%. However, as developing
countries’ improved water access increases by 1%, GGHE (as a proportion of GGE) decreases
by only.001%. While this result does not seem like a substantial effect in developing countries,
the remainder of the effect is captured by a shift in the intercept caused by the dummy variable
less than 11,905 USD.


The Journal of Politics & International Affairs

developing. For developing countries, the intercept is shifted upward 24.98 units, showcasing
the difference between developing and developed countries. Both the water and dummy variable
coefficients are significant at the 1% level. This statistical significance shows us that there is a
correlation between access to improved water sources and public health care spending when a
country’s development status is taken into account. Furthermore, our multivariate regression analyzes other variables, such as private health care spending and tobacco use. These variables are
also statistically significantly at a one and five percent level, respectively.
Put simply, access to improved water sources in developing countries does affect the amount of
public health spending. As water access increases, government expenditure on health decreases.
We verified our hypothesis, but only in instances where we looked at developing countries. In
developed nations, the amount of water access also affects public health spending. However, the
relationship is positively correlated, negating our hypothesis. Overall, the impact of water access
on public health spending is rather minimal. It is apparent that there are other, more influential
factors that determine the level of government expenditure on health.
It is important to note that we also performed an analysis on other variables, such as
GDP, mortality, the number of physicians in a country, private health care spending, and tobacco
use. Our analysis on GDP showed that there was a small, positive correlation between GDP and
health care spending. This correlation was statistically insignificant. Subsequently, we recognized
the importance of analyzing the number of physicians in a country. However, the data available
would have cut our sample size greatly. As noted above, mortality was also removed from our
regression due to lack of significance. Private health care spending and tobacco use showed significance in our model. Therefore, these variables also affect GGHE.
The analysis of water access on GGHE in this study showed us two distinct scenarios.
One in which a developing nation with better water access spent less public money on health.
Another scenario showed a developed nation with better water access and more government
expenditure on health. The evidence demonstrated throughout should be used to guide research
in developing nations on the improvement of water access in order to better the nations overall
health and thus lower GGHE.


Effects of Water Access on Government Health Care Spending

Appendix 1.1 List of Developing Countries
Burkina Faso
Central African Republic
Congo, Dem. Rep.
Cote d’Ivoire
Kyrgyz Republic
Papua New Guinea

Appendix 1.2 List of Developed Countries
Bosnia & Herzegovina
Marshall Islands
Czech Republic
New Zealand
South Korea
St. Lucia


The Journal of Politics & International Affairs

A., D., & R.M., S. (2005). Can water institutions be cured? A water institutions health index. Water
Science & Technology: Water Supply, 5(6), 17-40
Country List for IAWP Membership Fee Information. (2014). World Bank. Retrieved from http://www.
Eder, C., Schooley, J., Fullerton, J., & Murguia, J. (2012). Assessing impact and sustainability of health,
water, and sanitation interventions in Bolivia six years post-project. Revista Panamericana De Salud
Publica, 32(1), 43-48.
Etienne, C., Evans, D., Elovainio, R., & Humphreys, G. (2010). Where Are We Now? In The World
Health Report: Health Systems Financing: The Path to Universal Coverage. Geneva: World Health
Fewtrell, L., Kaufmann, R., Kay, D., Enanoria, W., Haller, L., & Colfordjr, J. (2004). Water, Sanitation,
And Hygiene Interventions To Reduce Diarrhoea In Less Developed Countries: A Systematic
Review And Meta-analysis. The Lancet Infectious Diseases, 42-52.
Frederiksen, Harald D. “Addressing Water Crisis In Developing Countries.” Journal Of Environmental
Engineering 131.5 (2005): 667-675. Business Source Complete. Web. 20 Oct. 2014.
Kirpich, Phillip Z. “Discussion Of “Addressing Water Crisis In Developing Countries” By Harald D.
Frederiksen.” Journal Of Environmental Engineering 132.6 (2006): 698.Business Source Complete.
Web. 20 Oct. 2014.
Peters, D., Garg, A., Bloom, G., Walker, D., Brieger, W., & Rahman, M. (2008). Poverty and Access to
Health Care in Developing Countries. Annals of the New York Academy of Sciences, 161-171.
Progress on Drinking Water and Sanitation: 2014 Update. (2014). WHO & UNICEF. Retrieved from _update.pdf
Prüss-Üstün A, Bos R, Gore F, Bartram J. Safer water, better health: costs, benefits and sustainability of
interventions to protect and promote health. World Health Organization, Geneva, 2008.
Tobacco-Related Mortality. (2014, February 6). Centers for Disease Control and Prevention. Retrieved
Walter, Elynn. “Water, Sanitation, And Hygiene: A Global Crisis With Real Solutions.”Dimensions
(1528820X) 15.1 (2013): 30. Publisher Provided Full Text Searching File. Web. 20 Oct. 2014.
Water and Sanitation. (2014). World Health Organization. Global Health Observatory. Retrieved from
Welfare Economics. (2014, January 1). University of Toronto. Retrieved from http://www.economics.
World Diarrhoeal Diseases. (2014). World Health Organization. Retrieved from
Yongsi, H. Blaise Nguendo. “Suffering for Water, Suffering from Water: Access to Drinking-water and
Associated Health Risks in Cameroon.”Journal of Health, Population and Nutrition 2010: 424.
JSTOR Journals. Web. 20 Oct. 2014.


Effects of Water Access on Government Health Care Spending


The Journal of Politics & International Affairs

Towards the Contradiction
of Legal Supremacy in the European Union
Cameron Mailhot1
Whether in terms of trade, rights, security, or employment, member states of the European Union
(EU) have grown increasingly unified and interdependent over the years. However, when examining these developments through the lens of the normative and practical values of domestic
constitutional supremacy, a theoretical and legal paradox emerges. In this paper I conduct an
analysis of independently collected legal sources and responses in order to map not only this
contradiction’s position in the EU but also the ways in which a subset of member states of the
EU have addressed it. In combining an examination of the constitutional structures, positions,
and elite responses in three EU member states—Germany, the United Kingdom (UK), and Romania—, I conclude that there has emerged a general framework of action by which these states
have attempted to approach this inconsistency by conducting themselves within the boundaries
of constitutionality and rule of law so as to fit a “legally square” peg into a “normatively round”
hole. These findings not only provide context for understanding the ways in which member states
of the EU have balanced a loyalty to constitutional supremacy with a membership in an increasingly supranational order but also lend empirical support for the observed normative value in
the supremacy of national constitutions. Moreover, this paper presents a broader understanding
of the ways in which a set of countries have addressed conflicts between normative beliefs and
concrete actions in a globalizing world, and it presents a possible qualitative framework for similar analyses in the EU and beyond.


early 70 years ago, then-Foreign-Minister Robert Schuman of France presented to his Western European counterparts a game-changing proposal: the establishment of a supranational

institution aimed at reducing interstate threats in Europe by engaging in collective action around
the coal and steel markets. The newly formed European Coal and Steel Community (ECSC) initiated the idea and framework of a united Europe. Though initially serving as a communal approach
to a common concern, over the following half century, the work of this body and its subsequent
institutions culminated in the movement of some of the most politically, militaristically, and eco1
Cameron Mailhot is a fourth-year undergraduate studying political science and German (minor)
at the University of Minnesota, Twin Cities. His major and minor fields of study are international relations
and comparative politics, respectively, with particular research interests in political conceptualizations of
human rights, global norms and governance, inter- and trans- national organizations, and European Union
politics. He would like to thank the political science department at the University of Minnesota for their
teaching and support.


Towards the Contridiction of Legal Supremacy in the EU

nomically aggressive states toward novel forms of cooperation never before seen on the war-torn
continent (“The History of the European Union”).
Forty-two years later in 1992, the 12 states of the European Communities, which developed from the ECSC, entered into an historic agreement, the Treaty of Maastricht, to form the
EU (European Community Members). The EU has, since its creation, expanded its direct vertical
influence into many more area of its member states’ domestic politics and economies. Not only
does it have a say in policies regarding migration, rights, and employment, but internal and external demands for harmonization and integration have led to increasingly binding supranational
Union law and court rulings, oftentimes requiring member states to implement or submit to decisions not necessarily supported by their politicians, officials, or their public (“EU Law Guide”).
Consequently, a noticeable transition towards legal supremacy of Union law has developed.
A robust discourse has emerged around the ways in which the EU’s legal system relates
to its member states. While some have discussed the positive implications of a supranational or
pluralist style of governing (e.g., Barber 2006, Douglas-Scott 2012, Pliakos 2007), others have
argued against it (e.g., Barents 2009, Borowski 2011). Meanwhile, various scholars have been
building a literature on the causes and effects of a loyalty stretched to two different levels of
government (Cruz 2008, Klamert 2014; Kwiecień 2005, Pernice 2002, Stiernstrom 2005), while
others have dedicated their attention to questions regarding past, present, and future complexities
in establishing a supranational constitutional structure for the Union (Alter 2003, Kumm 2005,
Stone Sweet 2009). Nevertheless, there still exist important considerations pertaining not only
to the recognition of multiple supreme legal systems and to the subsequent discord but also to a
break from norms and standards of legal ordering.
As it appears, this paper will focus on the ways in which member states of the EU have
addressed the theoretical and practical contradiction of constitutional supremacy and the supranational nature of the EU. To evaluate this situation, I examine three member states of the EU:
Germany, the United Kingdom, and Romania. While its constitution does not explicitly create a
contradiction of supremacy, Germany is nevertheless a founding member of the ECSC and has
developed a strong legal system based on a written constitution and a sense of rule of law, representing, in this regard, a model example for analysis. On the other hand, the UK operates under


