Principles of Macroeconomics F2020
Visualizing Welfare and Inequality
The goal of this assignment is to give you the opportunity to play around with some government
data and to introduce you to some neat tools. Along the way, we’ll think about GDP as a proxy
(substitute measure) for development and/or welfare, and make comparisons across
populations. Except that, to make life easier, we won’t use GDP. Instead, we’ll use GRP, Gross
Regional Product (the GDP equivalent for each province or region), and annual regional data
from China’s National Bureau of Statistics.
Files needed for this assignment:
Assignment Welfare and Inequality F2020.pdf (this document)
Economist 201809 Life in developing countries continues to improve – Daily chart.pdf (required
reading on the HDI)
Sample.xlsx (Excel file with multiple sheets for data, graphs, and examples)
Report Template.docx (Word doc to help with formatting the report)
Prov.dbf, prov.prj, prov.shp, prov.shx (files for use with GeoDa, a free GIS)
Before you start:
Check that the Excel file works well on your computer. There are three sheets, which you should
keep in order. The first has data. The second sheet has a scatterplot. The third sheet has sample
calculations on data.
Install GeoDa (https://geodacenter.github.io/download.html)
Links to data (National Bureau of Statistics of China)
Direct link to Annual Data by Region http://data.stats.gov.cn/english/easyquery.htm?cn=E0103
Maps (Use GeoDa for final submission) http://data.stats.gov.cn/english/mapdata.htm?cn=E0103
Links to software:
Other links for inspiration:
Best and most entertaining, by Hans Rosling and friends
Gross National Happiness
Background and preparation
1. OPTIONAL To get warmed up, take the Gapminder Test. Can you get a perfect score?
2. OPTIONAL The entire website is quite nice. Dollar Street is a neat way of showing how
different people live. There’s are many photos and details there if you select a family
and click Visit this Home. On other portions of the site, the data visualization is quite
sophisticated, as well.
Ex. Hit “Play” to see how Argentina has stagnated over the past 100 years.
3. Read the Economist’s article on the Human Development Index included in the
a. Details on calculating the HDI: https://www.calculators.live/humandevelopment-index
4. Go the ENGLISH website for China’s National Bureau of Statistics website and navigate
to the Regional Annual data (data > annual > regional (top bar) > annual data.
a. Only use annual data for this assignment.
b. Refer to the video available on NYU Classes for more details.
i. Contrary to what I said in the video. The original .xlsx file will work.
5. OPTIONAL In the provided Sample.xlsx for this assignment, create a new column on the
first page and calculate the GRP per capita in 2018 for each province using the
appropriate formula in Excel.
a. You will need to check the units for GRP and population on the NBS website for
these series (GRP under National Accounts, and regional population under
Population) by bringing up the information tooltip for each series.
b. Verify against the GRP per capita included in the sample. The differences should
6. Browse the data on the NBS website and, in Excel, calculate your own measure of
a. Based on your reading and sensibilities, think about what sort of data you would
like to use.
b. Look for relevant data series on the NBS website.
c. Copy and paste any series you use into your Excel sheet under a new column.
Data for 2019 is mostly complete, but many interesting series are only available
for earlier years. It’s fine if you need to use previous years’ data, either for all
series (i.e. construct a measure for 2018) or as needed (i.e. using latest available
data). However, avoid using any data from before 2010.
i. E.g. You might want to deflate income by regional price level. Notice
that NBS reports consumer prices in a way that is difficult to compare
over a longer timeframe without additional calculation.
d. From the series you copied, compute your new measure.
i. E.g. the HDI is a composite of four other measures, some of which have
been “normalized” in some way. Others illustrate development by
looking at changes in life expectancy and real GDP per capita (but how
should you weigh each aspect?).
ii. Recommendation 2020: For most cases, I suggest forming your
measure by taking the arithmetic mean of the normalized each data
series. To normalize each series, one compatible method is to scale the
values such that the minimum provinces value equals 0 and the
maximum province’s value equals 1. An example is included in the
Sample.xlsx file. Double-check that the values make sense with your
iii. TIP2020: If you still want to use a geometric mean for your measure
(not necessary), pay attention to the method of normalization for each
column. No province should have a value of 0 (or less) or the geometric
mean will have problems (since a geometric mean starts by multiplying
all the values together). Several normalization variations are included in
the Sample.xlsx file, and details used to calculate the HDI are available
online (e.g. https://www.calculators.live/human-development-index)
7. Compare your new measure to GRP per capita by plotting both series on a scatterplot.
a. You can replace the series on the second sheet of the Sample.xlsx and the
scatterplot should update automatically.
b. You will want to edit the scatterplot and save an image for use in your final
8. Load Geoda and open the prov.shp file. Merge your new data series into Geoda
(Table>Merge. Merge>Key current = NAME, Key data = Region, select all series if you
a. Watch the video on NYU Classes for step-by-step instructions
9. Visualize your measure on the map of China. Change the basemap for a little flash!
a. NBS has regional data maps as well, but not of your measure! Feel free to
explore with this online tool, but please use GeoDa to produce the map for your
b. Save an image of your map to use in your paper
After going through the steps above, you will produce a short paper documenting your work and
thoughts. Your submission must conform to the outline below.
