I’d like to share with you a few of the activities that we conducted in Mexico to fill in the open data census in four different cities (Mexico City, Guadalajara, Monterrey, Xalapa). It’s all anecdotal, but I hope you find it useful.
In Guadalajara, the event took place in the 9th of december in a local hackerspace called Hacker Garage. The municipalities that were chosen to work on were Guadalajara and Zapopan. More than 20 people assisted to the event, including government officials from the respective departments that promoted the use of open data (Departamento de Innovación Municipal de Guadalajara and Zapopan Lab).
We started with a tutorial on how to fill-in the different fields of the census. Afterwards, two teams were formed, where each team worked in a municipality. The teams went into the websites of Zapopan and Guadalajara and reviewed if those datasets were available, and at least in an Excel format, instead of just PDF. The census was completed.
The group was divided into two smaller groups, one for each municipality, and had people delve into the “transparency” section of the government sites, and even search on Google, but making sure the sites searched were from the local government. Within each municipality team, people gathered in subteams of 2-3 people to search for each of the datasets to be searched for.
The good stuff
Teamwork. Something that could have taken a long time to do by one or two people, was a good excuse to have our end of the year party. The task was finished at hand, but people were so excited they wanted to keep looking in the sites. A sense of awareness was created about the status of open data in each municipality and had a group event that was good for the community.
What can be improved
Getting organized civil society involved.The group of volunteers that worked on completing the census was a group of volunteers (with few or no expertise on specific topics) and some government officials. For future events, we believe it would be necessary to involve more NGOs and organizations to work on the field to provide a better insight into the completeness and relevance of the data available in each municipality.
In Monterrey, the event took place the 16th of december in a coworking space called The Pool. The event had a participation of around 20 assistants. The activities started with a small workshop about the datum that was going to be surveyed and the methodology that was to be followed to conduct the survey. The census was completed for three of the municipalities of Monterrey-metro area: Monterrey, San Pedro Garza García, and San Nicolás de los Garza. These municipalities were chosen since a.) they are the most important cities of the metro area, and b.) they have at least one major university.
We ran a one time event that tried to cover as much of the metropolitan area of Monterrey at once. We had a fairly good crowd (around 20 people) that we divided with as much experts to appraise the datasets related to each municipality. Each team was comprised sometimes of government officials/members (Of which we had 3 members that covered that profile), so in teams we tried to find as much information for the city that we could.
The good stuff
One of the most positive aspects was that participants gained a lot of expertise on finding data of each topical section, so that it was only natural that people would ask each other where the information was found. The capacities that were built will help us not only find data sources more efficiently, but also to assess their relevance and usefulness. In the long run, these efforts are strengthening civil society and will prepare to road to building better public policies regarding open data.
What can be improved
The ownership given to teams during the event can’t be replicated at home and we believe that would give it a whole new level of participation. That means, that one can claim to be a city revisor and acclaim that one will be responsible that the information is complete and curated. That would be really enabling and would allow for a more asynchronous dynamic.
In Mexico City, the event took place the 11th of december in a coworking space called Centraal and had 10 assistants.
The assistants were organized into groups to review how the cities were ranking up against each other in the census. A facilitator explained the parameters needed to upload a dataset.This event was organized jointly and with the support of SocialTIC, a non-profit organization that was in promoting the use of open data with journalists.
What can be improved
Getting the government involved. Mexico City pioneered the open data movement in Mexico. As such, they are well aware of the benefits and potential of opening data up. This can be seen in the Laboratorio para la Ciudad open data portal, which has continuously opened datasets and offered other digital services such as API endpoints. As such, the task of promoting an open data census shouldn’t have been that strange to government officials. However, participation from “the Lab” and other government agencies from Mexico City was close to none because there was a poor coordination and communication with them. We believe that we should’ve get them more involved in the process, perhaps inviting them to host the Open Data Census in their offices. We also believe that something that could be improved particularly in Mexico would be more closely coupled to the Open Knowledge Foundation, rather than making it feel something that Codeando México and SocialTIC were pushing. This could have given a little more relevance to the event.
Bringing in civil society. There was a small presence of organized civil society. It’s very important to bring in civil society to validate the relevance of the data that’s opened by the government. In a city like Mexico City, it’s not about opening up data (there’s lots of open data), but about it being really useful, accessible, and relevant to citizens.
Choosing a better time/date combination. In Mexico, during the month of December there are a lot of festivities (the traditional “posadas”). This usually makes it difficult to coordinate any large event, especially the ones that involve dedicating time to a civic cause. This also brings in major traffic problems in the city, in a city that usually has already major traffic problems, making it difficult for people to move to other parts of the city. In a city as large and busy as Mexico City, an event like this must be either be conducted during a non-labor day (like Saturday) or during an adjacent month (November or January).
In Xalapa, two activities took place on different days. The first activity was an “informal census”. This informal census included the parameters of the Open Data Census, but wasn’t filled-in.
The census was conducted by volunteers jointly with the Oficina para el Buen Gobierno (The Office for Good Government), but were never uploaded to the platform because some data was supposed to be confirmed by the government agency and wasn’t confirmed in time. The census data is annexed. One of the major challenges we encountered in this city is that we didn’t have local contacts to help us conduct the survey and with the right skills to conduct them. As such, the survey was pushed until March 5th.
On March 5th, the second activity for the Open Data Census was conducted during the Open Data Day activities. The activity consisted of a workshop on open data with government officials and members of civil society, which included a tutorial on what is the Open Data Census, why is it useful, and how to fill the survey. The workshop had an estimated participation of 70 assistants.
What can be improved
The Office for the Good Government is the team that coordinates the publishing of datasets to datos.gob.mx, the federal open data repository. There were a few things that could have been improved. For instance, the team in charge of publishing data is quite small and they were often unaware of all the data produced by the municipality that was requested by the census. Perhaps a better way to deal with this could be to get them involved from the beginning and not only invite them to the event. That way, they could be better prepared to provide resources to the participants of the census and would have a better idea on what to expect from it.
Working with the Office for the Good Government also taught us a few lessons:
1. There isn’t an open data culture inside the government. While they are aware of their relevance, they also feel that they are getting “exposed”. On a few occasions, they called someone, told them to send them an Excel or PDF file, uploaded it and felt that their task was done.
2. The quality of the data needs to be improved. Some datasets were opened in formats that weren’t truly open, such as Excel and PDF formats. Teams inside governments need to understand why data needs opened in accessible formats.
3. Opening data isn’t a single occurrence, it’s a on-going process. Teams inside the government believe that by only uploading datasets once, they’re already opening data. While this is partly true, the reality is that a lot of datasets weren’t relevant as they hadn’t been updated for months. A real commitment to opening data needs to exist inside governments.
Some of these lessons were tackled during the second activity, the workshop on Open Data. However, we believe that cultural changes inside the government do not happen overnight and they usually take time to settle in. During this year, the Codeando Xalapa brigade will be in charge of following-up on these matters and help them establish a permanent open data and open governance culture.