The Journal of Politics & International Affairs

a longstanding “small-c” constitution2 and therefore provides a unique context for addressing the
conflicting notions of the sovereignty of Parliament and the emerging legal supremacy of the EU.
In contrast, contemporary Romania is a relatively new country, and, much like other post-communist states in the EU, it dealt with the duality of seeking EU membership while simultaneously
engaging in state-building. Consequently, it is a timely example of the shared condition faced by
many Eastern European countries still seeking EU membership.
I will begin my analysis with a brief examination of the theoretical contradictions that
have arisen between nations’ constitutional systems and the supranational legal supremacy of the
EU. I will then transition into a discussion of a set of particularly pertinent events in each of the
three member states to highlight the ways in which they have approached this paradox. Through
this process, I conclude that member states have not, in fact, departed from a sense of loyalty to
the supremacy of national constitutions, as multiple scholars have considered in their discussions
on the EU (see, e.g., Bellamy & Castiglione 1997; Elazar 1998; Moravcsik 2001; Tsebelis &
Garrett 2001). Instead, as I will contend, the norm of constitutional supremacy remains, and the
member states under consideration have taken legal and theoretical steps to maintain this value
while meeting their obligations as members of the EU.
The Emergence of a Contradiction
Within a society ordered under a constitutional system, determining ultimate legal hierarchy is a
priori a relatively simple task. When concerns for legality arise, constitutions generally provide
not only the framework for sorting out the issue but also the ground on which such judgments
are legitimized (Kay 2011, 715). However, an increase in the number and influential authority of
international institutions in the past half-century has shone light on a newly emerging phenomenon: explicit legal pluralism cutting across domestic governments and international politics.
Whether in the domain of economics, war, or sovereignty, states have begun to transfer authority
and decision-making powers to inter- and supra- national institutions (“International Organizations”). The EU is an extreme example of such a condition, but it nonetheless provides a platform
on which discussion of this emerging trend can develop. To fully understand the ways through
Small-c” constitutions, as defined here, signify the intersubjective “body of rules, practices,
[norms, customs,] and understandings, written or unwritten, that…[shapes] who holds what kind of power, under what conditions, and subject to what limits” in a given territory (Law 2010, 377).


Towards the Contridiction of Legal Supremacy in the EU

which member states interact with this issue, it is necessary to have a basic understanding of the
current situation.
At its commencement, the ECSC, the first predecessor to the EU, was nothing more than
a weak institution tasked with both distributing and monitoring resources for rebuilding Western
Europe. Requirements within the ECSC, such as notification of the High Authority for any violations of free competition, were set in an air of goodwill, binding members insofar as they freely
remained party to the community’s treaty. The treaty established a court for adjudication, but its
role was restricted to clarification and interpretation of the treaties, and individual member states
retained great influence and autonomy in appointment procedures (European Commission, c).
The member states nevertheless moved forward in their coordination, expanding the
domain into energy (European Atomic Energy Community) and trade and investment (the European Economic Community). This was met by a corresponding transfer of legal supremacy to
international (secondary) lawmaking.3 Qualified Majority Voting procedure4 expanded into new
sectors, such as immigration, security, and foreign policy, and with the ratification of the Treaty
of Amsterdam of 1997, the European Court of Justice gained even more authority to produce
legally-binding rulings when EU law was in conflict with domestic law. The result has effectively been a reduction in the legal legitimacy of the legislation, constitutions, and courts under
which member states govern themselves (European Commission, a; European Commission, b;
European Commission, d). While the large portion of legislation in the EU is passed without a
direct contradiction to member states’ domestic laws, there nevertheless remain theoretical and
empirical instances of supranational requests for the change of national law, as highlighted below
(Barker 2001, 42). As a result, it has become more possible for various bodies in the Union to
have indisputable authority over the member states’ constitutional sovereignty.
A few questions may arise at this point. Most noticeably, why does this contradiction
matter? As long as a state is able to maintain relative order within its territory, why is it important
to consider from where a “sovereign” 5 state derives its legal order? The implications are two3
Wherein “secondary” lawmaking signifies the laws, decisions, and communications formed in
the framework of bodies, institutions, or interstate agreements in the institution outside of the various
treaties themselves.
Qualified Majority Voting, as used here, signifies a form of decision-making wherein legislative
support is determined by at least two distinct forms of majority approval (such as by a majority of votes
by legislators representing at least a majority of the total constituency).
5 Note: Though a highly contested concept with its own highly nuanced field of literature, for


The Journal of Politics & International Affairs

fold. First and foremost, if states submit to internationally binding standards without consistently
and explicitly consenting, the very essence of what is rhetorically understood as the standard for
states’ independence is undermined. Over 97% of member states of the United Nations function
under the supremacy of a written constitution, so a break from this order may signal not only
a break from norms of governing domestically but also an even more serious consideration for
new forms of international political ordering. Moreover, with such a change, states are no longer
(relatively) free to justify their independence in governing and administrating themselves, instead
bowing to a supranational usurping of domestic authority. This essence of sovereignty is essentially removed from where it is constitutionally located. Therefore, the contemporary situation in
the EU is an extreme challenge to the demarcated boundaries of legal sovereignty, and it is in this
sense worthwhile to consider such implications carefully.
Arguably just as difficult to pinpoint is the second theoretical concern: the normative
value of the constitution in domestic politics. The debate surrounding the role of constitutions in
a society is not new. Scholars discuss its normative role as a source of legal authority in societies, such as in establishing rule of law in a territory, stabilizing a transitioning and/or war-torn
society, and creating and protecting some set of shared values (e.g., Ghai & Galli 2006; Epstein,
Knight, and Shvetsova 2001; and Samuels 2006). This organic law is established to provide the
highest source to which citizens, lawmakers, and courts can refer any points of contention or
justify their actions—be it with individual or governmental violations, authoritative overlapping,
or disagreements as to the constitutionality of legislation, to name a few.
Empirical observations support this theoretical argument. As Atanas Slavov (2011)
emphasizes in his study on the emergence of constitutional law in post-communist Bulgaria, the
constitution was formed to stabilize the country through the societal norm of a single source of
higher law. It provided much-needed stability and order for a country and people in transition.
Moreover, insofar as rhetoric has a practical value, direct considerations of the language in various constitutions provides further confirmation: Article VI of the U.S. Constitution, for example,
gives itself the authority as the “Supreme Law of the Land,” while the Constitution of Germany
literally translates to “basic [or ground] law” (U.S. Constitution; Constitution of the Federal
simplicity and respect for the scope of this paper, “sovereignty” is understood here as legitimate, (however conceptualized and consequently determined) “supreme power or authority” and/or “the authority of a
state to govern itself…” (“Sovereignty”).


Towards the Contridiction of Legal Supremacy in the EU

Republic of Germany of 1949).
It is also necessary to consider the practical implications at the international level. While
it may be that what is happening is simply a sharing of power among domestic and supranational
legal frameworks, this claim does not address the instances in which these two levels directly conflict with each other and compete for jurisdiction. Civilians, politicians, lawmakers, and
other actors need a clear understanding of the legal system under which they operate, and with
the increasing complexities of supranational institutions’ role in domestic law, this is even more
crucial: the ambiguity of overlapping normative and empirical legal supremacy makes it all the
more difficult to not only engage in effective lawmaking but to correctly interpret and maneuver
the legal system.
It is therefore evident that the internal relationships of the EU raise concerns as to how
member states are able to balance the role of international law and legal hierarchy with domestic
constitutions. To address this concern, I will now move to an analysis of specific case studies
of Germany, the UK, and Romania to understand, if possible, how member states in the EU
have approached this issue. In independently gathering pertinent laws, court cases, comments,
and legal reports, I examine each case study from a wide range of viewpoints. Additionally, in
choosing these three countries—each with their own constitutional styles, regional histories, and
positions in the EU—I aim to balance a consideration for a wide variance in modes of approaching this issue with a worthwhile and accurate analysis. While the decisions and courses of action highlighted below do not characterize all relationships between national constitutions and
international law, they nevertheless provide a model overview of the features of the relationship
between national authority and supranational law for EU member states.
The German Case: Steadfast yet Flexible
While the German constitution, the Basic Law, provided a new legal source of domestic governance for the (West) German state following World War II, it is also one of the more experienced
constitutions in regards to interactions with European law, providing multiple sources from
which I am able to draw an analysis. Studies of the legal relationship between Germany and the
EU often discuss two normatively important court cases: Solange I and Solange II. In both cases,
the constitutional court of Germany ruled on the role of the domestic government in the imple-


