Mini project 2

Due by 11:59 PM on Monday, November 17, 2025

As of 2024, there are nearly 31 million refugees around the world. Refugee services and care remain a critical—but underfunded and often politically unpopular—public policy issue.

The UN Relief Agency, or UNHCR,¹ oversees and protects refugees. They collect detailed data on the number of forcibly displaced and stateless people around the world.

The {refugees} R package provides access to eight different datasets from the UNHCR and other refugee-focused agencies. You’ll be working with just one of them: population. This contains data on forcibly displaced and stateless persons by year, including refugees, asylum-seekers, returned refugees, internally displaced persons (IDPs), and stateless persons from 1951–2024.

For this mini project, you will use R, {ggplot2}, {dplyr}, and an external vector editing program like Inkscape, Affinity, or Adobe Illustrator create one plot that tells an interesting story from this UNHCR data. Full detailed instructions are included below.

Don’t worry! You’re not completely on your own! I’ve given you clean data, some starter code, and some possible ideas for summarizing and aggregating the data.

The data

This UNHCR data contains 19 columns. There were 16 columns in the original data. I used this code to add 3 more for you : date (useful for plotting), coo_region, and coa_region.

Columns in data. † = I added this column for you.
Variable	Description
`year`	The year (numeric, like 2020, 2021, etc.)
`date`^†	The date (a date, like 2020-01-01, 2021-01-01, etc.)^†
`coo_name`	Country of origin name
`coo`	Country of origin UNHCR code
`coo_iso`	Country of origin ISO code
`coo_region`^†	Country of origin UNCHR region^†
`coa_name`	Country of asylum name
`coa`	Country of asylum UNHCR code
`coa_iso`	Country of asylum ISO code
`coa_region`^†	Country of asylum UNCHR region^†
`refugees`	The number of refugees
`asylum_seekers`	The number of asylum-seekers
`returned_refugees`	The number of returned refugees
`idps`	The number of internally displaced persons
`returned_idps`	The number of returned internally displaced persons
`stateless`	The number of stateless persons
`ooc`	The number of others of concern to UNHCR
`oip`	The number of other people in need of international protection
`hst`	The number of host community members

Definitions

A country of origin (coo) is a country that sends out refugees, or one that refugees leave.
A country of asylum (coa) is a country that receives refugees, or one that refugees enter.

For example, in 2015, 115,604 refugees left Syria (coo) and entered Germany (coa):

unhcr_refugees |> 
  filter(coo_iso == "SYR", coa_iso == "DEU", year == 2015)
#> year              2015
#> coo_name          "Syrian Arab Rep."
#> coo_iso           "SYR"
#> coo_region        "Middle East and North Africa"
#> coa_name          "Germany"
#> coa_iso           "DEU"
#> coa_region        "Europe"
#> refugees          115604
#> asylum_seekers    81582
#> returned_refugees 0

Your general job

For this mini project, you will use R, {ggplot2}, {dplyr}, and an external vector editing program like Inkscape, Affinity, or Adobe Illustrator create one well-designed, enhanced plot that tells an interesting story from this UNHCR data. Specifically, you need to do these things:

Create a plot in R with {ggplot2} that shows something interesting from some aggregated or summarized version of the UNHCR data.
Save that plot as a PDF or SVG with ggsave() (just like in Exercise 10).
Open that exported plot in a vector editing program like Inkscape, Affinity, or Illustrator and add annotations and captions and other details to enhance and polish it and make it publication-worthy (just like in Exercise 10). Make sure you create a clear narrative/story.

I’d recommend doing as much polishing and refining as you can in R—make adjustments to the colors, scales, labels, grid lines, and even fonts, etc. Use the vector editing program to add annotations, change colors and fonts, and other final enhancements.

Full instructions

Open the project either on your computer or in Posit.cloud
- 02-mini-project.zip: If you’re using R on your own computer, download this file, unzip it, and double click on the file named 02-mini-project.Rproj
- Posit.cloud project: Use this link if you’re using Posit.cloud in your browser
Rename the Quarto file named your-name_mini-project-2.qmd to something that matches your name and open it in RStudio.
Explore the data and summarize it somehow. The raw data has more than 130,000 rows, which means you’ll need to aggregate the data (filter(), group_by(), and summarize() will be your friends). I include some ideas below.
Create one appropriate time-based visualization based on the data you summarized. Just make one plot. That’s all. One.
Save that plot as a PDF or SVG with ggsave()
Open that exported plot in a vector editing program like Inkscape, Affinity, or Illustrator and add annotations and captions and other details to enhance and polish it and make it publication-worthy. Make sure you create a clear narrative/story.
Write a memo (no word limit) explaining your process. I’m specifically looking for a discussion of the following:
- What story are you telling with your new graphic?
- How did you apply the principles of CRAP?
- How did you apply Kieran Healy’s principles of great visualizations or Alberto Cairo’s five qualities of great visualizations?
You can approach this in a couple different ways—you can write the memo and then include the full figure and code at the end, or you can write the memo in an incremental way, describing the different steps of creating the figure, ultimately arriving at a clean final figure.
Upload the following three (3) outputs to iCollege:
1. A PDF or Word file of your memo with your your final code, intermediate graphic (the one you create in R), and final graphic (the one you enhance) embedded in it, rendered with Quarto. Remember to use ![Caption](figure_file_name.pdf) to include external images in Markdown
2. A standalone PNG version of your final graphic. You’ll export this from the vector editing application.
3. A standalone PDF version of your graphic. You’ll export this from the vector editing application.

You will be graded based on completion using the standard ✓ system, but I’ll provide comments on how you use R and {ggplot2}, how well you apply the principles of CRAP, The Truthful Art, and Effective Data Visualization, and how appropriate the graph is for the data and the story you’re telling. I will use this rubric to make comments and provide you with a simulated grade.

