More Emigration Trends

I was curious about where people are going when they leave Denmark, and who they are. I got data from VAN2AAR and VAN1AAR on https://statbank.dk/

# Importing my libraries
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

from matplotlib.sankey import Sankey
import matplotlib.animation as animation

import squarify
import matplotlib.cm as cm

Ukraine

Ukraine is a popular destination for people leaving Denmark. As you can see, most people moving to Ukraine are from Ukraine.

ukraine_df = pd.read_csv('destination ukraine.csv', header=None, encoding='ISO-8859-1')
ukraine_df.columns = ['Destination', 'Year', 'Citizenship', 'Region Hovedstaden',	'Region Sjælland',	'Region Syddanmark',	'Region Midtjylland',	'Region Nordjylland']

ukraine_df.head()

	Destination	Year	Citizenship	Region Hovedstaden	Region Sjælland	Region Syddanmark	Region Midtjylland	Region Nordjylland
0	Ukraine	2023	Ukraine	928	678	960	924	631
1	Ukraine	2023	Non-Ukrainians	9	13	7	6	6

Sweden and Germany

A lot of internationals are moving to Sweden and Germany. Who are they and where were they living before?

germany_df = pd.read_csv('destination germany.csv', header=None, encoding='ISO-8859-1')
germany_df.columns = ['Destination', 'Year', 'Citizenship', 'Region Hovedstaden',	'Region Sjælland',	'Region Syddanmark',	'Region Midtjylland',	'Region Nordjylland']

germany_df.head()

	Destination	Year	Citizenship	Region Hovedstaden	Region Sjælland	Region Syddanmark	Region Midtjylland	Region Nordjylland
0	Germany	2023	Germany	862	92	533	339	106
1	Germany	2023	Non-Germans	642	160	477	499	170

# Plotting
regions = ['Region Hovedstaden', 'Region Sjælland', 'Region Syddanmark', 'Region Midtjylland', 'Region Nordjylland']
citizenship = germany_df['Citizenship']

# Set bar width and colors
bar_width = 0.5
colors = ['teal', 'lightblue']

# Calculate total heights for each region
total_heights = germany_df.iloc[0, 3:] + germany_df.iloc[1, 3:]

# Sort regions by total heights in descending order
sorted_indices = total_heights.sort_values(ascending=False).index
sorted_regions = [region for region in sorted_indices]

# Set the position of bars on the x-axis (fixing the positions variable to use sorted regions)
positions_sorted = range(len(sorted_regions))

fig, ax = plt.subplots()

# Create stacked bar chart with sorted regions
ax.bar(positions_sorted, germany_df[sorted_regions].iloc[0], bar_width, label=citizenship[0], color=colors[0])
ax.bar(positions_sorted, germany_df[sorted_regions].iloc[1], bar_width, bottom=germany_df[sorted_regions].iloc[0], label=citizenship[1], color=colors[1])

# Adding labels and title
ax.set_xlabel('Regions')
ax.set_ylabel('Population')
ax.set_title('Emigrants to Germany in 2023 by Citizenship and Region')
ax.set_xticks(positions_sorted)  # Fix: positions_sorted should be used here instead of positions
ax.set_xticklabels(sorted_regions, rotation=45, ha="right")  # Fix: sorted_regions instead of regions
ax.legend()

# Display the plot
plt.tight_layout()
plt.show()

sweden_df = pd.read_csv('destination sweden.csv', header=None, encoding='ISO-8859-1')
sweden_df.columns = ['Destination', 'Year', 'Citizenship', 'Region Hovedstaden',	'Region Sjælland',	'Region Syddanmark',	'Region Midtjylland',	'Region Nordjylland']

sweden_df.head()

	Destination	Year	Citizenship	Region Hovedstaden	Region Sjælland	Region Syddanmark	Region Midtjylland	Region Nordjylland
0	Sweden	2023	Sweden	774	62	88	88	41
1	Sweden	2023	Non-Swedes	549	57	48	61	23

# Plotting
regions = ['Region Hovedstaden', 'Region Sjælland', 'Region Syddanmark', 'Region Midtjylland', 'Region Nordjylland']
citizenship = sweden_df['Citizenship']

# Set bar width and colors
bar_width = 0.5
colors = ['teal', 'lightblue']

# Calculate total heights for each region
total_heights = sweden_df.iloc[0, 3:] + germany_df.iloc[1, 3:]

# Sort regions by total heights in descending order
sorted_indices = total_heights.sort_values(ascending=False).index
sorted_regions = [region for region in sorted_indices]

# Set the position of bars on the x-axis (fixing the positions variable to use sorted regions)
positions_sorted = range(len(sorted_regions))

fig, ax = plt.subplots()

