Description
seven063_ec386_final-3000 words term paper
Unformatted Attachment Preview
EC386 Health Economics
Autumn 2023
Lecturer: Nicholas Reynolds
UNIVERSITY OF ESSEX
Term Paper
The Assignment for EC386 takes the form of a term paper (extended essay) to be submitted in
mid/late January to FASer (you can find the exact due date by logging into FASer). Late
submissions receive a mark of zero, unless you receive approval through the formal Extenuating
Circumstances committee procedure. See the Undergraduate Economics Handbook, pages 28–9.
You are responsible for complying with the University’s regulations on academic offences. See
Undergraduate Economics Handbook, pages 55–8.
The term paper will be similar to the formative assessment you completed — with a focus on
critical evaluating the claims and evidence of an article or other source (see below for the list of
possible sources). The goal of the paper will be to assess the substantive claim and empirical
support made in the article/other source. You will also be asked to describe additional evidence
for or against the claims — this will require looking on your own for other things written on the
topic.
The term paper will be substantially longer than the formative assessment and should follow the
structure and format of a paper — with a logical structure, citations etc. It should be 2,500-3,000
words and should convey a thoughtful, critical engagement with the main source as well as
additional research into other sources on the same topic.
Read carefully the sections of the Undergraduate Economic Handbook: “A Guide to Good
Practice in Assessed Work” and “Making the Best of Your Essays, Term Papers and Projects”.
You are strongly advised to read and follow this guidance.
What your paper should cover
Your goal in this paper will be to describe and assess the argument made in the article/source.
The list of articles you can choose from are below.
Write the paper as if you are describing the argument to a scholar who is familiar with math,
analytical thinking, and economics but is not an expert in the particular topic covered.
Explain to them the argument made in the article in you own words.
Explain the important assumptions that underlie the argument. This may include theoretical
assumptions which are very difficult or impossible to test. And it may also include some
empirical facts which they cite to support their argument — but which may be disputed by other
scholars.
Give your own assessment of the likely validity of the assumptions and therefore of the
conclusions made in the argument. This will involve both critical thinking AND secondary
research of your own.
You can start this assessment by looking at the articles which your source cites. You should also
look yourself in Google scholar or other sources for additional research relevant to assessing the
authors’ argument.
Finally, describe additional tests that could be carried out to further assess the validity of the
author’s claim. The more concrete the better here. For example, you could describe a particular
empirical relationship that the author’s could examine to potentially falsify their theory. You can
also describe if any of these tests have already been conducted by other researchers — and if you
think they falsify the authors’ theory.
Possible topics
Below you will find a list of sources/topics. Pick ONE of the following sources to evaluate for
your paper:
1. The BCG vaccine does not protect against COVID-19: Applying an economist’s toolkit
to a medical question
Richard Bluhm, Maxim Pinkovskiy
https://voxeu.org/article/bcg-vaccine-does-not-protect-against-covid-19
See also:
Bluhm, R. and Pinkovskiy, M., 2021. The spread of COVID-19 and the
BCG vaccine: A natural experiment in reunified Germany. The
Econometrics Journal, 24(3), pp.353-376.
Related to:
Block 1: Causality and Research Design
Block 5: Coronavirus and the SIR model
2. Pregnancy during the pandemic
Hannes Schwandt
https://voxeu.org/article/pregnancy-during-pandemic
Related to:
Block 1: Causality and Research Design
Block 2: Socioeconomic Disparities in Health
Block 5: Coronavirus and the SIR model
3. Free to choose?
Marty Gaynor, Carol Propper, Stephan Seiler
https://voxeu.org/article/free-choose
See also:
Gaynor, M., Propper, C. and Seiler, S., 2016. Free to choose? Reform,
choice, and consideration sets in the English National Health Service.
American Economic Review, 106(11), pp.3521-57.
Gaynor, M., Moreno-Serra, R. and Propper, C., 2013. Death by market
power: reform, competition, and patient outcomes in the National Health
Service. American Economic Journal: Economic Policy, 5(4), pp.134-66.
