Migration2013

Introduction

The purpose of this report to explore the bilateral migration data available at KNOMAD. The Global Knowledge Partnership on Migration and Development (KNOMAD) is a global hub of knowledge and policy expertise on migration and development issues.

The dataset utilized in this report can be accessed here.

The report extracts data pertinent to the emigration from the USA for the exploratory analysis and plotting. A companion Shiny app allows the user to select a country for emigration or immigration data.

---

Data load and preparation

library(readxl)  #http://readxl.tidyverse.org/

## Warning: package 'readxl' was built under R version 3.4.3

library(dplyr)
library(data.table)

## 
## Attaching package: 'data.table'

## The following objects are masked from 'package:reshape2':
## 
##     dcast, melt

## The following objects are masked from 'package:dplyr':
## 
##     between, first, last

library(ggplot2)
library(plotly)


download.file("https://www.knomad.org/sites/default/files/2017-03/bilateralmigrationmatrix20130.xlsx","Mig2013.xlsx", mode="wb")
Mig2013 = read_excel("Mig2013.xlsx",range="A2:HJ219")
names(Mig2013)[1]="Country"  # shorten the col name

# Select the US for the focus of the exploratory analysis, 
#    

Mig2013US = Mig2013 %>% filter(Country == "United States")  # extracting data points pertinent to US

Mig2013US = melt(Mig2013US)

## Using Country as id variables

Mig2013US$Country = NULL  # remove the first col whose value is "United States" for all rows, thus meaningless

names(Mig2013US) = c("Country","NumMig")
Mig2013US = Mig2013US %>% filter(NumMig>0)

Now, the dataset needs to be augmented with country codes. With a quick Google search, I found a source for Country Codes. Unfortunately, it did not cover all countries represented in the dataset downloaded from KNOMAD. Further research led me to the countrycode R package which turned out to be a better choice.

CountryCodes=read.csv("https://raw.githubusercontent.com/plotly/datasets/master/2014_world_gdp_with_codes.csv")

names(CountryCodes)[1]="Country"

Mig2013US2= merge(x=Mig2013US,y=CountryCodes,by="Country",all.x=TRUE)
# ref: https://stackoverflow.com/questions/1299871/how-to-join-merge-data-frames-inner-outer-left-right
# This returned about 20 countries with no code.

# Fortunately,I found the countrycode package. 

library(countrycode)

## Warning: package 'countrycode' was built under R version 3.4.3

Mig2013US2 = Mig2013US %>% mutate(Code=countrycode(Country,"country.name","iso3c"))

## Warning in countrycode(Country, "country.name", "iso3c"): Some values were not matched unambiguously: Other North, Other South, World

# A much better result.  It is still coded north korea incorrectly.  No country codes for Other North and Other South.  This should be noted in the Shiny app.

Mig2013US3 = Mig2013US2 %>% 
               filter(!is.na(Code)) %>% # remove countries with no code
               mutate(Code=ifelse(grepl("^Korea,( +)Dem",Country),"PRK",Code)) %>% # fix north korea's code
               mutate(hoverpop = paste(as.character(round(NumMig/1000,1)),
                                       "K moved to ",
                                       Country,sep="")) # compose the tooltip text

Plotting - Emigration

Let’s plot a world map visualizing how many US citizens have moved to other countries.

borders = list(color= toRGB("black"),width=1)
map_options = list(
   scope="world",
   projection = list(type='Mercator'),
   showcoastlines = TRUE
   ) # borders
     
p3 = plot_ly(z   = ~Mig2013US3$NumMig,
            text = ~Mig2013US3$hoverpop,
            locations = ~Mig2013US3$Code,
            type ="choropleth",
            locationmode='world',
            color=~Mig2013US3$NumMig,
            colors="Blues",
            marker = list(line=borders)) %>%
   
     colorbar(title = "Migration Outflow") %>%
   
     plotly::layout(title='US citizens migrate out to (Hover for breakdown)', geo=map_options)
     
p3

Sort the data frame on the migration outflow numbers in descending order to list top 10 countries the US citizens moved to.

Mig2013US3.desc = Mig2013US3 %>% arrange(desc(NumMig))
head(Mig2013US3.desc,10)

##           Country NumMig Code                       hoverpop
## 1          Mexico 848576  MEX         848.6K moved to Mexico
## 2          Canada 316649  CAN         316.6K moved to Canada
## 3  United Kingdom 222201  GBR 222.2K moved to United Kingdom
## 4     Puerto Rico 188954  PRI      189K moved to Puerto Rico
## 5         Germany 146677  DEU        146.7K moved to Germany
## 6       Australia  89957  AUS         90K moved to Australia
## 7          Israel  80463  ISR          80.5K moved to Israel
## 8     Korea, Rep.  71817  KOR     71.8K moved to Korea, Rep.
## 9           China  71493  CHN           71.5K moved to China
## 10          Italy  60037  ITA             60K moved to Italy

---

Companion Shiny App

As mentioned earlier in the report, you may take a look at the Shiny app that allows you to select a country for emigration or immgration data.