# Create stacked bar chart with sorted regions
ax.bar(positions_sorted, sweden_df[sorted_regions].iloc[0], bar_width, label=citizenship[0], color=colors[0])
ax.bar(positions_sorted, sweden_df[sorted_regions].iloc[1], bar_width, bottom=sweden_df[sorted_regions].iloc[0], label=citizenship[1], color=colors[1])

# Adding labels and title
ax.set_xlabel('Regions')
ax.set_ylabel('Population')
ax.set_title('Emigrants to Sweden in 2023 by Citizenship and Region')
ax.set_xticks(positions_sorted)  # Fix: positions_sorted should be used here instead of positions
ax.set_xticklabels(sorted_regions, rotation=45, ha="right")  # Fix: sorted_regions instead of regions
ax.legend()

# Display the plot
plt.tight_layout()
plt.show()

Americans

Many Americans immigrate to Denmark but how many go back to the USA? Let's look at the trends over the past 17 years

american_in_df = pd.read_csv('americans arriving.csv', header=None, encoding='ISO-8859-1')

year_columns = {i: 2006 + i for i in range(1, 18)}  # 1 corresponds to 2007, 17 to 2023


american_in_df.rename(columns=year_columns, inplace=True)

# Drop the 'citizenship' column (which is the first one)
american_in_df.drop(columns=[0], inplace=True)

# Adding 'Migration' column with the value 'immigration'
american_in_df['Migration'] = 'immigration'

# Rearrange columns so 'Migration' is the first column
american_in_df = american_in_df[['Migration'] + list(american_in_df.columns[:-1])]

american_out_df = pd.read_csv('americans going home.csv', header=None, encoding='ISO-8859-1')

american_out_df.drop(american_out_df.columns[0:3], axis=1, inplace=True)
year_columns = {i: 2006 + i -2 for i in range(2, 22)}
american_out_df.rename(columns=year_columns, inplace=True)
# Adding 'Migration' column with the value 'emigration'
american_out_df['Migration'] = 'emigration'

# Rearrange columns so 'Migration' is the first column
american_out_df = american_out_df[['Migration'] + list(american_out_df.columns[:-1])]

combined_df = pd.concat([american_in_df, american_out_df], ignore_index=True)
combined_df.head()

	Migration	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022	2023
0	immigration	1980	2264	2372	2665	2721	2848	3249	3290	3760	3859	3760	3946	3810	1392	1504	3829	4168
1	emigration	1176	1928	1822	2134	1415	3053	2562	2900	2429	3469	2480	3346	3318	882	640	2247	2995

years_columns = combined_df.columns[1:18]  # Exclude 'Migration' 

# Separate immigration and emigration rows
immigration_df = combined_df[combined_df['Migration'] == 'immigration']
emigration_df = combined_df[combined_df['Migration'] == 'emigration']

# Calculate net migration: immigration - emigration
net_migration_values = immigration_df[years_columns].iloc[0] - emigration_df[years_columns].iloc[0]

# Set positions for the bars
years = np.arange(len(years_columns))

# Set the figure size
fig, ax = plt.subplots(figsize=(10, 6))

# Plot immigration as positive bars
ax.bar(years, immigration_df[years_columns].iloc[0], width=0.4, label='Immigration', color='#76c7c0')

# Plot emigration as negative bars
ax.bar(years, -emigration_df[years_columns].iloc[0], width=0.4, label='Emigration', color='#f28e8e')

# Plot net migration as a line on top
ax.plot(years, net_migration_values, label='Difference', color='#4b88a2', marker='o')

# Adding labels and title
ax.set_xlabel('Years')
ax.set_ylabel('Population')
ax.set_title('US Citizens Immigration and Emigration back to USA')
ax.set_xticks(years)
ax.set_xticklabels(years_columns, rotation=45)
ax.legend()

# Display the plot
plt.tight_layout()
plt.show()

Denmark is attractive as a destination for American immigrants but many go back to the USA. Note: I'm not counting the "serial expat" figures where Americans go to another country after Denmark.

Italians

I noticed something interesting in the statistics, in 2023, Italians were the biggest group expatriating to another country after trying Denmark. I wanted to dig into that a little bit. Where are they going?

