Data Visualization for Economics Research


Lyon Summer School in Empirical Research Methods - ENS de Lyon

Vincent Bagilet

2025-07-01

Introduction

Everywhere but not so simple

  • Data viz is everywhere

  • We work with data, we routinely (need to) visualize it

  • Seems pretty simple, we all know how to make graphs

  • Sure BUT there are a few things we need to think about when visualizing data

  • Once you pay attention to data viz, it is fun, instructive and satisfying!

Data viz can be obviously deceptive

…or difficult to interpret


They can be memorable and insightfull


…and also beautiful

Outline




1. Usefulness and importance


2. Key data viz principles


3. Building a graph


4. Data viz for research in economics

Usefulness and importance

Types of graphs



Graphs to explore


  • Analyze
  • Confirm


Graphs to explain


  • Inform
  • Convince



They have different goals and audiences

Explore: make sense of your data

  • Data often contain patterns; data viz can be tremendously helpful to identify them

  • But need to look at your raw data

  • That’s the role of the exploratory data analysis (EDA)

  • It may also help you formulating hypotheses (to test on other data sets)

Look at your raw data

Same relation, different patterns

Look at your raw data

Some plots (or summary statistics) help summarize but can also hide information

Explain: communicate your results

Country ISO 2020 2020 Ranking 2025 2025 Ranking Region Pattern
Afghanistan AFG 62.30 122 17.88 175 Asia-Pacific Lower
Albania ALB 69.75 84 58.18 80 EU & Balkans Lower
Algeria DZA 54.48 146 44.64 126 MENA Lower
Andorra AND 76.77 37 63.30 65 EU & Balkans Lower
Angola AGO 66.08 106 52.67 100 Africa Lower
Argentina ARG 71.22 64 56.14 87 Americas Lower
Armenia ARM 71.40 61 73.96 34 EECA Higher
Australia AUS 79.79 26 75.15 29 Asia-Pacific Lower
Austria AUT 84.22 18 78.12 22 EU & Balkans Lower
Azerbaijan AZE 41.52 168 25.47 167 EECA Lower
Bahrain BHR 39.87 169 30.24 157 MENA Lower
Bangladesh BGD 50.63 151 33.71 149 Asia-Pacific Lower
Belarus BLR 50.25 153 25.73 166 EECA Lower
Belgium BEL 87.43 12 80.12 18 EU & Balkans Lower
Belize BLZ 72.50 53 68.32 47 Americas Lower
Benin BEN 64.89 113 54.60 92 Africa Lower
Bhutan BTN 71.10 67 32.62 152 Asia-Pacific Lower
Bolivia BOL 64.63 114 54.09 93 Americas Lower
Bosnia and Herzegovina BIH 71.49 58 56.33 86 EU & Balkans Lower
Botswana BWA 76.44 39 57.64 81 Africa Lower
Brazil BRA 65.95 107 63.80 63 Americas Lower
Brunei BRN 50.35 152 53.47 97 Asia-Pacific Higher
Bulgaria BGR 64.94 111 60.78 70 EU & Balkans Lower
Burkina Faso BFA 76.53 38 52.25 105 Africa Lower
Burundi BDI 44.67 160 45.44 125 Africa Higher
Cambodia KHM 54.54 144 28.18 161 Asia-Pacific Lower
Cameroon CMR 56.72 134 42.75 131 Africa Lower
Canada CAN 84.71 16 78.75 21 Americas Lower
Cape Verde CPV 79.85 25 74.98 30 Africa Lower
Central African Republic CAF 57.13 132 60.15 72 Africa Higher
Chad TCD 60.30 123 51.89 108 Africa Lower
Chile CHL 72.69 51 62.25 69 Americas Lower
China CHN 21.52 177 14.80 178 Asia-Pacific Lower
Colombia COL 57.34 130 49.80 115 Americas Lower
Comoros COM 70.23 75 59.27 75 Africa Lower
Congo COG 63.44 118 60.58 71 Africa Lower
Costa Rica CRI 89.47 7 73.09 36 Americas Lower
Croatia HRV 71.49 59 64.20 60 EU & Balkans Lower