Related to:
Block 1: Causality and Research Design
Block 2: Socioeconomic Disparities in Health
Block 4: Health Care Policy
4. Long-run economic impact of in-utero shocks
Mazhar Waseem Timotej Cejka
https://cepr.org/voxeu/columns/long-run-economic-impact-utero-shocks
See also:
Cejka, T. and Waseem, M., 2022. Long-Run Impacts of In-Utero
Ramadan Exposure: Evidence from Administrative Tax Records.
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4082587
Related to:
Block 1: Causality and Research Design
Block 2: Socioeconomic Disparities in Health
Alternatively, you may select you own article that makes a theoretical claim about health
economics (broadly definied) to evaluate. If you choose your own piece to evaluate, you must
check with me ahead of time. A good place to look for alternative aricles would be voxeu.rog.
Email me with a possible term paper topic by 12th December 2022.
If I have not approved an alternative piece/topic by that date you must choose one of the
above topics!
How the paper should be structured
Your paper should be logically organized into sentences and paragraphs.
A good structure would be to include the following sections:
1. Introduction
Describe in one short paragraph what your paper will be about. Ie. that you will
describe and assess the article. Preview what your overall assessment of the argument
will be in 1-2 sentences.
2. Description of author’s argument and assumptions
Describe in your own words the arguments made by the author’s. State the
assumptions as you go.
3. Assessment of assumptions and claims
Assess some of the key assumptions and therefore provide your opinion of the likely
validity of the claims. Again, this is where references to other sources will be
essential.
4. Alternative possible tests
Describe additional tests that could be carried out to further assess the validity of the
authors’ claim.
5. Conclusion
Restate the argument of the source and your assessment of it.
You should end with a reference list citing the main source and other sources you used in your
paper. Use Harvard style — see the student handbook and library reference page.
Figures and equations may be helpful in making you argument, and if you include them they
should be included in the body of the paper.
Other Points
Check carefully the sections of the Undergraduate Economic Handbook: “A Guide to Good
Practice in Assessed Work” and “Making the Best of Your Essays, Term Papers and Projects”.
You are strongly advised to read and follow this guidance
Avoid academic offences
Never copy from any source unless the words copied are enclosed in “quotation
marks” and are properly referenced.
Avoid long quotations
Use quotations for a given purpose, such as stating someone’s position or view. Do
not let quotations be a substitute for your own exposition: do not be afraid to use your
own words.
Always provide bibliographical references, with full publication details (e.g. journal title, volume
and page numbers; working paper series; book publisher and location). List references
alphabetically by the first author’s surname.
Presentation matters
Pay attention to coherence, st
le and layout. Use word processing and double linespacing with a 12 point font. Write in
coherent, grammatically correct English and check spelling.
Keep to the word limit. Term papers should not exceed 3,000 words. In general, they should be
at least 2,500 words.
75
University of Essex
Department of Economics
Mark:
Registration Number:
EC386 Autumn Assignment
FEEDBACK SHEET
Contributing Factors
Mark
Range
90100%
80-89%
Overall quality
Structure
English
Analysis
Exceptional term
paper, highly original
Exceptional standard
of professionalism in
organising the
material
Displays exceptional
writing skills
Exceptional
competence in several
areas of economic
research*
Displays an excellent
and logical
understanding of the
material
Excellent standard of
professionalism in
organising the
material
Displays excellent
writing skills
Excellent competence in
several areas of
economic research*
Displays very good
grasp of the material
Very good standard
of professionalism
in organising the
material
Very good writing
skills
Very good
competence in several
areas of economic
research*
Displays a good grasp
of the material
Good standard in
organising the
material
Displays good writing
skills
Good competence in
several areas of
economic research*
Displays satisfactory
understanding of the
material
Satisfactory standard
in organising the
material
Displays satisfactory
writing skills
Displays satisfactory
levels of competence in
at least one area*
Potential
understanding of the
material but insufficient
demonstration of
competence
The term paper may
be unsatisfactory
organised
Displays
unsatisfactory writing
skills in several parts
Displays satisfactory
competence in one area
but other areas are
inadequate
Inadequate term paper
with insufficient
demonstration of
competence
The term paper is
poorly organised
The term paper is
overall very poorly
written
Inadequate competence
in all areas of economic
research
Wholly inadequate
term paper
The term paper is
very poorly organised
The term paper is
extremely poorly
written
Failure to demonstrate
competence in any area
of economic research
70-79%
60-69%
50-59%
40-49%
30-39%
0-29%
* eg theoretical analysis, data presentation and analysis, use of examples and case studies, critical
discussion of the literature.