# Source: VAN2AAR on https://statbank.dk/

italian_df = pd.read_csv('italian emigrants.csv', header=None, encoding='ISO-8859-1')

column_renaming = {
    0: 'Citizenship',
    1: 'Destination Country'
}

# Add year renaming to the dictionary
year_columns = {i: 2005 + i for i in range(2, 19)}  # Maps columns 2-18 to 2007-2023
column_renaming.update(year_columns)  # Update the dictionary with the year mapping

# Rename the columns in the DataFrame
italian_df.rename(columns=column_renaming, inplace=True)

italian_df.head()

	Citizenship	Destination Country	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022	2023
0	Italy	Faroe Islands	0	0	0	0	0	0	0	1	0	0	0	0	0	1	0	0	0
1	Italy	Greenland	0	0	0	0	1	0	1	0	0	0	1	0	2	3	1	1	2
2	Italy	Albania	0	0	0	0	0	0	0	1	0	0	0	0	2	1	0	0	0
3	Italy	Andorra	0	0	0	0	0	0	0	0	0	0	0	2	0	0	0	0	1
4	Italy	Belarus	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0

# Specify the year columns (2007-2023) to sum
year_columns = [i for i in range(2007, 2024)]  # This will give the list of year columns (2007 to 2023)

# Calculate the row-wise sum for the year columns
italian_df['Total'] = italian_df[year_columns].sum(axis=1)

# Filter out rows where the total sum is less than 100, and create a copy to avoid the warning
filtered_df = italian_df[italian_df['Total'] >= 100].copy()

# Sort the dataframe by the 'Total' column in descending order
sorted_df = filtered_df.sort_values(by='Total', ascending=False).reset_index(drop=True)

# Use the new method to get the colormap
cmap = plt.colormaps.get_cmap('cividis')  # You can also try 'plasma', 'cividis', etc.

# Normalize the 'Total' values to map to the colormap
norm = plt.Normalize(vmin=min(sorted_df['Total']), vmax=max(sorted_df['Total']))
colors = [cmap(norm(value)) for value in sorted_df['Total']]

# Create the treemap with the new color scheme
plt.figure(figsize=(12, 8))
squarify.plot(sizes=sorted_df['Total'], label=sorted_df['Destination Country'], color=colors, alpha=0.8)

# Add title and display
plt.title('Italian Emigration by Destination Country (Based on Total)', fontsize=16)
plt.axis('off')  # Remove axes for clarity
plt.show()

# After sorting and resetting the index, drop the 'Total' column once
sorted_df.drop(columns=['Total'], inplace=True)

# Define the years (2007-2023) as integers
years = [i for i in range(2007, 2024)]

# Calculate the total migration per destination across all years
sorted_df['Total'] = sorted_df[years].sum(axis=1)

# Get the top 5 destination countries based on total migration
top_5_destinations = sorted_df.nlargest(5, 'Total')

# Define a visually distinct color palette (for top 5 destinations)
colors = plt.colormaps.get_cmap('Set2')  # Tab10 provides up to 10 distinct colors

# Set the figure size
plt.figure(figsize=(12, 8))

# Plot stacked bars for each year
bottom_values = np.zeros(len(years))  # Initialize bottom values for stacking

for i, destination in enumerate(top_5_destinations['Destination Country']):
    plt.bar(years, top_5_destinations[top_5_destinations['Destination Country'] == destination][years].sum(), 
            bottom=bottom_values, color=colors(i), label=destination)
    # Update bottom values for stacking the next destination on top
    bottom_values += top_5_destinations[top_5_destinations['Destination Country'] == destination][years].sum()

# Add labels and title
plt.xlabel('Year', fontsize=12)
plt.ylabel('Total Migration', fontsize=12)
plt.title('Where Italians are Going (Top 5 Destination Countries)', fontsize=16)

# Force x-axis ticks to be integers (years)
plt.xticks(years, [str(year) for year in years])  # Convert years to strings to avoid float formatting

# Add legend
plt.legend(title='Destination Country', bbox_to_anchor=(1.05, 1), loc='upper left')

# Display the plot
plt.tight_layout()
plt.show()

# Create subplots arranged in a 2x2 grid
fig, axes = plt.subplots(2, 2, figsize=(10, 8), sharex=True, sharey=True)

# Flatten axes to easily iterate over them
axes = axes.flatten()

# Get the top 4 destination countries based on total migration
top_4_destinations = sorted_df.nlargest(4, 'Total')

# Plot for each of the top 4 destinations
for i, destination in enumerate(top_4_destinations['Destination Country']):
    ax = axes[i]
    # Plot the trend line (blue) for each destination without dots
    ax.plot(years, top_4_destinations[top_4_destinations['Destination Country'] == destination][years].values.flatten(), 
            color='blue')

    # Set title for each subplot with the destination name
    ax.set_title(destination)

    # Set x-axis to only show 2007 and 2023, with labels
    ax.set_xticks([2007, 2023])
    ax.set_xticklabels([2007, 2023])