Cuba CUB 36.19 171 26.03 165 Americas Lower
Cyprus CYP 79.55 27 59.04 77 EU & Balkans Lower
Cyprus North CTU 70.21 77 54.84 91 EU & Balkans Lower
Czechia CZE 76.43 40 83.96 10 EU & Balkans Higher
DR Congo COD 50.91 150 42.31 133 Africa Lower
Denmark DNK 91.87 3 86.93 6 EU & Balkans Lower
Djibouti DJI 23.27 176 25.36 168 Africa Higher
Dominican Republic DOM 72.10 55 69.87 43 Americas Lower
East Timor TLS 70.10 78 71.79 39 Asia-Pacific Higher
Ecuador ECU 67.38 98 53.76 94 Americas Lower
Egypt EGY 43.18 166 24.74 170 MENA Lower
El Salvador SLV 70.30 74 41.19 135 Americas Lower
Equatorial Guinea GNQ 43.62 165 48.68 118 Africa Higher
Eritrea ERI 16.50 178 11.32 180 Africa Lower
Estonia EST 87.39 14 89.46 2 EU & Balkans Higher
Eswatini SWZ 54.85 141 52.86 98 Africa Lower
Ethiopia ETH 67.18 99 36.92 145 Africa Lower
Fiji FJI 72.59 52 71.20 40 Asia-Pacific Lower
Finland FIN 92.07 2 87.18 5 EU & Balkans Lower
France FRA 77.08 34 76.62 25 EU & Balkans Lower
Gabon GAB 62.80 121 70.65 41 Africa Higher
Gambia GMB 69.38 87 65.49 58 Africa Lower
Georgia GEO 71.41 60 50.53 114 EECA Lower
Germany DEU 87.84 11 83.85 11 EU & Balkans Lower
Ghana GHA 77.74 30 67.13 52 Africa Lower
Greece GRC 71.20 65 55.37 89 EU & Balkans Lower
Guatemala GTM 64.26 116 40.32 138 Americas Lower
Guinea GIN 65.66 110 52.53 103 Africa Lower
Guinea-Bissau GNB 67.94 94 51.36 110 Africa Lower
Guyana GUY 73.37 49 60.12 73 Americas Lower
Haiti HTI 69.80 83 51.06 111 Americas Lower
Honduras HND 51.80 148 38.51 142 Americas Lower
Hong Kong HKG 69.99 80 39.86 140 Asia-Pacific Lower
Hungary HUN 69.16 89 62.82 68 EU & Balkans Lower
Iceland ISL 84.88 15 81.36 17 EU & Balkans Lower
India IND 54.67 142 32.96 151 Asia-Pacific Lower
Indonesia IDN 63.18 119 44.13 127 Asia-Pacific Lower
Iran IRN 35.19 173 16.22 176 MENA Lower
Iraq IRQ 44.63 162 30.69 155 MENA Lower
Ireland IRL 87.40 13 86.92 7 EU & Balkans Lower
Israel ISR 69.16 88 51.06 112 MENA Lower
Italy ITA 76.31 41 68.01 49 EU & Balkans Lower
Ivory Coast CIV 71.06 68 63.69 64 Africa Lower
Jamaica JAM 89.49 6 75.83 26 Americas Lower
Japan JPN 71.14 66 63.14 66 Asia-Pacific Lower
Jordan JOR 57.92 128 35.25 147 MENA Lower
Kazakhstan KAZ 45.89 157 39.34 141 EECA Lower
Kenya KEN 66.28 103 49.41 117 Africa Lower
Kosovo XKX 70.67 70 52.73 99 EU & Balkans Lower
Kuwait KWT 65.70 109 44.06 128 MENA Lower
Kyrgyzstan KGZ 69.81 82 37.46 144 EECA Lower
Laos LAO 35.72 172 33.22 150 Asia-Pacific Lower
Latvia LVA 81.44 22 81.82 15 EU & Balkans Higher
Lebanon LBN 66.81 102 42.62 132 MENA Lower
Lesotho LSO 69.55 86 52.07 107 Africa Lower
Liberia LBR 67.75 95 66.61 54 Africa Lower
Libya LBY 44.23 164 40.42 137 MENA Lower
Liechtenstein LIE 80.48 24 83.42 12 EU & Balkans Higher
Lithuania LTU 78.81 28 82.27 14 EU & Balkans Higher
Luxembourg LUX 84.54 17 83.04 13 EU & Balkans Lower
Madagascar MDG 72.32 54 50.80 113 Africa Lower
Malawi MWI 70.68 69 59.20 76 Africa Lower
Malaysia MYS 66.88 101 56.09 88 Asia-Pacific Lower
Maldives MDV 70.07 79 52.46 104 Asia-Pacific Lower
Mali MLI 65.88 108 48.23 119 Africa Lower
Malta MLT 69.84 81 62.96 67 EU & Balkans Lower
Mauritania MRT 67.46 97 67.52 50 Africa Higher
Mauritius MUS 72.00 56 67.31 51 Africa Lower
Mexico MEX 54.55 143 45.55 124 Americas Lower
Moldova MDA 68.84 91 73.36 35 EECA Higher
Mongolia MNG 70.39 73 52.57 102 Asia-Pacific Lower