The Journal of Politics & International Affairs

mentation of EU law, but with the latter overruling the former, a shift in relations between EU
and German governing institutions developed.
In 1962, the European Court of Justice (ECJ) made a landmark decision for the European Economic Community (EEC) with Costa v ENEL. In response to recent moves by the Italian
government to nationalize companies, Italian citizen Flaminio Costa objected to his government’s overtaking of the electric company ENEL in which he had stock, arguing that consolidation violated free market aspects of the 1958 Treaty of Rome and the Constitution of Italy.
Though the Constitutional Court of Italy ruled in support of the new laws, overturning the earlier
ones that contradict the movement toward nationalization, this matter was nevertheless brought
to the ECJ. As not only a ruling on this case but also a general clarification on legal hierarchy, the
ECJ ruled that member states cannot override any agreements, laws, or decisions made under the
legality of the Treaty of Rome, thus setting EEC law above national law (Costa v ENEL). The
ECJ further strengthened this norm in the 1970 Internationale Handelsgesellschaft case ruling in
favor of EU supremacy over challenges based on national laws or precedence (Beck).
However, the German courts took a bold step in overturning this international precedent.
The German Federal Constitutional Court (FCC) ruled in Solange I of 1974 that as long as Community law did not conflict with German law, it would be treated as national law. Nevertheless,
in the case that Community law were to infringe on German citizens’ constitutional rights or freedoms, the court decided that the latter prevailed over the former (Beck). This ruling is therefore
important as it detailed how the national government viewed its relationship with the EEC. At
this time, the German courts appeared to have held a higher regard for the supremacy of domestic law over international law.
Just 12 years later, the FCC overturned its previous ruling. In Solange II of 1986, the
FCC stated that as long as fundamental rights were generally protected, the FCC no longer had
jurisdiction to rule on cases pertaining to secondary law (Beck). Of even greater importance, this
highlights a changing attitude surrounding the supranational institutional order. No longer was
such law seen in the German courts as “secondary” to national legislation. Instead, there was an
acceptance of the shift of ultimate primacy from the constitutionally legitimized courts and legislatures of Germany to international institutions created through executive treaties.
Following the ratification of the Treaty of Lisbon in 2009, the German legislature passed


Towards the Contridiction of Legal Supremacy in the EU

two particularly pertinent laws: The Act Amending the Basic Law and The Act Extending and
Strengthening the Rights of the German Federal Parliament and the German Federal Council of
States in EU Matters. The former accepted the Treaty of Lisbon by amending articles 23, 45, and
93 of the constitution to allow for EU law, while the latter restricted the framework for lawmaking of the federal legislature (Wohlfahrt 2009, 1277). In a subsequent court case, the FCC
ruled the act accepting the treaty as constitutional, while it declared the act on restriction of the
legislature unconstitutional on the grounds that it limits the fundamental right of individuals to
participate democratically in lawmaking and governing under article 38 (1)[1] of the constitution
(Wohlfahrt 2009, 1277-1278).
I can draw a few conclusions from these decisions on the relationship between the German constitution and Union law. First, constitutionality was consistently invoked in deciding
such matters. Citizens and lawmakers in Germany regularly looked to their national constitution as the first step in considering issues of international law—even when such concerns have
already been transferred to a supranational institution. The normative role of the constitution
remained in Germany.
However, from its determined legal and judicial shifts, it is safe to conclude that the German constitution is not rigid. Instead, there have been amendments to the constitution to allow
for concessions of power from the FCC to the EU/Community. Yet, while international law may
prevail in practice, it is only through the framework of the national constitution that such change
is permitted. Thus, the focal point in arguments of legal primacy remains around its constitution,
which acts as an authoritative “hinge” to not only be a source of solidarity in changing times but
also adjusts to the opening “gate” of EU/Community law in domestic law.
The Case of the UK: Steady and Reluctant
Compared to the most other member states of the EU, the UK is unique not only in that it operates under a “small-c” constitution but also in that it has generally been an outsider among European powers, remaining staunchly independent in policy, identity, and governing in the face of
pressures to integrate and harmonize with its neighbors (Bogdanor; Gifford 2008). Thus, while
the UK has encountered—and still encounters—similar debates of the influence of EU/Community law in domestic law and politics, it has ultimately followed the path towards integration of


The Journal of Politics & International Affairs

Union law. Namely, the nature of the UK’s constitutional structure has resulted in a slower—yet
arguably easier—malleability of its sources of legal authority when confronted with implementation of or submission to international law (“Doctrine of Supremacy of European Union”).

One of the more famous court cases regarding Community/Union law in the UK is Re-

gina v Secretary of State for Transport, Ex Parte Factortame Limited and Others. Beginning in
the 1970s, specific allocations of fishing quotas were given to each member state of the European
Communities. However, multiple Spanish ships gained permission to fish in British waters by the
assigning of vessel ownership to a single British citizen. In an attempt to combat what it saw as
unfair claiming of British fishing quotas by foreign businessmen, the British Parliament passed
the Merchant Shipping Act 1988, which required that fishing companies wishing to pull from the
British quota have a majority of their owners holding British citizenship. Spain challenged this
law in UK courts on grounds of discriminatory practices. The House of Lords eventually upheld
the Court of Appeals’ ruling in favor of parliamentary supremacy, claiming that there “was no
power to grant an injunction against the Crown to suspend the application of an Act of Parliament” (Stiernstrom 2005, 9). Nevertheless, a reference was made to the ECJ in the ruling, and
more importantly, the court’s decision noted that any subsequent Community law or ECJ ruling
could overturn such an obstacle on the grounds that the UK, in joining the Community, voluntarily agreed to accept any limit of “sovereignty”— no matter if explicitly expressed in the treaties
or established through jurisprudence. In its ruling, the ECJ found that in such a case pertaining
to Community law, if the only obstacle to granting interim relief is a UK law, the national government must set aside such a statute. With a request for a preliminary ruling of the 1988 act’s
compatibility with Community law already in place, the House of Lords was required to—and
did—revoke the regulation (Stiernstrom 2005, 8-10; “Cases – EC Law;” Regina v Secretary of
State for Transport, Ex Parte Factortame Limited and Others).

With a similar attitude set forth by fellow MP Lord Bridge, the Supreme Court of the UK

upheld the emerging norm of EU primacy in R v The Secretary of State for Transport and another. The issue raised in this case was whether the proposed “HS2” rail system would meet requirements set forth by an EU directive in 2011 on facilitating public participation. Most notable from
the case is that the court recognized that the UK does not have a written constitution but rather
a set of constitutional instruments. Taking into consideration the decision in the Factortame case


Towards the Contridiction of Legal Supremacy in the EU

and texts surrounding European integration, Lord Neuberger and Lord Mance led a third judgment on the case that certain Union law is not superior over domestic law but rather a part of
domestic law. The court ruled in favor of such claims, arguing that the focus of the court case,
the European Communities Act of 1972, may be added to the “body” of constitutional texts of
the UK (Mark Elliott 2014; Hart, QC; R v The Secretary of State for Transport and another).
These examples show a change in the framework for discussion surrounding what comprises the constitution of the UK. Not unlike in Germany, these few important moments in UK
constitutional history highlight the continued pertinence of the constitution, effectively arguing
that it can and sometimes should be composed of legislation, agreements, and treaties formed
outside of the domestic legislative process. Like the distinction between constitutional “instruments” of the UK and general legislation, specific “instruments” of international institutions can
also reach “constitution” status in the UK.
Nevertheless, the discussion has been framed in the sense of maintaining a constitution
at the national level. Evidently, scholars and lawmakers still consider discussion of the unwritten, “small-c” constitution important in the UK, consistently judging it against EU/Community
law and vice versa. However, this has not resulted in a decrease in its normative value. Instead,
the movement of the discussion is to how such a constitution is defined, with the results of court
cases and decisions opting for an expansion to include possible external sources of law in the
constitution. As such, the constitution in the UK—however defined—is still relevant and of high
The Romanian Case: Legal Supremacy of the Constitution
Much like its other post-communist neighbors, Romania was met with simultaneous demands
to join the EU and to construct a national revolution. Through international intervention and in
efforts to integrate and harmonize with other states on the continent and the international community, its constitution includes various provisions respecting many regional and international
norms, institutions, and laws (Vereshchetin 1996, 30-32). In doing so, it has forced the national
government to address legal conflicts on several occasions, two of which are outlined here.
As a newly independent country, Romania made early strides in an effort to harmonize
with the Union’s preexisting policies, laws, and regulations. This is evident in Article 11 of


The Journal of Politics & International Affairs

the constitution, which states that any treaties ratified by the government are thereafter legally
binding under national law (Bondoc 2001, 33; Constitution of Romania of 1991). The public’s
position on this provision was unwavering under a national referendum: a year prior to accession
to the Union, 80% of the voting public supported an amendment to bring the legal precedence of
EU law under the Romanian constitution. In that same year, the Constitutional Court of Romania’s Decision No. 148 confirmed “the constitutionality of the proposal [of transfer of powers] by
arguing…that the sharing of the exercise of sovereign attributes with other states does not confer
on the EU ‘competence over competence…” (Hoffmeister 2007, 11). It ruled that the delegation of powers to another body was constitutional, as it is a relegation of lawmaking through the
constitution’s own institutional structure, thus making such a body reliant on, subservient to,
and secondary to the national government. In this sense, a feeling of constitutional supremacy is
maintained (Hoffmeister 2007, 11-12).
Additionally, the Romanian Constitution explicitly outlines the legal power of Union law.
Under Article 148 of the constitution, EU law has supremacy over national legislation. However,
similar to the 2006 court ruling on constitutionality regarding external law- and regulation-making body, Article 147(3) of the constitution also places international treaties secondary to the constitution. According to subsection 3, the Constitutional Court partakes in constant judicial review,
subjugating each treaty or agreement to review for constitutionality before being ratified. This
suggests that, once an international agreement is supported (including a treaty on accession to the
EU), any subsequently binding law, regulations, or directives are assumed to be constitutional
in Romania (Constitution of Romania of 1991; Hoffmeister 2007, 20-21). This therefore places,
as Hoffmeister (2007, 21) notes, EU law “somewhere between constitutional and statutory law.”
Only in becoming subservient to the constitution is it granted the right to supremacy.
Thus, as is the case in Germany and the UK, the Romanian constitution ultimately retains legal supremacy. In the same regard as the UK, Union law is permitted, under Romania’s
constitution’s delegation, to take precedence. It is through the sovereign authority of the constitution that supremacy of Union law is derived. Instead of grappling with supranational legislation
post-constitutional construction, the Romanian legal system has made special provisions allowing for the creation of a gap through which Union law can play a role within the country, thus
maintaining the value of the constitution domestically while meeting obligations internationally.


