First published on http://tugrul.dbsdataprojects.com on 24th of October 2016.
Being from a country where we had five elections/referendums in last 5 years, I have seen many different maps with the election results and I always wondered how they created those maps. I saw my first Fusion Table based map when Panama Papers hit the news. This map shows all the addresses from Ireland which were mentioned in Panama Papers and I remember admiring how quickly Gavin Sheridan created that map, literally 10 minutes after his first tweet about the addresses.
Although I was impressed with an efficient tool such as Fusion Tables, I haven’t used it until this year. My first attempt to create a map from a Fusion Table was during our Application of Cloud Technologies module, just a week before Data Management & Analytics class. It was a good preparation for Data Management & Analytics class and for this assignment. During the class, we have created a US Population Density Map and were asked to do a similar version of that map for Ireland.
To create the Irish Population Density map, I was given two different data and we have been asked to turn them into information. The first data set was from Central Statistics Office (CSO) website. 2011 Census population data was enough to get the population by county and gender. I have done a bit of cleansing, by removing break down of city data for big cities and get every county in one lane. Dublin, Cork, Galway, Waterford, and Limerick population were given by county and by City and County. I have removed those lines and I also added South Tipperary and North Tipperary data to one line. Also after uploading my KML file, I have realized that Laoighis was spelled differently in KML data, so I have changed that in my Fusion table into to Laois.
The second data set that I used was KML data to map out the county boundaries. Although we have been given a KML file, there were a few issues with that file so I have done a bit of research and found out that there is a website named as Ireland’s Open Data Portal which is supported by Government Departments and Public Bodies. I discovered a KML file with boundaries of counties published by Ordnance Survey Ireland. After downloading this KML file, I have uploaded it to Fusion Tables and merged it with Census data. After merging both tables, my map was ready. But to make my map more meaningful, I have decided to change filling colours based on population of the counties. I have checked the data and divided it into six different buckets. Dublin and Cork have had to their own buckets/colors as their population are far bigger than the rest of the countries. As you can see in the map below, each colour represents a different range of population.
As seen in the map, most of the population is located around the transport hubs Dublin, Cork, Galway and their neighbour counties such as Kildare, and Limerick. At the same time these counties have a great connection to the transport by sea. As %100 of the transportation was through sea via ferries and ships before planes become so popular, it is not a big surprise to see counties with that connection has more populated areas around them. These three counties constitute %44.52 of the Irish population. That unequal distribution of the population makes it harder to live in small counties as most of the government budget and investments are made around more populated areas. As an example, almost all motorways are directly linked to Dublin. Although there are two other Motorways (M18 & M20), they are just part of a national road and very limited when you compare to the rest of the other motorways.
This graph also emphasizes the fact that, if you are young male living in Ireland, you should either be living in Dublin or Waterford as these are the top two counties where there are more women than men.
More People or More Houses?
After gathering this data and map, I saw that 2016 Census preliminary data have been published by CSO. Therefore, I wanted to check population increase in the past five years to explain if there is a real housing crisis. I have downloaded the 2016 Census data and removed the columns except Total Number of Houses and Population data by county to see the change between 2011 and 2016.
Ireland has the 3rd biggest population increase, after Luxembourg and Cyprus, in EU between 2006 and 2015. Irish population has grown %10 within that time frame according to Eurostat data. Irish population has increased %3.70 from 2011 to 2016 while housing stock increased just %1.41. In 2011, for every 100 people there were 44.08 houses available whereas in 2016 this number dropped to 42.51. I have created another map with my new data and tried to show differences between population increase and housing supply increase. Dublin has the biggest difference after Longford. Longford’s population increased by %4.64 while house stock was decreased, but the entire population of the county is only 40,810. Same numbers are high for Dublin as Dublin’s population has increased %5.68 while house supply went up by only %1.32. On the other hand, Mayo, Donegal, and Sligo ‘s population were decreased although more houses were built. This information shows that people are still leaving counties that are not linked to big cities.
Census 2016 data shows that Government needs to act to increase housing supply. Although “Budget 2017” might increase the investment in new houses, the new Help to Buy Scheme for first time and new home buyers resulted in an increase of the current house prices. Government might need to take other measures such as decreasing VAT for residential construction for a limited period.
Conclusion – Fusion Tables
Google Fusion tables are very useful way to store online data and merge different data tables to create a map. Although it isn’t powerful as MS Access or SQL tables, the ability to create a map based on your data makes it very convenient. It can be used by businesses and newsroom, journalists etc. to make meaningful visualizations.
References & Sources: