This Week: Three data lessons from Kenya’s 2009 census
I stumbled upon a fascinating post about birth certificates in Kenya last week. According to Brett Keller’s post, since many Kenyans don’t have a birth certificate and might not be aware of the exact date they were born, the 2009 census provided a list (pg. 60) for each region of big events that individuals might know took place near their birth date. This context helped citizens to better estimate their date of birth and provided the Government of Kenya with some important data, although accuracy was limited.
Reading this was a refreshing reminder of the context within which we work and, I feel, showcases some key ideas to remember when we start talking about our desired data revolution.
Start where you are. This method used will not provide accurate birth dates. However, given the current limitations, it does provide a better understanding of the country’s demographics and can still be used for decision makers, given the caveat of data quality. At times we are so worried about the quality of data collected that we want to wait until it’s perfect. We need to remember that there are still things we can learn from imperfect data. It’s usually not too far off and can provide a helpful picture of things. It also highlights where we need to improve our data quality.
One of the key points made during the Talking About a Data Revolution event at the World Bank last week, was that making data public tends to help improve data quality because more eyes are on it, and more voices can speak up about it. Getting over the fear of publishing imperfect data can help lead us to better data.
Courtesy Photo via DevEx, Steven Shapiro / World Bank
Make it easier for your future self: don’t repeat the same mistakes. This example shows a great need for Kenyans to improve birth certificate availability to its citizens and should be quickly addressed. Let’s not repeat our same mistakes, but instead let's adapt our processes to make it easier to get the high quality data we need.
The DATA Act in the US just passed the Senate and would improve access to government expenditures in order to hold government accountable. This data previously has not been publicly available in a usable format. "The DATA Act takes a structured data model that has delivered unprecedented accountability in stimulus expenditures and applies it across all domains of federal spending," said Data Transparency Coalition Executive Director Hudson Hollister. Read more about the progress of the DATA Act here.
Think about your audience first. The creators for the census knew their audience: citizens who had no birth certificates, little record keeping, and a focus on the goings on in their local community. This allowed them to create a method for data collection that would make sense to their audience. This is applicable to data collectors and data users alike. Who is our audience, and what is the context they work in? Are our tools usable for government technical staff? Are we asking questions that they can even answer? Are we answering the questions that they would ask?
This will be especially important for the new “Data Native” generation who hold high data expectations. We aren’t always are own audience and adjusting for context will always be key.
Taryn Davis is an Assoicate with Development Gateway based in Washington D.C.