28/11/2022
Comments on the assignment:
This was a very strong term paper. The clarity of the writing stood out in particular. The description of
the arguments made in the assigned source was very clear. There is also a strong attempt at critical
assessment of the author’s claims, with references to outside sources.
That said, there is some room for improvement in terms of the critical assessment. It is clear that the
student understands the idea of confounders and the description of other factors — such as changes
in nutrition — driving the results is a nice start on this. However, it is important to note that these
alternative explanations would have to change discontinuously at exactly the birth-month cutoff of the
educational policy. Similarly, the reference to a possible vaccination campaign is a good point — but
again this campaign would have to have a very very specific timing. A great next step would’ve been if
the student had done some research on the timing of prominent vaccination campaigns themselves.
Some of the critiques are also of things that are not really essential assumptions of the author’s
argument. For example, the idea that lifestyle decisions can affect chronic conditions is not an
essential assumption — more just a motivation of why might think to look for an effect of education
on chronic conditions.
28/11/2022
EC386: Introduction to Health Economics
and Policy
Term Paper Article: The causal effect of education on chronic
health: Evidence from the UK (2020)
Word Count: 2,684
1
Introduction
This term paper will describe and analyse the arguments presented in the article regarding the
relationship between education and chronic health. A strong positive statistical association
between the two variables are identified in existing studies. However, when studying the causal
effect of education on chronic health there has been varied views as to whether causality is
present or not. Therefore, the literature has investigated the causal relation of schooling on
chronic health conditions by utilising educational reforms which occurred in the United
Kingdom. The claims of the research will be evaluated, hereafter, alterative tests will be
suggested to falsify or further validate the study.
Description of Arguments
Janke et al. (2020) claim there is a negative correlation between the level of education attained
and chronic health conditions. However, there is no significant causal effect of an additional
year of education on chronic health outcomes with the exemption of diabetes and
cardiovascular disease.
The correlation between the two variables may be derived from the causal link of education on
chronic health conditions or a third variable affecting the relationship. To understand and
distinguish this, the authors utilised UK educational reforms as an instrument for education.
With the use of national reforms, quasi-experimental research designs are applied. This is as
random assignment is not present here, however, two groups of individuals exist: pre-reform
and post-reform.
The first reform was implemented in 1972. A change in policy meant individuals born after a
cut-off date (September 1957) were not allowed to leave school until the age of 16. Whereas,
2
for students who were born before the cut-off date, the minimum school leaving age was 15.
As a result of the educational reform, students received an extra year of schooling (25% of the
population of pupils who were born after September 1957 were affected). Two groups are
present: students born before the cut-off date and students born after. Additionally, those
affected by the 1972 reform would be in their “late middle age” at present time, where the
development of chronic ill health is likely.
The second reform occurred in 1990. There was a growth in education attainment due to the
consequences of policy changes. This affected students who were in cohorts from 1970 to 1976.
However, mainly affected whom may have left at the minimum school leaving age (16 years
old) and those who were anticipated to complete secondary education but not expected to start
higher education. Though, not compulsory, the educational expansion still resulted in more
individuals who were able to participate in higher education which raised the average schooling
years by 1.5 years in a decade. In addition, “entering their 40s” are those affected by the 1990
educational expansion, where chronic ill health “begin to manifest”.