Towards the Contridiction of Legal Supremacy in the EU

Lessons & Conclusions
I can confidently derive two conclusions from these three case studies. First and foremost, in
contrast to what may appear on the surface, the role and precedence of the constitution—whether
written or unwritten—has remained undiminished in these member states. While there have been
concerns of superiority, authority transfer, and the placement of international law in national legislation, there remains a consistent invocation of national constitutionality. Unlike the aforementioned arguments focused on the formation or recognition of a supranational European constitution, these three model countries within the Union still hold a value in maintaining some sense
of domestic constitutional supremacy. As a result, this translates to a perception of EU law as an
external actor that must meet the varying contextual demands of each state under consideration.

Second, there nevertheless exists a general pattern of how these three member states have

molded the two levels of legal order together. All three countries have—be it in legal maneuvering, power relinquishing, or enlargement—adjusted or solidified their own national governments
to include or permit the legitimation of Union law. However, this does not weaken the argument.
Instead, it is safe to conclude that this occurs in an air of constitutionality. Instead of treating
each individual Union law as something purely external and supranational to the constitution,
these states have begun to consider such legislation as part of or subservient to their national
constitutions. When EU member states interact with Union law, it is through the authorization of
each individual state’s constitution, and when Union laws or rulings reach past national legislation, it is again because of the authorization of each states’ constitution. Consequently, legality
and supremacy is retained at the national level.
--In this paper, I have been able to provide a brief documentation and analysis of the
important aspects of the EU and its predecessors as they relate to the constitutions of Germany,
the UK, and Romania. As outlined before, these member states hold a beneficially wide range of
unique characteristics of constitutionality and statehood, allowing for wider applicability of this
paper’s findings. Thus, I am able to conclude through my research that these member states are,
in fact, willing to address the paradox of overlapping legal pluralism in the EU through various
manners—of which all dovetail at the maintenance of the theoretical and legal supremacy of the
national constitution. Be it through consistent invocation of constitutionality to the expansion of


The Journal of Politics & International Affairs

the “constitution” to the delegation of separate powers, the normative and theoretical value of a
constitution is maintained in these three states.
Consequently, it is possible to consider the broader applicability of my findings. In
developing a better understanding of how states have similarly approached the paradox of legal
pluralism in the supranational arena with norms of constitutional supremacy, this paper contributes further empirical support for the observed diffusion and role of national constitutions, and it
provides a possible framework for similar analyses in the EU and beyond. Nevertheless, as the
EU and the rest of the increasingly interdependent world continue to evolve, additional research
in a similar vein is necessary to further conceptualize variances and document how states can
and do approach these issues. Such work will present a better understanding of these conditions
as they continue to unfold, and it will necessarily provide improved applicability, guidance, and
support for theoretical and practical considerations in the future.


Towards the Contridiction of Legal Supremacy in the EU

Alter, Karen J. Establishing the Supremacy of European Union Law: The Making of an International Rule
of Law in Europe. Oxford University Press, 2001.
Barber, N.W. 2006. “Legal Pluralism and the European Union.” European Law Journal 12(3): 306-329.
Barents, René. 2009. “The Precedence of EU Law from the Perspective of Constitutional Pluralism.”
European Constitutional Law Review 5(3): 421-446.
Barker, J. Craig. 2001. International Law and International Relations (International Law for the 21st
Century). London: Bloomsbury Academic Publishing, 2001.
Beck, Gunnar. 2011. “The German Constitutional Court versus the EU: self-assertion in theory and
submission in practice—Euro Aid and Financial Guarantees. Part 1..” Eutopia Law, 24 October
2011. (14 March 2015).
Bellamy, Richard & Dario Castiglione. 1997. “Building the Union: The Nature of Sovereignty and the
Political Architecture of Europe.” Law and Philosophy 16(4): 421-445.
Bogdanor, Vernon. 2012. “British Euroscepticism.” Social Europe. 2012/04/
british-euroscepticism/ (March 2015).
Bondoc, Lucian. 2001. “European Integration and Constitutional Law: The Situation in Romania.” In
European Integration and Constitutional Law, ed. the European Commission: 31-38.
Borowski, Martin. 2011. “Legal Pluralism in the European Union.” In Law and Democracy in Neil
MacCormick’s Legal and Political Theory: The Post-Sovereign Collection, ed. Agustín José
Menéndez and John Erik Fossum. Springer: 185-209.
“Cases – EC Law.” 2015. Sixth Form Law. cases_ec_law.htm.
(15 March 2015).
Central Intelligence Agency. 2014. Constitution. In The World Facebook. /library/
publications/the-world-factbook/fields/2063.html (21 March 2015).
Constitution of the Federal Republic of Germany of 1949. 1949.
Constitution of Romania of 1991. 1991.
Costa v. ENEL. 1964. 6/64 European Court of Justice.
Cruz, Julio Baquero. 2008 “The Legacy of the Maastricht Judgment and the Pluralist Movement.”
European Law Journal 14: 389-422.
“Doctrine of Supremacy of European Union.” http://www.
administrativelaw/doctrine-of-supremacy-of-european-union-administrative-lawessay. php#ftn13
(15 March 2015).
Douglas-Scott, Sionaidh. 2012. “Justice and Pluralism in the EU.” Current Legal Problems 65: 83-118.
Elazar, Daniel J. 1998. “The New Europe: a Federal State or a Confederation of States?” Swiss Political
Science Review 4(4) (1998): 119-138.
Elliott, Mark. “Reflections on the HS2 case: a hierarchy of domestic constitutional norms and the
qualified primacy of EU law.” 23 January 2014. 01/23/markelliot-reflections-on-the-hs2-case-a-hierarchy-of-domestic-constitutional-norms-and-the-qualifiedprimacy-of-eu-law/ (15 March 2015).
Epstein, Lee, Jack Knight, & Olba Shvetsova. 2001. “The Role of Constitutional Courts in the
Establishment and Maintenance of Democratic Systems of Government.” Law & Society Review
35(1): 117-164.
“EU Law Guide.” (3 April 2015).


The Journal of Politics & International Affairs

European Commission. “Court of Justice of the European Union.” (3 April 2015).
European Commission. “Monitoring application of Union law.” applying-eulaw/index_en.htm (10 March 2015).
European Commission. “Treaty Establishing the European Coal and Steel Community, ECSC Treaty.” affairs/treaties/treaties_ecsc_en.htm (11 March
European Commission. “Supranational Decision-Making Procedures.” http://www.europarl.
ftu/pdf/en/FTU_1.4.1.pdf (11 March 2015).
European Community Members. 1992. Treaty on European Union. 7 February.
“Ghai, Yash & Guido Galli. 2006. “Constitution-building Process and Democratization: Lessons
Learned.” International Institute for Democracy and Electoral Assistance.
Gifford, Chris. 2008. The Making of Eurosceptic Britain: Identity and Economics in a Post-Imperial
State. London: Ashgate Publishing, 2008.
Hart, David, QC. “Highspeed rail, Parliament, and the EU Courts.” January 2014. http://ukhuman (15 March 2015).
“The History of the European Union: The European Citizenship.” 2003. http://www.historiasiglo
europe/anteceden2.htm (1 March 2015).
Hoffmeister, Frank. 2007. “Constitutional Implications of EU Membership: A View from the
Commission.” Croatian Yearbook of European Law and Policy 3: 59-97.
“International Organizations.” fotw/flags/int.html (22 March 2015).
Kay, Richard S. 2011. “Constituent Authority.” The American Journal of Comparative Law 59(3): 715761.
Klamert, Marus. 2014. The Principle of Loyalty in EU Law. Oxford: Oxford University Press.
Kumm, Matthias. 2005. “The Jurisprudence of Constitutional Conflict: Constitutional Supremacy in
Europe before and after the Constitutional Treaty.” European Law Journal 11(3): 262-307.
Kwiecień, Roman. 2005. “The Primacy of European Union Law over National Law Under the
Constitutional Treaty.” German Law Journal 6(11): 1479-1496.
Law, David S. 2010. “Constitutions.” In The Oxford Handbook of Empirical Legal Research, eds. Peter
Crane and Herbert Kritzer. Oxford: Oxford University Press: 376-398.
Moravcsik, Andrew. 2001. “Federalism in the European Union: Rhetoric and Reality.” In The Federal
Vision: Legitimacy and Levels of Governance in the US and the EU, eds. Kalypso Nicolaïdis and
Robert Howse. Oxford: Oxford University Press: 161-187.
Pernice, Ingolf. 2002. “Multilevel Constitutionalism in the European Union.” European Law Review 8:
Pliakos, Asteris. 2007. “National Parliaments and the European Union: Necessity of Assigning a
Supranational Role.” European Review of Public Law 3(65): 757-785.
R (on the application of HS2 Action Alliance Limited) (Appellant) v The Secretary of State for Transport
and another (Respondents). 2014. 3 UKSC 2014. Supreme Court of the United Kingdom.
Regina v Secretary of State for Transport, Ex Parte Factortame Limited and Others. 1999. House of
Samuels, Kirsti. 2006. “Post-Conflict Peace-Building and Constitution Making.” Chicago Journal of
International Law 6(2): 663-682.
Slavov, Atanas. 2011. “Challenges to Constitutional Supremacy in a New Democracy: A Critical Study of
Bulgaria.” CAS Sofia Working Paper Series 4: 1-30.