1. Title: YourMeasure: A GDP Alternative
a. You can give your measure a snappy name
2. Author: Your name (NetID and email)
3. Submission date
Structure (Report Template.docx has been included to help with this stage):
1. One-paragraph abstract (i.e. a summary of 100 – 150 words)
a. About 5 sentences
i. Suggested template (can copy if stuck): Though widely used, Gross
Domestic Product (GDP) is an imperfect measure of aggregate welfare.
In this paper, I construct a novel index, [YourMeasure], which
incorporates [concepts that motivated you], using publicly available
data from the National Bureau of Statistics of China. Compared to Gross
Regional Product per capita, YourMeasure [summarize the result of the
comparison in the scatterplot with a few words]. Based on YourMeasure,
[a few words summarizing the main patterns from the mapping
exercise]. [Any other concluding sentence].
b. This is the most important paragraph. Any other student who did this
assignment will be able to understand more than 50% of what you did and
found from this paragraph.
c. Obviously, you will write this paragraph LAST (even if it’s at the beginning).
2. Introduction (not technical) 2 – 3 paragraphs. Super short paragraphs are fine (this is not
a real paper)
a. Motivate your measure (what were you considering when you came up with it)
b. Preview the interesting results
c. Often this section functions as an extended abstract.
3. Methods: Someone reading this section should be able to compute your values from the
a. Describe in detail the construction of the measure.
b. Be sure to name of the various series you used from the NBS. What they
measure (if the name is unclear), and why you chose them. It also helps to give
some sense of the data. One strategy is to compare values for familiar or
extreme regions (e.g. disposable income in Shanghai or Beijing versus Xinjiang
c. Any calculations you used to get your new measure. For brevity, you do not too
be extremely detailed here, unless you feel you did something that really
requires attention. Otherwise, imagine that the specific equations might go the
appendix (not needed for us).
i. E.g. Each category was normalized to a value between 0 and 1, and the
final index was computed from a weighted average, with W and X
weighted 1/3 each, and Y and Z weighted 1/6.
4. Results section (as few as two paragraphs with two figures)
a. Comparison to GRP subsection
i. Scatterplot figure with caption goes here
ii. Summarize what how YourMeasure compares to GRP according to the
iii. You may interpret a bit here. Save lengthy interpretation for Discussion
b. Looking at welfare across China using YourMeasure
i. Your Geoda map image with caption goes here
ii. Describe any patterns or trends from the graph. What does
YourMeasure “say” about wellbeing in China.
iii. You may interpret a bit here. Save lengthy interpretation for Discussion
5. Discussion section
a. Based on the results, is there anything you want to say? Support your comment
i. E.g. How might you interpret the patterns or trends identified on your
map? Can you suggest any causes or consequences? This is the time to
bring in considerations beyond the data and computation, such as
current events, history, problems or successes.
ii. E.g. You were limited by the data available. Is there some other data
you might like that would help you to improve your measure?
iii. E.g. Put another way, this is a “What did I learn” sort of section
6. Conclusion (Optional)
a. Generally very brief. Most of us are tired by now.
i. Except for those of us who start reading from the end, which is
definitely something people do with papers. So the lazy way is to restate
the abstract or intro highlights here.
b. Any takeaway for the reader?
c. Based on your results, does your new measure add much insight beyond just
GRP (or GDP)?
Figures: Minimum two (one scatterplot or equivalent. One map from Geoda). Maximum three (I
know the maps are fun). You may make a figure with multiple panels (e.g. a map of your
measure in panel A and a map of GRP in panel B to compare and contrast geographically), but
that would be far beyond the needs of the assignment.
Citations: Should not need to use any, but just be consistent if you do.
Total length: < 1500 words (~150+300+200+200+300). Less, preferably. Grading: The writing may be difficult, especially the introduction. Do not spend too much time writing for this assignment. Focus your attention on the abstract, which will be read more carefully, using the template as a starting point. The two figures are also important, so that we can see that you attempted the basic assignment. Otherwise, concise, even simple sentences expressing ideas that are logically organized and developed will be more than sufficient. Grammar and style will not be heavily penalized (or praised), if at all, as long as we can read it. The key points approximately correspond to each section of the paper: a. Intro: We can understand what you wanted to do and why b. Method: We can understand how you constructed the measure and it makes sense with what you wanted to do i. I think this second part is the trickiest. It doesn’t make sense to simply add a GRP per capita value in the thousands of yuan to an average pollution number quoted in 2.5 PPI. c. Result 1: We can understand how your measure compares to GDP (GRP) i. scatterplot d. Result 2: We can understand what your measure says about welfare across China i. Data map e. Discussion & Conclusion: We believe that YOU understand what you did and what it means Format: PDF for the report. Other files in a .zip. Submit on NYU Classes for all students. In-person students may also submit a stapled hardcopy of the report (only) to Jingchao during recitation.