Montenegro MNE 66.17 105 72.83 37 EU & Balkans Higher
Morocco MAR 57.12 133 48.04 120 MENA Lower
Mozambique MOZ 66.21 104 52.63 101 Africa Lower
Myanmar MMR 55.23 139 25.32 169 Asia-Pacific Lower
Namibia NAM 80.75 23 75.35 28 Africa Lower
Nepal NPL 64.90 112 55.20 90 Asia-Pacific Lower
Netherlands NLD 90.04 5 88.64 3 EU & Balkans Lower
New Zealand NZL 89.31 9 81.37 16 Asia-Pacific Lower
Nicaragua NIC 64.19 117 22.83 172 Americas Lower
Niger NER 71.75 57 57.05 83 Africa Lower
Nigeria NGA 64.37 115 46.81 122 Africa Lower
North Korea PRK 14.18 180 12.64 179 Asia-Pacific Lower
North Macedonia MKD 68.72 92 70.44 42 EU & Balkans Higher
Norway NOR 92.16 1 92.31 1 EU & Balkans Higher
OECS CSS 76.22 44 68.08 48 Americas Lower
Oman OMN 56.58 135 42.29 134 MENA Lower
Pakistan PAK 54.48 145 29.62 158 Asia-Pacific Lower
Palestine PSE 55.91 137 27.41 163 MENA Lower
Panama PAN 70.22 76 66.75 53 Americas Lower
Papua New Guinea PNG 76.07 46 58.35 78 Asia-Pacific Lower
Paraguay PRY 67.03 100 56.84 84 Americas Lower
Peru PER 69.06 90 42.88 130 Americas Lower
Philippines PHL 56.46 136 49.57 116 Asia-Pacific Lower
Poland POL 71.35 62 74.79 31 EU & Balkans Higher
Portugal PRT 88.17 10 84.26 8 EU & Balkans Lower
Qatar QAT 57.49 129 58.25 79 MENA Higher
Romania ROU 74.09 48 66.42 55 EU & Balkans Lower
Russia RUS 51.08 149 24.57 171 EECA Lower
Rwanda RWA 49.66 155 35.84 146 Africa Lower
Samoa WSM 81.75 21 69.28 44 Asia-Pacific Lower
Saudi Arabia SAU 37.86 170 27.94 162 MENA Lower
Senegal SEN 76.01 47 59.43 74 Africa Lower
Serbia SRB 68.38 93 53.55 96 EU & Balkans Lower
Seychelles SYC 71.34 63 68.56 45 Africa Lower
Sierra Leone SLE 69.72 85 66.36 56 Africa Lower
Singapore SGP 44.77 158 45.78 123 Asia-Pacific Higher
Slovakia SVK 77.33 33 71.93 38 EU & Balkans Lower
Slovenia SVN 77.36 32 74.06 33 EU & Balkans Lower
Somalia SOM 44.55 163 40.49 136 Africa Lower
South Africa ZAF 77.59 31 75.71 27 Africa Lower
South Korea KOR 76.30 42 64.06 61 Asia-Pacific Lower
South Sudan SSD 55.51 138 51.63 109 Africa Lower
Spain ESP 77.84 29 77.35 23 EU & Balkans Lower
Sri Lanka LKA 58.06 127 39.93 139 Asia-Pacific Lower
Sudan SDN 44.67 159 30.34 156 Africa Lower
Suriname SUR 82.50 20 74.49 32 Americas Lower
Sweden SWE 90.75 4 88.13 4 EU & Balkans Lower
Switzerland CHE 89.38 8 83.98 9 EU & Balkans Lower
Syria SYR 27.43 174 15.82 177 MENA Lower
Tajikistan TJK 44.66 161 32.21 153 EECA Lower
Tanzania TZA 59.75 124 53.68 95 Africa Lower
Thailand THA 55.06 140 56.72 85 Asia-Pacific Higher
Togo TGO 70.67 71 48.03 121 Africa Lower
Tonga TON 72.73 50 68.39 46 Asia-Pacific Lower
Trinidad and Tobago TTO 76.78 36 79.71 19 Americas Higher
Tunisia TUN 70.55 72 43.48 129 MENA Lower
Turkey TUR 49.98 154 29.40 159 EECA Lower
Turkmenistan TKM 14.56 179 19.14 174 EECA Higher
Uganda UGA 59.05 125 37.61 143 Africa Lower
Ukraine UKR 67.48 96 63.93 62 EECA Lower
United Arab Emirates ARE 57.31 131 26.91 164 MENA Lower
United Kingdom GBR 77.07 35 78.89 20 EU & Balkans Higher
United States USA 76.15 45 65.49 57 Americas Lower
Uruguay URY 84.21 19 65.18 59 Americas Lower
Uzbekistan UZB 46.93 156 35.24 148 EECA Lower
Venezuela VEN 54.34 147 29.21 160 Americas Lower
Vietnam VNM 25.29 175 19.74 173 Asia-Pacific Lower
Yemen YEM 41.75 167 31.45 154 MENA Lower
Zambia ZMB 63.00 120 57.33 82 Africa Lower
Zimbabwe ZWE 59.05 126 52.10 106 Africa Lower

