Feowl

Feowl wants to measure power cuts. What exactly is a power cut and how can you measure it? We asked the question a few days ago. It was time to ask the people of Douala what they thought about the issue.

@Pirhoo and @nicolaskb, with the invaluable help of our driver Nicaisse, spent a full day walking and driving across Douala, asking over 50 people in 34 shops which problems they had with electricity. A sort of mini-baseline survey.

The 4 neighborhoods where the survey was conducted

Biases in data collection are too numerous to count:

• Suspicion. Quite a lot of the people we met thought we were from the power utility, AES-Sonel, and that our pretending to be journalists was just a clever trick. Showing a press card helped a bit.
• Shyness. Some respondent (mostly women) agreed to answer but mumbled only a few words. 
• Selection. We did not randomly select the people we talked to, as we mostly walked along the street we parked in (going inside the slums, were roads aren’t tarred, would have required a 4WD, or more time). We also talked only to businesses. Finding households would have been more time consuming.

Despite these shortcomings, we now have a clearer idea of the problem. Here’s what we’ve learned.

People do remember power cuts (well, sort of)

Power cuts are such a common occurrence that we thought most people would not be able to remember exactly how many power cuts they had the day before. But when we repeated the question, making it clear that we were interested in the precise time of the outage, 9 people out of 10 could remember.

Now, for most people, a 10-minute power cut isn’t remarkable. They often fail to report them, so we had to make it clear that we were interested in every power cut, even small ones.

Power cuts are definitely independent events

Last time, we wondered if power outages were linked or if we could consider that each of them was a distinct event. It turns out that power can be cut from so many different sources that city- or neighborhood-wide power cuts are not the real thing. Most people get their electricity from neighbors, so that the secondary line can fail even if power is still up elsewhere, for instance.

One respondent mentioned that the power line to his shop were hanging so low that it was not uncommon for trucks to tear it away. Another pointed out that the rain could cause power to fail due to the poor state of the poles.

Data about power cuts is missing

It sounds obvious, but power cuts seem to be very hard for people to estimate (just like inflation in richer countries, which is notoriously hard to assess by oneself). We asked whether the situation was better now compared to 3 years ago. Several professional in the energy sector told us that things improved in the past 2 years, as the 86 MW Dibamba plant opened (the US embassy in Yaoundé says the same thing in the cables opened by Wikileaks). Despite this potential improvement, respondents gave diverging answers when asked if things had improved.

More interestingly, we asked respondents which season was worse, expecting that the vast majority of them would say that things were better during the rainy season, when the reservoirs of the dams are full. Although a third of respondents gave this answer, it isn’t as clear cut as we expected.

Another possibility is obviously that power cuts are also independent in the long run, as individual causes outweight city-wide blackouts. In this case, we will need to individually assess any progress, to be able to say that “for X inhabitants of Douala out of 10, the situation improved over the last Y months”. However, what we heard pointed in the direction of neighborhood-wide trends.

Voltage is the main problem

Besides power cuts, respondents consistently pointed out to variations in voltage as one of the most important issues. Owners of electronic appliances (computers, fridges etc.) all complained about destroyed assets due to voltage below or above 220V.

Results

The data (get the raw data on Google Drive) shows a large dispersion of answers (no 2 respondents gave us similar answers) but coherent data on a per neighborhood basis.

Feel free to browse through the data and share the insights you find!