Towards the Contridiction of Legal Supremacy in the EU

“Sovereignty.” Def. 1 and 1.1. Oxford Dictionaries, Oxford Dictionary Online, n.d. 28 April 2015.
Stiernstrom, Martin. 2005. “The Relationship Between Community Law and National Law.” Jean
Monnet/Robert Schuman Paper Series 5(33): 1-13.
Stone Sweet, Alec. 2009. “Constitutionalism, Legal Pluralism, and International Regimes.” Faculty
Scholarship Series. Paper 1295.
Tsebelis, George & Geoffrey Garrett. 2001. “The Institutional Foundations of Intergovernmentalism and
Supranationalism in the European Union.” International Organization 55(2): 357-390.
U.S. Constitution. 1789.
Vereshchetin, Vladlen S. 1996. “New Constitutions and the Old Problem of the Relationship between
International Law and National Law.” European Journal of International Law 7(1): 29-41.
Wohlfahrt, Christian. 2009. “The Lisbon Case: A Critical Summary.” German Law Journal 10(8): 12771286.


The Journal of Politics & International Affairs

French Aid for a French France
Matthew Cullom1 & Andrea Moneton2
Does immigration to France affect the allocation of French foreign aid? We hypothesize that France
sends more foreign aid to countries that are large sources of immigration in order to stem the tide of migrants. In addition, we expect that as a country sends more immigrants to France, the French government
responds by allocating more aid to that country. Using immigration data from the Institut National des
Études Démographiques and foreign aid data from the OECD, we find that bilateral immigration flows
have less of an effect on aid allocation than we predict. Our research suggests that while there is no clear
causal relationship between bilateral aid and migration in the French case, the French government does
send significantly more foreign aid to countries that send France more immigrants. However, as more
immigrants enter France from a specific country of origin, France does not respond by allocating that
country more aid. Thus, our research suggests a more nuanced relationship between aid and migration.


oes immigration to France affect France’s allocation of foreign aid? While some scholars
have looked at the relationship between bilateral foreign aid and immigration, we look to

focus exclusively on France in order to discover if the French government allocates larger amounts
of Official Development Assistance (ODA) to countries that are larger sources of immigrants. We
concentrate on France due to its well-known immigration problems that have caused French public
opinion to gradually become more critical towards immigrants over the last half-century (IPSOS,
2011: 5). Because popular opinion often influences both public and foreign policy, we investigate
whether domestic anti-immigration sentiment might affect French foreign aid priorities.
Many scholars study the connection between immigration and foreign aid and, in some
cases, suspect that any increase in foreign aid to a country would act as disincentive for that
country’s citizens to emigrate (Berthélemy 2002). Consequently, we hypothesize that the French
Matthew Cullom is a second-year undergraduate studying at Georgetown University’s School of
Foreign Service. He is from Virginia and is studying International Economics with a focus on international development.
Andrea Moneton is also a second-year undergraduate studying at Georgetown University’s
School of Foreign Service. She is a French-American student studying International Politics with a focus
on the development of international institutions.

Mathew and Andrea would like to thank Professor James Vreeland for facilitating the project by
providing us with data on foreign aid as well as for his guidance and inexhaustible patience. Finally, they
hope their research will prove especially relevant in light of the recent refugee crisis in Europe and the
Middle East. Their thoughts are with all those who have been displaced by the conflict in Syria.


French Aid for a French France

government, faced with rising anti-immigration sentiment, would increase foreign aid to countries that are large sources of immigrants in order to stem immigration and please its constituents.
If this hypothesis holds, it would seem the French government actively wants to decrease immigration and view foreign aid as an effective tool to do so.
Foreign aid’s effectiveness at reducing immigration is presumably tied to its ability to
promote development and, in turn, reduce push factors that drive people to emigrate. The effectiveness of aid with regards to economic development has been widely debated. Some scholars
claim that aid may not always end up in the hands of those in need and doubt its effectiveness at
ameliorating the quality of life in recipient counties (Boone 1996; Burnside and Dollar 2000).
Thus, the effect of aid on immigration might be weaker than we predict. France might not see
ODA as an effective tool to reduce inflows of migrants, and so immigration flows to France
might not significantly alter the distribution of French ODA. Moreover, if we do find a correlation, we must still address potential issues of endogeneity.
These challenges, however, have not thwarted attempts to research the effect of immigration on foreign aid disbursements. Azam and Berlinschi (2010) produce the most comprehensive
study and find that many international organizations do in fact account for optimal migration
flows when distributing multilateral foreign aid (Azam and Berlinschi 2010). However, because
France awards 65 percent of ODA through bilateral channels, we choose to focus instead on
bilateral aid flows. (French Ministry of Foreign Affairs and International Development 2013;
hereafter, France Diplomatie) In this regard, we enter uncharted territory as no previous studies
focus exclusively on immigration’s effect on the disbursement of French ODA.
This may be due to the fact that French immigration records are spotty. A 1978 French
law forbids censuses from requesting information on race and ethnicity, and French policies in
general limit acknowledgement of the multi-ethnic nature of the nation’s population (Bleich
2001). This consequently limits the development of large comprehensive datasets on immigration. We use a relatively small dataset from Institut National des Études Démographiques (2013;
hereafter INED) that provides 360 country-year observations from the largest sources of immigrants to France from 1994 to 2008. We assume negligible immigration from other countries not
accounted for in INED’s dataset, which, while increasing observations to 585, may produce some
systematic error in our study.


The Journal of Politics & International Affairs

We gather country-year data on our dependent variable from the Organization for Economic Cooperation and Development (2012; hereafter OECD) records on bilateral foreign aid
flows as used by Vreeland and Dreher (2014). This comprehensive dataset does not reduce our
number of observations.
Descriptive statistics suggest a positive correlation between our two variables – larger
sources of immigration receive significantly more French aid. After controlling for factors known
to affect the allocation of ODA such as real gross domestic product (GDP) per capita, polity, pariah status, language, and ex-colony status, ordinary least squares (OLS) regression corroborates
descriptive data. Yet, country fixed-effects regression does not maintain statistical significance,
implying that as a country sends more immigrants to France, the French government does not
respond by allocating that country more ODA.
The paper proceeds as follows. Section 2 examines France’s unique history of immigration policy and official foreign aid priorities. Section 3 then analyzes literature on the determinants and effects of both immigration and foreign aid flows and relates these studies to research
connecting these variables. Section 4 describes the nature of our data, while section 5 details our
methodology and results. Finally, we conclude in section 6 by discussing implications of our
findings and avenues for further research.
Background French Aid and a French France
France has a long and controversial history with immigration issues. Our background elaborates
on immigration in France in the second half of the 20th century so our reader may better understand the reasoning behind our research question and hypothesis.
The controversy over immigration to France has peaked over the course of the past few
decades, specifically since the 1970’s. France enjoyed a period of economic prosperity traditionally labeled Les Trente Glorieuses, or the “Glorious Thirty,” from the end of World War II until
the oil crisis and subsequent economic downturn of 1973. During this period, France’s industrial
sector flourished, creating a strong pull factor for potential immigrants, especially for those living
in Maghrebin ex-colonies like Morocco and Algeria.
Because WWII decimated France’s male, working age population, the French government instituted a lax immigration policy allowing most immigrants to work and reside in France.


French Aid for a French France

Towards the beginning of the 1970’s, however, the economy began to suffer and the French population became less receptive to immigrants. Even before the 1973 recession, more than half of
France’s population believed that the best way to fight unemployment was to limit immigration
(Blanc-Chaléard 2001, p. 71). In response, the French government reformed immigration policy, and enacted much stricter laws. Following the recession, France halted established programs
meant to recruit foreign labor and made it increasingly risky for companies to hire illegal immigrants (Hamilton 2004).
The evolution of public opinion engendered the rise of the far-right party, Le Front National (FN) and its controversial leader Jean-Marie Le Pen. Throughout the 1970’s the FN never
received more than one percent of the popular vote in national elections, but its popularity steadily rose throughout the 1980’s and 1990’s (Marthaler 2008). While not a single-issue party, the
FN and its members care most about immigration. Its heavy-handed stance on the issue attracts
most of its members (Carvalho 2014, p. 99). Over the course of our study, from 1994 to 2008,
at least 90% of FN voters thought there were too many immigrants in France at any given time
(Marthaler 2008).
Not only did many French citizens blame immigrants for the struggling economy, but
they also accused these “outsiders” of diluting France’s cultural exceptionalism. Jean-Marie Le
Pen referred to immigration as an “invasion” which “dilutes the…identity of our people.” (Le
Pen 2001) Indeed, a significant portion of the French population perceives multiculturalism as
a threat to national culture. In 2002, Jean-Marie Le Pen reached the second and final round of
French elections for the first time in the FN’s history. During this time, half of all voters considered Le Pen to be most competent to deal with immigration levels, which had risen from 150,000
in 2000 annually to 200,000 by 2003 (Carvalho 2014, p. 98).
At the turn of the century, France once again modified its immigration policy. In 2003,
then-Minister of the Interior and future president Nicolas Sarkozy spearheaded a law on immigration that would not only stem the flow of immigrants, legal and illegal, but also help integrate those already on French territory. In 2007, he proposed a new law requiring immigrants
attempting to reunite with family members (a loophole around immigration restrictions) to take
DNA tests proving their familial ties. Seventy-five percent of the French population believed
this initiative would be effective in combatting illegal immigration (Marthaler 2008). This policy