Getting your point across

Data viz as a rhetorical tool

Data viz as a rhetorical tool

The power of data viz



We can easily see patterns presented in certain ways, but if they are presented in other ways, they become invisible [..]

Following perception-based rules, we can present our data in such a way that the important and informative patterns stand out. If we disobey the rules, our data will be incomprehensible or misleading.


Ware, C. (2012). Information Visualization, Third Edition: Perception for Design

Data viz can be misleading

  • We have briefly discussed that before

  • There is a breadth of ways in which they can be misleading

  • Charts can be wrong. They can also be correct BUT misleading

  • See Defense Against Dishonest Charts on Flowing Data

Map projections




Cutting 0 on the y-axis

Key data viz principles

Theory

  • There is actually a lot of theory behind data viz:

    • Perception,
    • Colors,
    • Design,
    • etc

  • Worth learning about it and being aware of key principles

  • Leverage it to make better data viz

Perception

Ebbinghaus illusion

Relative differences matter

Law of simultaneous contrast

Gestalt principles



  • How our brain interprets what we see

  • How it organize visual information

  • How we group elements together

  • Use them to highlight some patterns and downplay others

Data-to-ink ratio




  • Introduced by Edward Tufte

  • In a nutshell: avoid clutter

  • Erase non-essential and redundant information

  • “Above all else show the data”

  • Sometimes good to break these rules

Data-to-ink ratio


\(\text{data-ink ratio} = \dfrac{\text{data ink}}{\text{total ink on graph}}\)


Graph aesthetics

  • Why make nice looking visualizations?

  • To trigger interest, to intrigue, to catch the eye

  • That affects how people perceive information

  • Nice looking visuals may be more memorable

  • In that sense, not everything is chartjunk

Pretty and memorable

Engaging the audience

Graph aesthetics in academia



  • Pretty graphs are useful in data journalism and so on, but what about academia?

  • They are also only more pleasing to look at, they also make readers want to engage more with them

  • Maybe better to keep the design rather minimalist

  • Pretty does not always mean non-simple. Simple graphs have value.

  • Opinion on credibility partly based on aesthetics

“This paper did not receive the care it deserved” comment given to a now senior researcher when they submitted a paper with a sketchy graph

Credibility and aesthetics

“Originality” VS “familiarity”

  • Original graphs may trigger interest

  • Familiar graphs may convey the point more easily

  • My take:

    • Use the best type of graph, regardless of its originality/familiarity

    • If it is different from what people are used to, make it easy to read

Building a graph

Know what chart types exist

Steps to choose a chart type

  • What do you want to show in your graph?

  • What is the main message you want to convey?