This study found both the reforms used to analyse the causal effect of an additional schooling
year on chronic conditions indicated no statistical significance with the exemption of diabetes
and cardiovascular disease. Figure 1 presents six graphs for each named chronic condition. The
relationship between the number of months born before or after the 1972 reform, and the
chronic illness score, the number of conditions and the proportion with health problems is
represented. The cut-off point is 0 months born before/after the reform. If there is a clear
indication of discontinuity around the cut-off point (reform date), it means a causal relation
exists between years of education and chronic disease. At the reform date, Figure 1 illustrates
a sign of discontinuity for reporting any chronic conditions, and cardiovascular disease and
3
diabetes, panel (a) and panel (e), respectively. However, this is not constant for the number of
chronic conditions, musculosketal conditions, mental health conditions and neurological
conditions. Whilst, for the remaining panels a discontinuity is not displayed. Figure 2
represents the educational expansion reform in relation to the proportion with health problems
and their month and year of birth. The variables have a negative association as the proportion
of health problems increases as the birth year decreases (the individuals are older). However,
there is no distinct discontinuity, so no causality is present. Importantly, the study expresses
the chances of having diabetes is decreased by an additional schooling year by 3.6 and 0.9
percentage points for individuals affected by the 1972 reform and 1990 expansion,
respectively.
Assessment of Assumptions and Claims
The authors’ claims on the relationship between education and chronic health are heavily based
on a range of assumptions. Even if one assumption is disproven, the research is at risk of losing
credibility.
The main assumption Janke et al. (2020) made is that poor “lifestyle decisions”, which is
possibly influenced by the level of education attained, lead to many chronic illnesses. This
assumption corresponds to Grossman’s (1972) theory “more educated people are more efficient
producers of health”. To prove this theory, the causal effects of education on chronic health
were examine by using educational reforms as an instrumental variable. The use of
instrumental variables in quasi-experiments help in eliminating “the potential bias resulting
from” observable and unobservable confounders (Martens et al. 2006). Further increasing the
credibility of the research method, as when a change is detected after the reform, then the
results are assumed to be attributed to the additional year in education. However, the 1972
4
reform requiring pupils to remain for an additional school year may not have caused the
discontinuity in reporting any chronic illness (Figure 1, panel a) through education. In contrast,
a third variable such as diet or exercise could explain the causality on reporting any chronic
illness. For instance, at the time school meals may have been nutritious or improved in
nutrition. Additionally, Janke et al. (2020) also assume poor diet develop many chronic
illnesses. Consequently, it is acceptable to suggest the discontinuity present in Figure 1 (panel
a) from the 1972 reform could have be caused by “improved nutrition in schools” (Mazumder
2008). Therefore, the assumptions of the use of the 1972 reform as a credible instrumental
variable for education and education influencing chronic diseases reported can be falsified.
However, this is only true if a strong confounder (e.g., nutrition school meals) and weak
instrumental variable is present (Martens et al. 2006).
The evidence used to support the claims made are entirely based on data from the United
Kingdom (a high-income developed nation). Both the educational reforms utilised were the
consequences from policy related decisions the UK government undertook. Accordingly, the
UK Quarterly Labour Force Survey was used to observe the health outcomes of the reforms.
The assumption of this is that the results collected from one nation can be generalised to fit all.
Conversely, the results of the study may differ if another country was to be used to analyse the
causal effects of education on chronic illness. Using the 1997 reform that raised the compulsory
schooling years by three years in Turkey (a middle-income country), many authors studied the
causality of schooling on health. Cesur et al. (2014) and Balgati et al. (2019) found schooling
had no causal effect on self-reported health and suffering any health issues in the last six
months, respectively. The results centred on Turkish data and policy differ from that of UK
data. This is since Janke et al. (2020) claim education has an effect on any chronic condition
(Figure 1, panel a), whereas Cesur et al. (2014) claim education has no effect on self-reported
5
health. There is limited literature analysing the causation of education on chronic conditions in
Turkey, therefore general health issues and self-reported health is discussed here instead.
