Viewer's Choice
Job Gains and Losses
by Furqan Farooqui
(Last updated Nov 16th, 2015)

Introduction

Dear readers! Here's my analysis of a visualization I found on the web, as part of my Viewer's Choice submission for CS 424. This particularly interesting visualization shows us how the recession affected various sectors, and lets us see a bird's eye view of which industries are seeing job growth, as well as which ones are getting straitened.
Link to visualization

What's in the Data?

The data is mainly a record of the number of jobs in various sectors over the past several decades. These jobs have been organized into various high-level sectors. The other piece of data collected is the unemployment rate over time, for various demographics.

The data has been sourced from:

Highlights

There's a lot to like in this visualization. There's so much information packed into one nice graph, without making it look cluttered. At first glance, one can easily see the significant impact of a recession. The recession has been clearly marked too, which is a nice touch. The data that hasn't been changed has been shown seperately as grey boxes, separate from the main data. The data points are shown as boxes as opposed to continuous representations, which makes it even more appealing and accurate in my opinion.

As with any good visualization, this one is interactive too. Hovering over a data point shows a pop up of information showing the info in text. It simultaneously highlights all the datapoints for that sector, to be able to see the trends over time for that sector.

Clicking on any column shows even more data separately as a bar chart, neatly placed in a table to be able to show even more information. Also, the bar chart provides us additional visualization goodness, by showing relatively how bad some sectors faired compared to others, which wasn't fully appreciable with just colours shown in the grand over graph above.

Finally, to sum it up, we contrast the sector growth and shrinkage with the impact it had on unemployment, i.e., the number of people affected. This is shown in two graphs, one as a color-coded graph, that complements the graph above nicely, as well as a simple line graph to show relative highs and lows in unemployment over time. There are various controls below that to change gender, age and other demographic parameters to see how various sections of society were affected.

Improvements

No visualization is perfect, so there are always ways to make it more useful, or add more features. Below are some of the points of improvement that I could think of:

⚫ Highlighting a datapoint, one can notice that sometimes, sectors seem to spread around wildly from the origin. It's hard to tell which sectors seems to be really struggling for a prolonged period of time, and which ones just saw a short spike in sectors failing. Maybe it would help if we could adjust the individual columns' time-interval to be less granular.

⚫ In the barchart, we can see the size of individual sectors. But it would've been nice if the growth of shrinkage of jobs in those sectors was shown visually in the barchart itself, as opposed to just a percentage.

Audience

The users for this visualization are many. Economists, casual curious observers, and maybe even researchers. This visualization does do a great job of making data more accessible to a wide variety of people. Recession is no longer just a word, but one can see visually the impact in hard numbers.

Insights

This visualization helps us answer several questions.

Which sectors were worsts affected? This can be seen by looking at the high level graph, and by hovering over the red dots to see the individual sectors.

It also answers questions about individual sectors, and one can even deduce trends for those sectors in the future.

Other interesting answers include trend about how unmployment hits various genders and demographics.