Centre for Public Interest Advocacy
This workshop happened shortly after the Spending Data Handbook sprint and was a great opportunity to do a test run on the material.
What we covered
Some of the topics we covered were:
- An introduction to DataWrapper for making simple charts and web visualisations
- Kartograph for making elegant maps
- Scraping using ScraperWiki
- Using Optical Character Recognition to get data out of PDFs
- Cleaning data using Google Refine
Also present at this workshop were the team from Expert Grup in Moldova, prior to the launch of their Budget Stories project.
The Budzeti.ba project is still in progress and is due for a full launch in Autumn 2013. The aim of the project is to provide a one-stop shop for budget information in a format which is accessible for citizens.
Comprehensibility of budget data
The training in Bosnia was a trigger to build on the material in the Spending Data Handbook by developing some guidelines on how to make data published by governments more accessible for citizens. In practice, this is often a process of simplification and aggregating large datasets so as not to overwhelm the viewer. One of the key methodologies which was used to produce Where Does My Money Go? was to think carefully about how we aggregated the data and opting for functional classifications.
Demanding data in functional classifications: why and what’s difficult?
- Countries such as Bosnia do not publish Citizen’s Budgets in the first place. This means that a functional classification has to come entirely from civil society, leading to worries from the CSO that the interpretation of the data may be contested.
- Some countries do not group their data by functional classifications. This is important, as the average citizen is more likely to want to know what money is spent on (i.e. what services they got in exchange for their tax money) than, for example, which government department is spending the money, which is all it is possible to infer from many budgets.
- For the purposes of visualisations such as OpenSpending’s, organisations such as CPI must classify the complex information contained in budgets for themselves in a form which is accessible and yet not overwhelming for citizens. There are practical implications to this. Having more than 10 top-level items in a budget, for example, results in a very cramped visualisation, and there are only so many categories a person can take in at any one time. For visualisation purposes, the CPI team classified the data in a schema similar to the internationally recognised Classifications of Functions of Government, as was used in the UK project Where Does My Money Go?.
More information about the classification methodology they used can be found in the blog post: Bosnian Budgets - grouping data by categories people care about.