Pretend data to appear to support a pretend argument

From How Distrust of Government by Marginalized People Fuels Conflict in Africa by Hany Abdel Latif and Mahmoud El Gama.  The subheading is Improving people’s relationships with state institutions and ensuring access to services can foster conditions for greater peace and social cohesion in sub-Saharan Africa

The accompanying chart show violent violence by geographical location in Africa.

Click to enlarge.

The value is in the chart, not in the accompanying article. 

Sub-Saharan Africa suffers from persistent security challenges, and Sahel countries such as Burkina Faso, Chad, Niger, Mauritania, and Mali are particularly afflicted by protracted conflicts and humanitarian crises.

While various factors can fuel conflict, our research shows that discontent with state institutions among marginalized groups is a key driver of unrest in the region. Such distrust reflects perceptions that governments fail to address equity issues and inclusive growth—including the fair allocation of natural resources and human capital development.

Institutional failures aggravate feelings of exclusion where some people or groups are systematically denied rights, opportunities, or resources that may be available to other segments of the population. This invites conflict by undermining the principles of fairness and inclusivity vital for sustainable development.

As the Chart of the Week shows, conflict is often concentrated near national borders where there tend to be more limited or insufficient public services, fostering feelings of exclusion. These conflict-prone hotspots pose substantial security and stability risks in both directly affected and neighboring countries.

Basically a bunch of social justice nattering.  Their solution is to improve trust in government by addressing exclusion.  Pretty much a recipe for disaster.  

But does the map even support their argument?   Not obviously.  It is a fascinating map.  But most the violence is not at the borders.  Most of it is in areas of dense population.  Most of it is concentrated within Nigeria, Ethiopia, Kenya, Mali, Niger, Burkina Faso, Uganda, Rwanda and Burundi.  

In fact, the map has nothing to do with Trust in Government and virtually nothing to do with borders.  And nothing to do with their argument.  It is interesting data as non sequitur.  

That gets you to look more carefully at the details.  I had assumed that violence was referring to deaths.  Not so.  The map is recording instances of riots, battles, explosions, violence against civilians, and protests.

OK.  This went from interesting to suspect.  Definitions and data collection become critical.  What is the distinction between a protest and a riot?  Do we have confidence that protests and riots equally reported everywhere?  What about kidnappings?  

The map is showing activities, not outcomes.  Where are the deaths occurring?  And how many.  I would guess that they are probably reasonably correlated with the red areas on the map but it would be useful to have confirmed.  

I go to IISS The Armed Conflict Survey of 2022: Sub-Saharan Africa Regional Analysis.  The data is better than with the IMF but still not specific.  

I could keep digging but why?  If those making the argument can't be bothered to put some parameters to it, to measure it, then why should I do that work for them?  It seems to me that the latent implication is that the absolute numbers won't support the argument they want to make.  

Fair enough.  The world is full of unsupported posturings.  No need to invest time in all of them.