Matthew K. Ribar

Banner image featuring the hashtag #OpenAccess on a green background above the text 'American Political Science Review' on a blue background.

#OpenAccess from @apsrjournal.bsky.social -

Land, Power, and Property Rights: The Political Economy of Land Titling in Sub-Saharan Africa - https://cup.org/46qHNnT

- @mattkribar.bsky.social

#FirstView

Pierre, a cat, lays upside down.

Big thanks to my supportive advisors, the various groups who funded this research, my wife, and one particularly fussy editor:

The field survey in Cote d’Ivoire process-traces three intermediate steps: (1) strong chiefs lead to more titles because Ivorian chiefs have an incentive to facilitate rather than impede; (2) chiefs capture land management and exclude the local out-group; and (3) chiefs benefit from this capture.

But households aren’t the only actors here; traditional chiefs either facilitate titling (if the state devolved land governance, meaning that chiefs can capture these institutions) or impede it (if titling remains centralized).

First, if farmland isn’t valuable enough (or if there are limited returns to the kind of agricultural investment that titling facilitates, like fertilizing land or planting trees), then households won’t bother to title.

Back to the question, I marshal 170,216 household-level observations of titling across 22 African countries from both the DHS and LSMS programs, along with a geospatial strategy to measure land values and a field survey in Cote d’Ivoire.

Control over land tenure can also act as a reservoir of authority for traditional elites—meaning that property rights are a crucial arena in which to test when traditional elites complement or substitute for the state.

First: why should you care about land titling? Some recent Nobel prizes suggest that good institutions are a necessary condition for economic development. Where the agricultural sector still employs many households, secure land tenure is the ‘institution’ that counts.

This figure shows the fraction of land-owning households who possess a formal land title for at least one parcel across 22 African countries. Data are from the Demographic and Health Surveys as well as the Living Standards Measurement Surveys.

Abundant research says formalizing land ownership is good for households, and land titles are available on-demand through much of Africa. But the below figure shows both low take-up overall and high amounts of variation within states. What gives?

I’m thrilled that my job market paper, “Land, Power, and Property Rights: The Political Economy of Land Titling in sub-Saharan Africa,” is hot off the press and open-access at @apsrjournal.bsky.social. A quick thread featuring more than you ever wanted to know about land tenure.

We present our new preprint titled "Large Language Model Hacking: Quantifying the Hidden Risks of Using LLMs for Text Annotation". We quantify LLM hacking risk through systematic replication of 37 diverse computational social science annotation tasks. For these tasks, we use a combined set of 2,361 realistic hypotheses that researchers might test using these annotations. Then, we collect 13 million LLM annotations across plausible LLM configurations. These annotations feed into 1.4 million regressions testing the hypotheses. For a hypothesis with no true effect (ground truth $p > 0.05$), different LLM configurations yield conflicting conclusions. Checkmarks indicate correct statistical conclusions matching ground truth; crosses indicate LLM hacking -- incorrect conclusions due to annotation errors. Across all experiments, LLM hacking occurs in 31-50\% of cases even with highly capable models. Since minor configuration changes can flip scientific conclusions, from correct to incorrect, LLM hacking can be exploited to present anything as statistically significant.

🚨 New paper alert 🚨 Using LLMs as data annotators, you can produce any scientific result you want. We call this **LLM Hacking**.

Paper: arxiv.org/pdf/2509.08825

This looks really interesting!

Rubio said no children died? Add that to the pile of lies. At least Musk admitted “he gave zero fucks.”

And all this? For a bunch of AI-generated hallucinations and outright lies about “corruption” at USAID.

By the way, in the initial wave of AI-powered program cuts, Lewin, Rubio, and Musk pinballed between cutting production contracts, then restoring production while cutting shipping, then cutting shipping while restoring production, etc.

Starting production takes time, arranging shipping takes time, the shipping itself takes time. That was true before they fired everybody and its even more true now.

This means food goes undelivered. Delays pile up. Much of the world’s supply of ready to use therapeutic food (RUTF) is made in either Lubbock, TX or Rhode Island, far from where it is used.

But Rubio, Musk, and Jeremy Lewin fired single person at USAID who knew how this process works. There are a handful of contracts officers (high-level) and program assistants (entry level) left, but none of the core staff who do the work (i.e. save lives).

A $45 Treatment Can Save a Starving Child. US Aid Cuts Have Frozen the Supply

This article highlights the fundamental lie behind Rubio’s claim that humanitarian aid is still flowing. Have grants been signed? A couple (massive cuts to humanitarian food). Has money been obligated? It’s starting to trickle. www.nytimes.com/2025/08/15/h...

Look, I hate to get all 1938 here, but this is exactly the same concession that Hitler made over Czechoslovakia.

Do Moderates Do Better? Uncovering Bias in Split Ticket’s WAR Scores

We took a close look at Split Ticket's WAR metric, which has become influential in Democratic circles for suggesting moderates significantly outperform progressives.

Our finding: The metric contains systematic biases that overstate the advantage of moderation. A corrected model shows no advantage.🧵

E.J. Antoni has more insurrections to his name than published papers

i mean, a genuinely striking thing about trump is that he basically never makes an appearance in the actual city outside

The Demographic and Health Survey’s death, and potential revival, explained Why this obscure survey from USAID has life-or-death stakes for global health. The Gates Foundation announced $25 million in emergency funding, but the long-term solution is far from clear.

Excellent article by Pratik Pawar on the consequences of cutting the Demographic and Health Surveys, and efforts to revive them. I'm quoted briefly.

Trump and Republicans know that America’s presence and influence around the world is critical.

Yet they’re destroying USAID and other foreign programs, ceding our global leadership to China.

The worst part about this "interview" is that all of Lewin's opinions are very obviously post hoc justifications. He went in with presumably ketamine fueled rage, destroyed everything, and made up a policy justification after.

At one point, several WASH advisors within BHA were debating if using the word "sewage" or "poop" was more likely to get their work past the DOGE bros.

Meanwhile Farritor and Lewin are using a shitty AI to determine what aid continued, and then canceling it all anyways. This led to a lot of dumb decisions at BHA, like transport contracts for food assistance being canceled, but not the food assistance itself, and vice versa.

Pete Marocco was fired from Trump 1 for being too crazy and trying to unilaterally dismantle his corner of USAID. My best guess is that Marocco was covering his ears and saying "lalala" the whole time. I'd love to see him deposed.

This is what I was hearing from inside USAID at the time---it was Farritor and Lewin unilaterally making these decisions. The problem is that the director of state foreign assistance, Pete Marocco, was probably giving these guys cover.

How the Threat of Trump’s Highest Tariff Derailed an African Nation

When the Trump administration said "trade not aid," it turns out they don't mean trade either:

www.nytimes.com/2025/08/01/w...