The Journal of Politics & International Affairs

change prompted us to investigate into whether France might be using other tools, specifically
foreign aid, to decrease immigration.
Today the FN Program advocates even stricter immigration laws: the party wishes to
regain control of France’s borders by reconsidering France’s place in the Schengen Zone, inside
which all people circulate freely. (Front National 2014) The party also opposes the aforementioned loophole, which allows an immigrant already in France to reunite with his or her (immediate?) family members. It is clear, however, that this is more an issue of culture than of economics for the FN, and that the root of the problem is no longer an economic crisis but an evolving
mentality of intolerance.
Given the importance of the immigration debate in France, we question whether the
government, in addition to using immigration laws, employs foreign aid as another tool to reduce inflows of unwanted immigrants. According to France Diplomatie, foreign aid represents
0.45% of France’s GDP, a high proportion compared to the global average of 0.31%. In addition,
France contributes approximately 10% of total global ODA, awarding €10 million of foreign
aid in 2008. Because France is a major foreign aid donor, it is important for the public both in
France and in recipient countries to understand what determines the government’s aid allocations. France Diplomatie (2013) officially cites the determinants of its ODA disbursements as the
1. Promoting peace, stability, human rights and gender equality
2. Equity, social justice and human development
3. Sustainable, job-rich economic development

4. Protecting the environment and global public goods
The French government has also pledged to meet the eight Millennium Development Goals
established by the United Nations Development Programme. Most of these relate to health, education, poverty, and environmental stability and fall under the umbrella of France’s four national
goals. Specifically, the French government explicitly states its priorities lie in Africa, a pledge
corroborated by data. Indeed, 55% of French ODA goes to African states (with a total of 41%

French Aid for a French France

going to sub-Saharan Africa). The government labels 16 African countries “poor priority countries.” For these countries, France adds one interesting goal that is not included in the Millennium Development Goals: promotion of democracy. Many of these countries are ex-colonies, have
Muslim-majority populations, or have adopted French as their official language. Some exhibit
multiple or all of these traits. These observations and existing literature (Rioux 2005) lead us to
consider these factors as control variables during data analysis.
Beyond official priorities, France may also consider strategic interests when allocating ODA.
Many scholars have found statistically significant relationships between ODA and country specific strategic considerations. The following section explores many traditional determinants of
foreign aid while also analyzing past studies connecting immigration with ODA.
Determinants of French Foreign Aid: Is Immigration a Factor
Why would immigration affect the allocation of foreign aid?
We arrive at our hypothesis by making four main assumptions:
1. People from poor developing countries are most likely to immigrate to rich countries. The
larger the gap in living conditions, the more incentive a citizen of the developing nation
has to migrate.
2. Foreign aid, in theory, improves the standard of living in recipient countries. Accordingly,
any inflow in foreign aid would reduce the marginal benefit of immigration and the incentive to immigrate by reducing push factors.
3. High-levels of immigration impose significant perceived and sometimes material costs on
host countries’ populations, so these countries aim to restrict net immigration.
4. Developed countries allocate ODA with their own strategic interests in mind.

Given these assumptions, donor countries facing unwanted immigration would naturally
allocate foreign aid strategically to lower immigration levels by donating more ODA to countries
that send the most immigrants. Considering France’s history and traditional anti-immigration
sentiment, we hypothesize that the French government would use the same technique when

The Journal of Politics & International Affairs

dispersing foreign aid and award larger quantities of foreign aid to countries that send France the
most migrants.
Significant literature pertains each point. First, we look at the determinants of immigration. Scholars have extensively researched determinants of immigration. Mayda (2009) states
that real GDP per capita of the destination country is a major pull factor and is positively correlated with high emigration rates for lower income countries, thus corroborating our first assumption.
In order to affect this major determinant of immigration, ODA must promote development (and thus real GDP per capita) in practice, rather than theory. Current research challenges
the theory that aid stimulates economic development. Burnside and Dollar (2000) find that aid
only promotes economic growth in countries that implement sound public policy. On the other
hand, Boone (1996) finds that “aid does not significantly increase investment, nor benefit the
poor as measured by improvements in human development indicators.” In addition, Djankov,
Montalvo, and Quernol (2006) claim that empirical evidence does not support the development
hypothesis. These studies produce troubling implications regarding immigration, for if ODA does
not benefit citizens in recipient countries, then it might not reduce the push factors that drive
them to emigrate.
Despite this potential challenge to our research, we move on to our third assumption and
consider the effects of immigration on developed nations like France in order to see if anti-immigration policies would be desirable. Conflicting studies plague the debate. On one hand, Borjas
(2003) claims that increased immigration produces considerable negative economic effects on a
nation’s domestic workforce. On the other hand, Ortega and Peri (2009) argue that immigration
actually increases employment without “crowding out natives.” Yet even if Ortega’s study holds
true, immigrants are easy scapegoats for economic troubles and can destabilize a formerly cohesive society. Fetzer (2010) illustrates this problem and poses the following question: “Could one
not also consciously or unconsciously rationalize one’s pre-existing nativism by claiming that
foreign workers are competing with natives for jobs?” As we saw in the background study, this
type of nativism is especially prevalent in France. The FN and much of the French population
accuse immigrants of contributing to the country’s economic and social decline. Thus, one must
recognize the multi-faceted issues immigration presents. Governments must not only weigh eco-


French Aid for a French France

nomic consequences of immigration, but also its impact on cultural identity and social stability.
Finally, if the French government has legitimate reasons to restrict the entry of migrants,
it might reasonably alter its allocation of foreign aid to decrease immigration. In order for this to
hold true, governments must strategically distribute aid as we assume above. Extensive scholarly
research shows donors’ strategic considerations do, in fact, affect the allocation of ODA. Maizels
and Nissanke (1984) assert bilateral aid flows fit best with the donor-interest model. Hoeffler
(2008) quantifies the effect of donor interest by measuring trade patterns and UN voting patterns
and finds it accounts for 12% of the variance in aid. On the other hand, recipient need, measured
by GDP per capita, and recipient merit, measured by economic growth, polity, and human rights
account only for a mere 2% and 1% of aid variance respectively.
When considering determinants specific to French foreign aid, researchers find that countries with economic and cultural ties to France receive significantly more aid. Rioux (2005) finds
that France allocates more ODA to countries whose official language is French, and Hoeffler
(2008) discovers the same effect for ex-colonies. Similarly, Schraeder, Hook, and Taylor (1998)
find that France rewards relatively large trading partners, many of whom are French speaking
ex-colonies, with significantly greater aid. This literature corroborates our fourth assumption on
foreign aid, and leads us to test whether, in addition to these economic and cultural variables,
France might also consider immigration when allocating ODA.
While we hypothesize that ODA reduces emigration from recipient countries due to its
positive impact on development, scholarly research reveals a more nuanced relationship. Berthélemy (2009) claims that for countries below a certain GDP per capita level, bilateral aid actually
complements bilateral migration and aid becomes a substitute when a country crosses a certain
threshold of income, dis-incentivizing further emigration. Furthermore, Gayton-Fregoso and
Lahiri (2000) argue that aid, when given in small quantities, also marginally increases illegal
Still, Azam and Berlinschi (2010) find that overall there exists an empirical trade-off between aid and migration, finding a significant correlation between the two variables as illustrated
in Figure 1.
The researchers argue that donors optimize flows of aid to significantly reduce inflows of
immigrants. In theory, this would require much collusion between various actors. Accordingly,


The Journal of Politics & International Affairs

they even suggest that donor countries employ multi-lateral institutions such as the European
Union or United Nations to mask this activity. Although this study concentrates on multilateral
aid flows, it corroborates our hypothesis by suggesting that donors do in fact take into account
immigration flows when distributing aid. It also implies that France might allocate aid via the EU
and UN to hide a potentially anti-immigration foreign aid policy. Vandercasteelen (2012) builds
on Azam and Berlinschi (2010) to show that this correlation between aid and immigration holds,
but only when considering bilateral aid flows. He finds this correlation to be robust, even when
accounting for country fixed effects. Vandercasteelen offers two potential explanations for this
correlation: First, donor countries might use bilateral aid as an anti-immigration tool. Second,
immigrants, once they establish a community in a destination country, successfully lobby the
government to provide more foreign aid for their home countries, an argument also present in
Bermeo and LeBlang’s work (2009).

Figure 1: Immigration as a function of ODA

Figure 1 shows a significant correlation between the number of immigrants a recipient country
sends to a donor, and the amount of aid it receives.
We look to contribute to existing research, while concentrating on specifically French
bilateral aid flows. We hope to corroborate prior research by Azam and Berlinschi (2010) and


French Aid for a French France

Vandercastleen (2012) by giving a concrete example of a specific country that takes into account
bilateral migration flows in their allocation of bilateral foreign aid.