    # Add 0 and 1200 to y-axis, with labels
    ax.set_yticks([0, 1200])
    ax.set_yticklabels([0, 1200])

# Set a common scale for all subplots (same x and y limits)
plt.xlim([2007, 2023])
plt.ylim([0, 1200])

# Add a tight layout for better spacing
plt.tight_layout()

# Display the plot
plt.show()

Let's take a look at net migration for Italians, as in, total immigration figures compared to emigration

italian_in_df = pd.read_csv('italians arriving.csv', delimiter=';', header=None, encoding='ISO-8859-1')

year_columns = {i: 2006 + i for i in range(1, 18)}  # 1 corresponds to 2007, 17 to 2023


italian_in_df.rename(columns=year_columns, inplace=True)

# Drop the 'citizenship' column (which is the first one)
italian_in_df.drop(columns=[0], inplace=True)

# Adding 'Migration' column with the value 'immigration'
italian_in_df['Migration'] = 'immigration'

# Rearrange columns so 'Migration' is the first column
italian_in_df = italian_in_df[['Migration'] + list(italian_in_df.columns[:-1])]

italian_out_df = pd.read_csv('italian out.csv', header=None, encoding='ISO-8859-1')


year_columns = {i: 2006 + i for i in range(1, 18)}  # 1 corresponds to 2007, 17 to 2023

italian_out_df.rename(columns=year_columns, inplace=True)
# Adding 'Migration' column with the value 'emigration'
italian_out_df['Migration'] = 'emigration'

# Rearrange columns so 'Migration' is the first column
italian_out_df = italian_out_df[['Migration'] + list(italian_out_df.columns[:-1])]

italy_combined_df = pd.concat([italian_in_df, italian_out_df], ignore_index=True)
italy_combined_df

	Migration	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022	2023	0
0	immigration	866	928	998	1094	1122	1347	1597	2050	2064	2091	2122	2239	2453	2291	3292	3978	3432	NaN
1	emigration	460	540	800	741	827	731	919	1278	1160	1256	1608	1716	1821	1635	1480	2057	2306	Italy

combined_df = italy_combined_df
years_columns = combined_df.columns[1:18]  # Exclude 'Migration' 

# Separate immigration and emigration rows
immigration_df = combined_df[combined_df['Migration'] == 'immigration']
emigration_df = combined_df[combined_df['Migration'] == 'emigration']

# Calculate net migration: immigration - emigration
net_migration_values = immigration_df[years_columns].iloc[0] - emigration_df[years_columns].iloc[0]

# Set positions for the bars
years = np.arange(len(years_columns))

# Set the figure size
fig, ax = plt.subplots(figsize=(10, 6))

# Plot immigration as positive bars
ax.bar(years, immigration_df[years_columns].iloc[0], width=0.4, label='Immigration', color='#76c7c0')

# Plot emigration as negative bars
ax.bar(years, -emigration_df[years_columns].iloc[0], width=0.4, label='Emigration', color='#f28e8e')

# Plot net migration as a line on top
ax.plot(years, net_migration_values, label='Net Migration', color='#4b88a2', marker='o')

# Adding labels and title
ax.set_xlabel('Years')
ax.set_ylabel('Population')
ax.set_title('Immigration, Emigration, and Net Migration Over Time of Italians')
ax.set_xticks(years)
ax.set_xticklabels(years_columns, rotation=45)
ax.legend()

# Display the plot
plt.tight_layout()
plt.show()

Not many Italians are staying in Denmark. What is going on with Argentina though?

import matplotlib.pyplot as plt
import numpy as np

# Years from 2007 to 2023
years = [i for i in range(2007, 2024)]

# Italians moving from Argentina (2007-2023)
from_argentina = [3, 5, 8, 5, 5, 8, 10, 7, 13, 25, 35, 65, 215, 197, 469, 596, 354]

# Italians moving to Argentina (2007-2023) - negative
to_argentina = [-1, -5, -2, -4, -4, -2, -5, -7, -8, -7, -11, -22, -64, -59, -109, -271, -226]

# Calculate net migration (from - to)
net_migration = np.array(from_argentina) + np.array(to_argentina)

# Create the area chart
plt.figure(figsize=(10, 6))
plt.fill_between(years, to_argentina, color="lightblue", alpha=0.5, label="Italians Moving To Argentina")
plt.fill_between(years, from_argentina, color="blue", alpha=0.5, label="Italians Moving From Argentina")

# Plot the net migration as a grey line
plt.plot(years, net_migration, color="grey", linewidth=2, linestyle="--", label="Difference")

# Add labels and title
plt.xlabel('Year')
plt.ylabel('Number of Migrants')
plt.title('Migration of Italians To and From Argentina (2007-2023)')
plt.legend()

# Display the chart
plt.show()

Argentina has a huge Italian diaspora, 62% of the population of Argentina have Italian heritage. A lot of them have dual citizenship, and of course it is easier to move to Denmark on an EU passport.