Hence, it is understandable that the mentioned comparison is not the best analysis.
Alternatively, the results may also vary if a range of countries were to be studied. Fonseca et
al. (2018) used “variation in the timing of educational reforms” among fourteen OCED nations
to assess the causal effect of schooling on health outcomes such as chronic illness. They found
an additional year in compulsory education decreases the chances of “reporting poor health”
and “having any chronic illness” by 6.58 and 4.4. percentage points, respectively. However, as
most OECD countries are high-income nations, this alternative study can also be considered to
be heavily centred around developed nations. Overall, the authors’ claims are made on the
assumption that the results obtained from using the UK as a case study can be simplified to
reflect the general causation. However, this can be falsified by examining the causal effects of
educational reforms on the same health outcomes in other countries and obtaining differing
results.
Self-reported data from surveys is used in the study to obtain health conditions of individuals.
The assumption made is that this data is accurate and truthful. However, reporting errors in the
form of bias can limit the data. When reporting health conditions, individuals may report what
they may feel is socially common or acceptable. In addition, self-reported answers are open to
subjectivity and differing interpretations. “False negative reporting” decreases by around 7
percentage points for individuals with degree level education (Johnston et al. 2009). This
indicates more educated individuals are likely to have more knowledge about their health
(Janke et al. 2018). Alternatively, less educated individuals might misjudge their symptoms
and not recognise their health conditions. As a result, the assumption of using accurate data is
falsified. Furthermore, this questions the validity of the authors’ claims. For instance, using the
6
1972 reform, an extra schooling year did not have a causal effect on the number on chronic
conditions (Figure 1, panel b). However, there is a probability that individuals born before the
reform (who did not have to undertake a compulsory additional schooling year) may
underestimate their health conditions. Hence, the data values of number of conditions may be
larger than plotted in panel b for individuals born before the reform date. If this is true then a
break around the reform date will be present, displaying schooling having a causal effect on
the number of chronic conditions (health).
Possible Alternative Tests
To evaluate the validity of the claims made by Janke et al. (2020) based on the relationship
between education and chronic conditions, additional tests can be conducted.
By utilising the two educational reforms, this research study only measures the quantity of
education. For example, only the number of years (time) in school is focused on. However,
schooling can also be measured through the quality of education as well. The quality of
education refers to creating awareness on chronical diseases, their causes and how to reduce
your chances of becoming ill. Henceforth, students will be able to learn how to and may follow
a healthy living lifestyle. The application of two different education measures helps in the
assessment of whether the research method used is highly reliable or can be falsified. For
example, Manios et al. (1999)(2002) studied the relationship between health education
(measured by quality of education) and health. A six-year primary school-based intervention,
developed from the “Social Learning Theory”, was used to inform and encourage first grade
students in Crete to partake in healthy eating and activities. As a result, beneficial “changes in
chronic disease risk factors” were observed for the intervention group. Unfortunately, this
study does not identify causation, but only the positive statistical association between the two
7
variables. Accordingly, the findings of Manios et al. (2002) partially support the claims made
by Janke et al. (2020); education has a positive association with good health and negative
association with chronic ill health. However, the strength of the claim’s validity would have
increased if the two different research approaches assessed causality and led to the same health
outcome-related findings.
Additionally, strong confounding variables (such as vaccination and nutrition) may correlate
with both education and chronic conditions. If confounders are ignored, problems will arise
through biasing and incorrectly estimating the causal effect of schooling on chronic illness. As
mentioned before, these confounders may have influenced the research and the discontinuities
present. Therefore, education may not have had a causal effect on reporting any chronic
condition (Figure 1, panel a), and cardiovascular disease and diabetes (Figure 1, panel e).
Researching whether there was an introduction to national vaccination programmes close to
when the 1972 reform was implemented, helps with falsifying the findings of Janke et al.