Our project relies fully on the analysis of quantitative data. To measure our dependent variable,
we use data on French ODA originally collected in 2012 by the OECD as used by Vreeland and
Dreher in The Political Economy of the United Nations Security Council (2014). This dataset
contains 4,743 country-years of data on French foreign aid as measured in millions of US dollars,
from 1951 to 2009. Sixty five percent of this aid was awarded through bilateral channels, which
allow France to pursue its own agenda; this explains our decision to focus on bilateral flows.
Moreover, the OECD’s dataset provides thousands of country-year observations on potential confounding variables such as polity, GDP per capita, and pariah among others.
For our independent variable, volume of immigration to France, we draw from a dataset
collected by the Institut National des Etudes Démographiques (INED), the French government
institution charged with studying the evolution of French demographics. The dataset contains
375 observations ranging from 1994 to 2008. Inopportunely, INED combines some groups of
immigrants from countries in close proximity to each other into groups such as “ex-Yugoslavia”
or “Sub-Saharan Africa,” thus eliminating the countries included in those groups from our study.
We end up with a small panel of data with 360 country-year observations. Each observation
represents the annual number of immigrants entering France from a specific country of origin.
Overall, the immigration data limits the span of our research because of its lack of detail and
narrow time frame.
Many countries receive French ODA, but do not send enough immigrants to be included in INED’s dataset. Because we consider INED’s dataset to be reliable, we assume negligible
immigration from countries that are not included in it. We assume the amount of immigrants
entering France from these countries is zero, with a few exceptions. The dataset groups together
17 countries labeled “Sub-Saharan Africa,” and gives the individual data for 6 of those countries.
It would be wrong to assume that the other 11 countries send no immigrants to France over the
span of our panel, so we label these countries as missing in the dataset. While all of this more
than doubles our number of observations to 585, it undoubtedly produces some systematic error.


The Journal of Politics & International Affairs

In addition, we choose to ignore flows of illegal immigrants – another potential source
of bias. Although these immigrants undoubtedly exist in relatively large numbers, their flows are
difficult to measure. As a result, we narrow our study to focus on legal, documented immigration
Figure 2 depicts our data. It shows that in country-years with high levels of immigration
(over 8,000 immigrants) France allocates more ODA – a finding that supports our initial hypothesis.
Figure 2

Figure 2 shows the distribution of ODA to countries that send, high, medium, and low
amounts of immigrants to France. As shown in Figure 2, for years during which a country sends
over 8,000 immigrants to France, the French government provides an average of over 250 million dollars in ODA. For years during which a country sends 4,000 to 8,000 immigrants those
countries receive an average of just over 150 million dollars of ODA. The same is true for years
during which a country sends less than 4,000 immigrants. This discrepancy suggests that France
allocates more ODA to large source-countries.


French Aid for a French France

Regression Analysis with Control Variables
To further examine the relationship between immigration to France and its disbursement of
ODA, we turn to regression analysis. Initially, bivariate regression results support our hypothesis:
the French government allocates more aid to countries that send more immigrants to France. This
correlation is statistically significant at the 1 percent level, as indicated by Table 1.
Table 1.



But, is this positive correlation robust? As discussed in the literature review, scholars

have identified many socio-economic variables that affect the disbursement of ODA. To ensure our results are not spurious, we control for many potential confounding variables: polity
(regime-type), inter-state war (defined as a conflict with over 1000 deaths), language, religion,
pariah status (defined as a country under international sanctions), GDP per capita, population,
and US military assistance.
Alesina (2000) finds that a country’s level of democracy significantly affects the allocation of ODA, and that democracies receive more aid on the margin. To control for this factor, we
use the polity2 score, which assigns countries a score from -10 to 10. Countries that receive a 10
are the most democratic; countries that receive a -10 are the most totalitarian.
We use two dichotomous variables to account for pariah status and interstate war as used
in Vreeland and Dreher (2014). We define pariah states as countries under international sanction
in a given year. So, if a country is under sanction in a specific year we assign it a value of 1, and
other countries receive a value of 0. We predict these pariah states receive significantly less ODA
than cooperative members of the international community. War is another dichotomous variable,
in which countries engaged in interstate conflicts with over one thousand deaths receive a value
of 1, while others receive a value of 0. We control for this variable because donor countries feel
that ODA’s potential is maximized when a country is at peace.
For cultural variables, we generate two collinear, dichotomous variables to account for
language and religion. Given that Rioux (2005) finds countries with French as their official language receive more French ODA, we control for this factor, assigning countries whose official


The Journal of Politics & International Affairs

language is French a value of 1 and others a value of 0. Likewise, we are curious to find a relationship between religion and foreign aid, as French xenophobia often targets Muslims. We create a dichotomous variable, labeled “Muslim” for countries with Muslim-majority populations,
to see if the French government allocates different amounts of ODA to these states. We assign
countries with Muslim-majority populations a value of 1 and others a value of 0.
We also control for economic potential, which Schraeder, Hook, and Taylor (1998) argue
significantly affects how France allocates ODA. We thus employ the natural log of GDP per capita to control for this factor. In later research, we hope to also control for French foreign direct
investment (FDI), which would further isolate immigration as the main variable of interest.
Population might also affect France’s allocation of ODA. We would expect countries with
larger populations to receive more ODA. We thus control for the natural log of population.
Ordinary Least Squares (OLS) regression corroborates bivariate results and shows immigration’s positive effect on French disbursement of ODA. Results hold at the 1 percent significance level. Countries that send more immigrants to France receive significantly more French
ODA. However, the nature of OLS regressions leave unanswered questions about how France
responds to changing flows of immigration. We would expect that as a specific country sends
more immigrants to France, the French government would respond by allocating said country of
origin more ODA.
To resolve this question, we turn to a fixed effects regression analysis. Results show
significance does not hold. While OLS regression suggests France sends significantly more ODA
to countries that send more immigrants, fixed effects regression indicates that, within a certain
country, France does not change its quantities of foreign aid to meet changing immigration flows.
The fact that France’s largest sources of immigrants have strong economic and cultural ties to
France– i.e. Algeria, Morocco, and Tunisia, etc. – might account for the small variance in aid
levels over time.
Table 2 depicts regression results. The first column shows findings after OLS regression
analysis, whereas the second displays country fixed effects regression results.
Immigration’s significant effect does not hold upon fixed effects regression. However,
although immigration to France no longer significantly affects the allocation of French ODA,
our data suggests many of our control variables do - specifically polity, pariah, ex-colony status,


French Aid for a French France

and language. In addition, it is noteworthy that as the number of observations increases, polity’s
significant relationship holds. Table 2 reveals these relationships.
Table 2




Pariah Status
Interstate War (>1,000 deaths)
Real GDP per Capita (Logged)
Polity 2 score
U.S. Military Assistance (2011
Log of Population



Fixed Effects
OLS Omitting
Immigration Immigration








French Language
















Number of observations


To further analyze the effect of pariah and polity on aid disbursement we omit immigration statistics to triple our number of observations. Our dataset grows to over 1800 observations.
This erases the significance of pariah status in both fixed effects and OLS regression. Polity’s
effect on French ODA becomes positive and significant at the 10 percent level upon fixed effects
regression. However, descriptive statistics present contradictory findings, as indicated by Figure

The Journal of Politics & International Affairs

Figure 3

We divide the polity scale into 2 parts, defining autocracies as countries that receive a
score between -10 and 6, and democracies as countries that receive a score between 6 and 10.
The entire sample of OECD data shows that, on average, France gives more aid to autocracies.
The same trend occurs even when we eliminate countries with a GDP per capita of more than
6,000 USD in order to eliminate observations about relatively developed countries that receive
negligible amounts of French ODA. However, OLS regression results (as shown in Column 4 of
Table 2) contradict these findings and indicate that other confounding variables may account for
the aforementioned descriptive finding. When incorporating control variables into OLS regression, we find that France does not provide significantly more aid to more autocratic or more democratic countries, contrary to what Alesina (2000) asserts. However, France does in fact reward
countries that democratize as indicated by fixed effects regression: as a specific country becomes
more democratic, it receives significantly more French ODA at the 10 percent level.
To further examine our three collinear dichotomous variables: religion, ex-colony status,
and language, we omit immigration to increase the number of observations. Table 3 shows the
following OLS regression results


French Aid for a French France

Table 3

French language
Number of observations

OLS Regression

OLS Regression
Omitting FRIM

Table 3 notes that the effect of ex-colony status and French language are significant at the 1 percent level.

It appears the French government rewards countries that have cultural ties to France. If

a country adopts French as an official language or is a former French colony, it receives significantly more French ODA. Both correlations hold at the 1 percent significance level. This demonstrates France’s commitment to preserving French culture and promoting it abroad, emblematic
of its cultural exceptionalism. France clearly considers cultural factors important when allocating
foreign aid. Still, one must consider that cultural ties often overlap with economic connections
that are known to affect how France disburses ODA (Schraeder, Hook, and Taylor 1998).
Our regression results generate interesting implications for France’s foreign aid and immigration
policy. Our findings coincide relatively well with France’s official foreign aid priorities. France
includes the promotion of democracy as one of its goals when dispersing foreign aid specifically
to the 16 Poor Priority Countries. We find France rewards countries that democratize, confirming this official pledge. Considering France’s unique history with democracy dating back to
the 18th century French Revolution, polity’s significant yet contradictory effect on French ODA
represents a curious avenue for further research. On the other hand, France does not list cultural
preferences in its official statements, while data suggests it does in fact reward countries that
possess cultural ties to France (although these could be attributed to overlapping economic rela-