02-mini-project_rubric.xlsx

For this assignment, I am more concerned with the design. Choose good colors. Choose good, clean fonts. Use the heck out of theme(). Add informative design elements in Illustrator/Inkscape/Affinity. Make it look beautiful and CRAPpy. Refer to the design resources here.

Please seek out help when you need it! You know enough R (and have enough examples of code from class and your readings) to be able to do this. Your project has to be turned in individually, and your visualization should be your own (i.e. if you work with others, don’t all turn in the same graph), but you should work with others! Reach out to me for help too—I’m here to help!

You can do this.

Data to possibly use in your plot

Data quality

Though the data goes back to 1951, many earlier numbers are loose estimates. For example, the number of refugees in the United States is a constant 500,000 (± 1–2,000) every year from 1951 to 1977; it starts showing actual numbers in 1978.

I’d recommend filtering the data for your visualization to start later than 1951 (like 1980, or 1990, or 2000, or whatever makes sense for the story you’re telling).

This data is incredibly rich and can be aggregated and collapsed and reshaped and visualized in all sorts of interesting ways. Here are some possible ideas:

Total refugees/asylum seekers/IDPs in one country of asylum (or more) over time
Total refugees/asylum seekers/IDPs from multiple countries of origin over time
Top N countries of origin in one country of asylum (or top N countries of asylum that one country of origin sends people to)—this could be used in something like a bump chart
Flows of refugees between regions over time (i.e. how many refugees from East Africa end up in Europe vs. the Americas vs. Asia, etc.)
Changes in counts of refugees sent or received in a country (or countries), so you can see when there are spikes and drops

You can visualize this stuff in many different ways! A line chart, a slope graph, a bump chart, a map, and so on. Do something neat!

Here are some examples of using different {dplyr} functions to filter, group, and summarize the data. PLEASE DON’T USE THESE VERBATIM. I don’t want to see a plot showing the count of refugees in Spain since 1990 or a plot of asylum seekers from Cambodia and Myanmar since 2000 in your final project (since I show that exact code below). Use these as general ideas.

library(tidyverse)
unhcr_refugees <- read_csv("data/unhcr_refugees.csv")

Total refugees in Spain every year since 1990

refugees_in_spain <- unhcr_refugees |> 
  filter(coa_iso == "ESP", year >= 1990) |> 
  group_by(year) |> 
  summarize(total = sum(refugees))
refugees_in_spain
## # A tibble: 35 × 2
##     year total
##    <dbl> <dbl>
##  1  1990  8303
##  2  1991  4005
##  3  1992  4040
##  4  1993  4990
##  5  1994  5351
##  6  1995  5840
##  7  1996  5672
##  8  1997  5512
##  9  1998  5918
## 10  1999  6708
## # ℹ 25 more rows

Changes in the total number of refugees in Spain every year since 1990

refugees_in_spain_diff <- unhcr_refugees |> 
  filter(coa_iso == "ESP", year >= 1990) |> 
  group_by(year) |> 
  summarize(total = sum(refugees)) |> 
  mutate(diff = total - lag(total))  # Subtract the previous value from each value
refugees_in_spain_diff
## # A tibble: 35 × 3
##     year total  diff
##    <dbl> <dbl> <dbl>
##  1  1990  8303    NA
##  2  1991  4005 -4298
##  3  1992  4040    35
##  4  1993  4990   950
##  5  1994  5351   361
##  6  1995  5840   489
##  7  1996  5672  -168
##  8  1997  5512  -160
##  9  1998  5918   406
## 10  1999  6708   790
## # ℹ 25 more rows

Total asylum seekers from Cambodia and Myanmar over time

refugees_from_se_asia <- unhcr_refugees |> 
  filter(coo_name %in% c("Cambodia", "Myanmar"), year >= 2000) |> 
  group_by(coo_name, year) |> 
  summarize(total = sum(asylum_seekers))
refugees_from_se_asia
## # A tibble: 50 × 3
## # Groups:   coo_name [2]
##    coo_name  year total
##    <chr>    <dbl> <dbl>
##  1 Cambodia  2000   301
##  2 Cambodia  2001   304
##  3 Cambodia  2002   465
##  4 Cambodia  2003   461
##  5 Cambodia  2004   501
##  6 Cambodia  2005   856
##  7 Cambodia  2006   668
##  8 Cambodia  2007   415
##  9 Cambodia  2008   212
## 10 Cambodia  2009   220
## # ℹ 40 more rows

Top 5 countries of origin in France (could be used in something like a bump chart)

top_coos_in_france <- unhcr_refugees |> 
  filter(coa_name == "France", year >= 2014) |> 
  group_by(year, coo_name) |> 
  summarize(total = sum(refugees)) |> 
  group_by(year) |> 
  slice_max(total, n = 5)
top_coos_in_france
## # A tibble: 55 × 3
## # Groups:   year [11]
##     year coo_name                             total
##    <dbl> <chr>                                <dbl>
##  1  2014 Unknown                              60000
##  2  2014 Sri Lanka                            23966
##  3  2014 Dem. Rep. of the Congo               13727
##  4  2014 Russian Federation                   13644
##  5  2014 Serbia and Kosovo: S/RES/1244 (1999) 12119
##  6  2015 Unknown                              68443
##  7  2015 Sri Lanka                            24220
##  8  2015 Russian Federation                   14195
##  9  2015 Dem. Rep. of the Congo               14182
## 10  2015 Serbia and Kosovo: S/RES/1244 (1999) 12500
## # ℹ 45 more rows

Footnotes

Formerly known as the UN High Commissioner for Refugees; they changed their name but kept the acronym.↩︎