(2020). This is by revealing the causal effects of vaccination programmes around the time of
the 1972 reform, on chronic health exists. Similarly, examining policies to encourage healthy
eating during the time of the 1972 reform may also support to falsify the authors’ claims.
Conclusion
To summarise, the article uses evidence from the UK to examine the causal relation of an extra
year in education on chronic ill health. Utilising the 1972 reform the authors claim no causal
effect of schooling years exists on the number of chronic conditions, musculosketal conditions,
mental health conditions and neurological conditions. On the other hand, there is a statistically
significant effect of education on reporting any chronic condition, and reporting cardiovascular
disease and diabetes. This implies education is specifically “protective” for chronic illnesses:
8
diabetes and cardiovascular disease. Using the educational expansion reform, Janke et al.
(2020) claim no causal link is present for educational attainment on health problems. These
claims are supported based on many assumptions. Therefore, the dismissal of one assumption
may invalidate the results obtained. For example, the assumption that the utilised instrumental
variable (compulsory schooling reform) is strong could be violated if an unobserved
confounder is later discovered. This challenges whether the findings concerning the causality
is accurate.
On the whole, there are still mixed interpretations around the subject of the causal effects of
education on chronic conditions. Majority of the claims made by Janke et al. (2020) had no
causal relation between the two variables. However, the cardiovascular disease and diabetes
cluster indicated education had a causal effect on them. For that reason, policymakers and
educational settings could further promote education (through raising compulsory schooling
years or creating intervention programmes) in aims to shrink the incidence of the diseases.
References
Baltagi, B. H., Flores-Lagunes, A. & Karatas, H. M., 2019. The effect of education on health:
Evidence from the 1997 compulsory schooling reform in Turkey. Regional Science and
Urban Economics, 77, pp. 205-221.
9
Cesur, R., Dursun, B. & Mocan, N., 2014. The impact of education on health and health
behavior in a middle-income, low-education country. No. w20764.
Fonseca, R., Michaud, P.-C. & Zheng, Y., 2020. The effect of education on health: evidence
from national compulsory schooling reforms. SERIEs: Journal of the Spanish Economic
Association, 11(1), pp. 83-103.
Grossman, M., 1972. On the Concept of Health Capital and the Demand for Health. Journal
of Political Economy, 80, pp. 223-255.
Janke, K., Johnston, D., Propper, C. & Shields, M. A., 2018. The Causal Effect of Education
on Chronic Health Conditions. IZA Discussion Paper No. 11353.
Janke, K., Johnston, D., Propper, C. & Shields, M. A., 2020. The causal effect of education
on chronic health: Evidence from the UK. [Online]
Available at: https://voxeu.org/article/causal-effect-education-chronic-health-evidence-uk
[Accessed 07 January 2021].
Johnston, D., Propper, C. & Shields, M. A., 2009. Comparing subjective and objective
measures of health: Evidence from hypertension for the income/health gradient. Journal of
Health Economics, 28(3), pp. 540-552.
Manios, Y. & Kafatos, A., 1999. Health and Nutrition Education in Elementary Schools:
Changes in health knowledge, nutrient intakes and physical activity over a six year period.
Public Health Nutrition, 2(3a), pp. 445-448.
Manios, Y., Moschandreas, J., Hatzis, C. & Kafatos, A., 2002. Health and nutrition education
in primary schools of Crete: changes in chronic disease risk factors following a 6-year
intervention programme. British Journal of Nutrition, 88(3), pp. 315-324.
Martens, E. P. et al., 2006. Instrumental Variables: Application and Limitations.
Epidemiology, 17(3), pp. 260-267.
Mazumder, B., 2008. Does education improve health? A reexamination of the evidence from
compulsory schooling laws. Economic Perspectives, 32(2), pp. 2-16.
Appendix
Figure 1 – Prevalence of chronic conditions for months-of-birth around the 1972 reform cut-off (Janke
et al, 2020)
10
11
Figure 2 – Prevalence of chronic conditions for months-of-birth around the education expansion
reform (Janke et al, 2020)
12
Purchase answer to see full
attachment