The Journal of Politics & International Affairs

tionships). These results are not surprising as they confirm previous literature and reflect France’s
unique sense of cultural exceptionalism.
Over the course of our study, from 1994 to 2008, the French government does appear
to want to stem the tide of immigrants, as evidenced by stricter immigration policy and public
sentiment. Moreover, descriptive statistics show countries that send more immigrants to France
do receive more foreign aid, suggesting that foreign aid might be viewed and used as a tool to
decrease immigration inflows.
Nevertheless, regression results conflict. On one hand, OLS regression results support descriptive statistics and suggest France may in fact strategically allocate more aid to countries that
send more immigrants. However, if we were to claim immigration undoubtedly affected France’s
foreign aid flows, we would expect that as a specific country sends more immigrants to France,
the French government would respond by allocating the country larger quantities of aid. Fixed
effects regression analysis does not support this claim. Still, our finding that France allocates
significantly more aid to countries that provide larger numbers of immigrants supports Vandercasteelen (2012) and Azam and Berlinschi (2010) despite the fact that country-fixed effects are
not statistically significant. We would be wrong to ignore OLS regression results as we account
for traditional ODA determinants as well as uniquely French factors. It might be the case that
France targets optimal immigration flows, while not significantly increasing or decreasing aid to
a specific country over time.
Alternatively, considering that Azam and Berlinschi (2010) suggest multilateral institutions often have a large role in altering migration flows, France could leave most anti-immigration aid policies to the EU. Scholars find that multilateral institutions like the EU and UN, do
take into account migration flows (Azam and Berlinschi 2010). Since the EU has recently advocated for a “global approach” to immigration including “measures to fight illegal immigration
[such as] overseas development, managing demand for skilled labor, and action against traffickers,” it seems that in some cases international institutions do account for immigration flows when
distributing aid (Brady 2008). Tracking and analyzing the EU’s aid flows with regard to immigration represents an interesting avenue for future research.
Our research contributes to an emerging concentration on the relationship between aid
and migration. While our scope reduces the size of our data, the narrow focus allows us to ex-


French Aid for a French France

plore the nuances of French foreign aid disbursement and its peculiar tie with immigration. We
hope to continue research on the aid–migration tradeoff, a term coined by Azam and Berlinschi
The French and wider European public would benefit from more research on this relationship. While many assume their country allocates foreign aid as a form of constructive charity, governments undoubtedly have different goals in mind. It is essential that citizens realize
that government foreign aid policy has substantial effects on the global macro-economy, which
in turn affects phenomena as divisive and controversial as immigration: a dollar of foreign aid
could have a similar effect as a dollar’s worth of chain-link fence and barbed wire.


The Journal of Politics & International Affairs

Alesina, Alberto, and David Dollar. “Who Gives Foreign Aid to Whom and Why?” Journal of Economic
Growth 5, no. 1, 33-63. Accessed November 10, 2014.
Azam, Jean-Paul, and Ruxanda Berlinschi. “L’aide contre l’immigration.” [Aid against immigration]
Revue d’économie du développement 23, no. 4 (2009): 81-108.
Bermeo, Sarah Blodgett, and David Leblang. “Foreign Interests: Immigration and the Political Economy
of Foreign Aid”.” In Annual Meeting of the International Political Economy Society. 2009.
Berthélemy, Jean-Claude, Monica Beuran, and Mathilde Maurel. “Aid and Migration: Substitutes or
Complements?” World Development 37, no. 10 (2009): 1589-1599.
Blanchard, Marie. Histoire De L’immigration. [History of Immigration] Paris: La Découverte, 2001.
Bleich, Erik. “Race Policy in France.” The Brookings Institution. May 1, 2001.
Brady, Hugo. EU migration policy: An A-Z. London: Centre for European Reform, 2008.
Burnside, A. Craig, and David Dollar. “Aid, policies, and growth: revisiting the evidence.” Working Paper
3251, World Bank Policy Research, 2004.
Boone, Peter. “Politics and the effectiveness of foreign aid.” European economic review 40, no. 2 (1996):
Carvalho, Joao. Impact of extreme right parties on immigration policy: comparing Britain, France and
Italy. Vol. 20. Routledge, 2013.
Collier, Paul, and Anke Hoeffler. “Aid, policy and peace: reducing the risks of civil conflict.” Defense and
Peace Economics 13, no. 6 (2002): 435-450.
De Haas, Hein. “International migration, remittances and development: myths and facts.” Third World
Quarterly 26, no. 8 (2005): 1269-1284.
De Haas, Hein. “The determinants of international migration: Conceiving and measuring origin,
destination and policy effects.” (2011).
Development Assistance Committee (DAC) Data, 2012. Distributed by the Organization of Economic
Cooperation and Development (OECD). Paris.
“Une Europe Au Service Des Peuples Libres. [A Europe in Service of Free Peoples]” Le Site Officiel Du
Front National. Accessed December 8, 2014.
Fetzer, Joel S. “Economic Self-interest or Cultural Marginality? Anti-immigration Sentiment and
Nativist Political Movements in France, Germany and the USA.” Journal of Ethnic and Migration
Studies 26, no. 1 (2000): 5-23.
“Development Assistance.” French Ministry of Foreign Affairs and International Development. http://
Gaytan-Fregoso, Helena, and Sajal Lahiri. “Foreign aid and illegal immigration.” Journal of Development
Economics 63 No. 2 (2000): 515-527.
IPSOS, Global Views on Immigration. IPSOS Survey. August 1, 2011. Accessed November 13, 2014.
Hamilton, Kimberly, Patrick Simon, and Clara Veniard. “The Challenge of French Diversity.” Migration
Policy Institute. November 1, 2004. 
Heinrich, Tobias. “When Is Foreign Aid Selfish, When Is It Selfless?” The Journal of Politics 75 No. 2
(2013): 422-435
Hoeffler, Anke, and Verity Outram. “Need, Merit, or Self-Interest-What Determines the Allocation of
Aid?” Review of Development Economics 15 No. 2 (2011): 237- 250.
Institut National des Etudes Demographiques (INED), Data. Immigration régulière de long terme par


French Aid for a French France

groupes de nationalités et pour les nationalités les plus representées, par année d’admission au
séjour [Regular long term immigration by nationality and year of admission], 2008. Distributed by
INED, Paris.
Lahiri, Sajal, and Pascalis Raimondos‐Møller. “Lobbying by ethnic groups and aid allocation.” The
Economic Journal 110 No. 462 (2000): 62-79.
Le Pen, Jean-Marie. “Speech on Terrorism National Identity, Immigration, and Insecurity.” Lecture, Blue,
White, Red Party, Paris, December 13, 2001.
Maizels, Alfred, and Machiko K. Nissanke. “Motivations for Aid to Developing Countries.” World
Development 12, no. 9 (1984): 879-900.
Marthaler, Sally. “Nicolas Sarkozy and the Politics of French Immigration Policy.” Journal of European
Public Policy 15, no. 3 (2008): 382-97.
Mayda, Anna Maria. “International Migration: A Panel Data Analysis of the Determinants of Bilateral
Flows.” Journal of Population Economics 23, no. 4 (2009): 1249-274.
“L’aide Bilatérale De La France.” [France’s Bilateral Aid] France Diplomatie. http://www.diplomatie.
Ortega, Francesc, and Giovanna Peri. “The Causes and Effects of International Migrations: Evidence
from OECD Countries, 1980-2005.” Working Paper No. 14833, National Bureau of Economic
Research, 2009.
Rioux, Jean-Sebastien, and Douglas A. Van Belle. “The Influence of Le Monde Coverage on French
Foreign Aid Allocations.” International Studies Quarterly 49, no. 3 (2005): 481-502.
Schraeder, Peter J., Steven W. Hook, and Bruce Taylor. “Clarifying the Foreign Aid Puzzle: A Comparison
of American, Japanese, French, and Swedish Aid Flows.” World Politics 50, no. 2 (1998): 294-323.
Tandonnet, Maxime. Migrations: La Nouvelle Vague. [Migration: a new wave] Paris: Harmattan, 2003.
Paxton, Pamela, and Stephen Knack. “Individual and Country-level Factors Affecting Support for Foreign
Aid.” International Political Science Review 33, no. 2 (2011): 171-92.
Portes, Alejandro, and Jozsef Borocz. “Contemporary Immigration: Theoretical Perspectives on Its
Determinants and Modes of Incorporation.” International Migration Review 23, no. 2 (2007):
606-30. Accessed November 10, 2014.
Vandercasteelen, Joachim, 2012. “The Impact of International Migration on Foreign Aid Flows,” Master’s
thesis, Katholieke Universiteit Leuven.
Vreeland, James Raymond, and Axel Dreher. The Political Economy of the United Nations Security
Council: Money and Influence. Cambridge University Press, 2014.


Submission of Manuscripts
The Journal of Politics & International Affairs (JPIA) welcomes submissions from undergraduates
of any school, class, or major. We seek to publish manuscripts of the highest quality and papers
for publication are generally exceptionally written, with well-developed theses, and exhibit
articulate arguments with original analysis or research. Submissions can include: Full academic
papers with original research based on new primary data or literature review with strong
argumentative input, policy analyses or evaluations, or short academic book reviews.
Papers are typically 10-20 pages in length and have been written for an upper-level course or
abridged forms of an undergraduate research thesis. Manuscripts for consideration should
include an abstract of approximately 150 words. Citations and references should follow the
American Political Science Association Style Manual for Political Science. All references must be
complete, accurate, and up-to-date for submissions to be considered. References in manuscripts
should be in-text citations in the author-date format with a complete bibliography.
Those who submit papers may be asked to revise their manuscript before and after it is accepted
for publication. Papers can be submitted digitally on the journal’s website at where
you can also find submission guidelines including a digital copy of our style manual. You will be
asked to provide your full name, university, class rank, a short biography, and contact information
(mailing address, e-mail address, and phone number). Although papers are encouraged and
accepted on a rolling basis, they will only be considered for publication during each publication