  • With the same data, you can:

    • Tell a lot of different stories

    • Emphasize different points

  • Making a graph = choosing a lighting for your data

1 dataset, many graphs, many stories

Type of relationship to show



In your graph, you may want to show a:


  • Distribution
  • Evolution over time
  • Magnitude
  • Part of a whole
  • Ranking
  • Geographical patterns
  • Flow
  • Correlations
  • Deviation

Graph type decision tree

Example data

country year lifeExp
France 1952 67.410
France 1957 68.930
France 1962 70.510
France 1967 71.550
France 1972 72.380
France 1977 73.830
France 1982 74.890
France 1987 76.340
France 1992 77.460
France 1997 78.640
France 2002 79.590
France 2007 80.657
Japan 1952 63.030
Japan 1957 65.500
Japan 1962 68.730
Japan 1967 71.430
Japan 1972 73.420
Japan 1977 75.380
Japan 1982 77.110
Japan 1987 78.670
Japan 1992 79.360
Japan 1997 80.690
Japan 2002 82.000
Japan 2007 82.603
Niger 1952 37.444
Niger 1957 38.598
Niger 1962 39.487
Niger 1967 40.118
Niger 1972 40.546
Niger 1977 41.291
Niger 1982 42.598
Niger 1987 44.555
Niger 1992 47.391
Niger 1997 51.313
Niger 2002 54.496
Niger 2007 56.867

A concrete example

Legible text

Title, clear axis labels and source

Scale of the y-axis

Stylize

An alternative story

Graphs in an oral presentation

  • Explain orally what your graph represents!

    • What is on the x-axis?

    • What is on the y-axis?

    • What is the message you want to convey?

Avoid using many different colors


If need more than 7 colors or so:


  • Use another graph

  • Group categories together

Use intuitive colors

Colorblind-friendly visualizations

Use gray. Emphasize.

Label directly

Order your data

Data viz caveats

Sometimes you may break the rules




9 colors


BUT


  • Labeled directly
  • Shade reinforces grouping

Nice data viz

Nice data viz

Pay attention to data viz

  • If you start paying attention to data viz when you see them, you will see what works, what does not

  • It is a process that takes place in the “background”: you will learn quickly and not realize it

  • You will see nice looking graphs (and hopefully enjoy it)

  • You will build better and more impactful graphs i

Summary of concrete recommendations



Take-away messages

  • Build legible, understandable and nice looking graphs.

  • Have a title and explicit axes; present them.

  • Limit the number of colors you use. Use gray.

  • Label your graphs directly, add annotations.

  • Think twice before cutting the y-axis

  • Overall, facilitate the retrieval of information.

Data viz in economics

When do we use graphs in econ?

  • As rhetorical visualization tools for models

  • To explore our data

  • To check the validity of our models

  • As diagnostics

  • To communicate results

Specificities of viz in economics

  • We often have to deal with a massive number of observations

  • We often want to display a specific type of graph: estimation output

  • Our analysis are often based on identifying assumptions that we can sometimes check through graphs

  • We often use very complex models. Visualization can help understand what we are actually estimating

Large numbers of observations

Opacity and subsamples

Use heatmaps instead of scatter plots

Table of estimation outputs

(1)
log(pop) 0.172
(0.021)
log(gdpPercap) 0.034
(0.014)
Num.Obs. 1704
R2 0.935
R2 Adj. 0.929
RMSE 0.06
Std.Errors by: country
FE: country X
FE: year X

Whisker plot of estimation outputs

Whisker and distribution

Graphs hierarchy

  • We make graphs for different audiences:

    • For you (and your future self): can be quite rough on the edges, but you will want to be able to understand it in the future

    • For presentations: you have some leeway for explaining orally your graphs

    • For the paper: there is only a couple of graphs in a paper; make them perfect

Main take-away points

Data viz at large



  • Data viz is powerful, harness its power

  • It can be super insightful or equally deceptive

  • It can make your point memorable

  • It can also be truly beautiful

  • Leverage perception and data viz principles

How to build a graph?



Take-away messages

  • There are many rules in data viz

  • The main goal is to facilitate the transmission of your message

  • What is the main point you want to convey?

  • Choose (one of) the right graph types

  • Explain your graphs, orally and by facilitating reading, on your graph

  • BUT avoid clutter

Data viz for economics

  • Most data viz principles and ideas also apply to economics and academia in general

  • There are however some specificities

  • In particular, some graphs and types of analyses are specific to academic research

  • Data viz can be extremely useful for research and communication

Thanks