Rolf Pendall

BPJ 1994--Nonpoint source pollution: Getting to the nonpoint. Do site-by-site approaches to stormwater kill urban density? @kateontransport.bsky.social your post also makes me happy because you were such a great Building Resilient Regions co-conspirator. Thanks for this legacy. 2/2

This makes me happy on so many levels--my first 2 sole-authored articles were in BPJ. 1993: Clavel & Forester--shall ever the twain meet? Little did I know they were already mixing it up at Cornell, that I'd join them there 5 years later, and that you'd be my RA a few years after that. 1/

I’m kidding, mostly

Next stop Phoenix. It’s a slippery slope…water flows uphill to money.

2/2 In particular, LLMs tend to overgeneralize findings and implications in ways that the original authors tended to avoid, perhaps because they didn't have enough evidence to draw such conclusions.

Screen grab of article title and authors, "Generalization bias in large language model summarization of scientific research," Uwe Peters and Benjamin Chin-Yee

Abstract: "Artificial intelligence chatbots driven by large language models (LLMs) have the potential to increase public science literacy and support scientific research, as they can quickly summarize complex scientific information in accessible terms. However, when summarizing scientific texts, LLMs may omit details that limit the scope of research conclusions, leading to generalizations of results broader than warranted by the original study. We tested 10 prominent LLMs, including ChatGPT-4o, ChatGPT-4.5, DeepSeek, LLaMA 3.3 70B, and Claude 3.7 Sonnet, comparing 4900 LLM-generated summaries to their original scientific texts. Even when explicitly prompted for accuracy, most LLMs produced broader generalizations of scientific results than those in the original texts, with DeepSeek, ChatGPT-4o, and LLaMA 3.3 70B overgeneralizing in 26–73% of cases. In a direct comparison of LLM-generated and human-authored science summaries, LLM summaries were nearly five times more likely to contain broad generalizations (odds ratio = 4.85, 95% CI [3.06, 7.70], p < 0.001). Notably, newer models tended to perform worse in generalization accuracy than earlier ones. Our results indicate a strong bias in many widely used LLMs towards overgeneralizing scientific conclusions, posing a significant risk of large-scale misinterpretations of research findings. We highlight potential mitigation strategies, including lowering LLM temperature settings and benchmarking LLMs for generalization accuracy."

LLMs suck at lit review, and they're getting worse at it. Transparently obvious from today's news (HHS's nonsense MAHA report!) and backed with recently published peer-reviewed research by Peters & Chin-Yee: doi.org/10.1098/rsos...

A little more precision and nuance would be nice to see. 2/2

Can you please expand on what you mean by “skepticism of private development”? Also, you go from “urban planning” (a set of practices) to “planners”—which planners do you mean? Where? Which sectors? Professional planners work on all sides of development. 1/2

2/2 Small suburbs in TX have often incorporated to avoid high density housing. This bill could reinforce such fragmentation and exclusionary HOA formation.

Thanks for highlighting this bill. By my count, TX SB 15 will apply to about 16 cities total; it also applies only where HOAs don’t impose higher minimum lot sizes. In short, it could reinforce spatial inequality and/ or hasten gentrification.

This atrocious trend in police killings is entirely driven by red states, as the chart shows. Police killings plateaued in blue states and then fell.

5/5 How can we overcome the political unpopularity of broad-based taxes after 50+ years of concerted efforts to erode the sense that we're all in this together?

4/ At least since Prop 13, property taxes have been undermined by state & local backlash, but demand for local public goods has grown. People want stuff but they won't pay for it, but they don't want other people to get stuff they aren't paying for.

3/ But local fiscal systems are path-dependent and overdetermined by higher-level laws & constitutions (state or national, depending on the country). Has local property-tax dependence strengthened anywhere? How has that occurred? @cbgoodman.co -- examples?

2/ "Many cities around the world can and should take more advantage of the broad-based land value capture mechanism already at its disposal, the property tax. Local efforts to raise revenue should focus on overcoming its political unpopularity instead of devising new more complicated programs."

Challenges to Equitable and Effective Land Value Capture: Lessons from Mexico City — Urban Affairs Review Scholars and practitioners argue that local governments should capture rising land values to fund services and infrastructure, especially when the value results from public and collective action. Yet ...

1/ Property taxes are the best value-capture mechanism, per Aurora Echevarria & @elpaavo.bsky.social based on intensive study of Mexico City. Excellent summary of their arguments on @urbanaffairsreview.bsky.social blog: www.urbanaffairsreview.com/uar-archive/...

Ageism leads many people to conclude that older drivers are inherently unsafe. This misconception needs to be corrected, and framing needs to shift. S. Rosenbloom has written extensively about this; also see a recent JPL review by Alex Li journals.sagepub.com/doi/abs/10.1...

Report: Young driver fatality rates have fallen sharply in the US, helped by education, technology A new report says that crash and fatality rates among drivers under 21 have fallen dramatically in the U.S. during the past 20 years but young drivers are still the riskiest group behind the wheel.

You might want to update this; your data are from 2013-14 I think. Crashes among young drivers are way too high, but they've come down significantly thanks in part to evidence-based policymaking. apnews.com/article/youn...

The Deadly Story of the U.S. Civil Service : Throughline When James Garfield won the Presidency in 1880, Charles Guiteau got ready to accept his new government job. No one had actually offered him a job – but he'd campaigned for Garfield, so he assumed he'd...

Indirectly related to the article: Spoils system vs. civil service figured in the assassination of James Garfield. Check it out on Throughline (a consistently great NPR podcast): www.npr.org/2025/04/24/1...

Amazing and sobering work. Congratulations.

Women leading the settlement house movement advocated for public toilets over 100 years ago. See Maureen Flanagan’s “Constructing the Patriarchal City” for examples of feminist urbanism of the early 1900s in London, Dublin, Toronto, & Chicago.

I promised my friend @rpplan.bsky.social that I would activate my account here to share my thoughts about the larger economic, political and global aspirations behind the tariffs, written for @chathamhouse.org

www.chathamhouse.org/2025/04/trum...

This is so wrong. I admire your restrained tone. Thank you for starting the work, and best wishes finding alternative funding sources.

Interesting reading. What implications for housing policy? Households are aging and shrinking but it’s expensive to adapt the stock. Would it be cost-effective to socialize costs of retrofits for energy efficiency, health, & safety?

Are Federal Infrastructure Dollars Meeting Your Community’s Needs? Data analysis tracking whether the billions in funding that the US has committed to its roads, bridges, housing, and other infrastructure are getting to the communities that need it most.

New data tool launch today 🚨 @urbaninstitute.bsky.social's Federal Infrastructure site.

It includes detailed information on funds from 110 programs across all states, counties & metro areas.

Tool pinpoints the support that communities receive for transportation, housing, energy, water & broadband.

Please, please, if you report on this: the mandate to affirmatively further fair housing is literally codified in the Fair Housing Act of 1968 and neither Trump nor HUD can “repeal” this law. The rule defined how to enforce the law, but the law and obligation remain. www.yahoo.com/news/hud-sec...

OK, I guess we'll have to agree to disagree.

Hypothetical: would you characterize a 15% IZ mandate (without additional public subsidy) adopted simultaneously with a 30% upzoning and a new transit station as "unfunded"?

Even if every tool could be a weapon, that doesn't mean we should ban every tool because of potential abuses, right? Do you support banning IZ or regulating it? Either way, the tool users are motivated to exclude by all means necessary and will find some other way to block housing.

This makes sense, but there's an unfortunate tendency to extrapolate from California to everywhere else. Also, folx at one national think tank are broadcasting anti-IZ perspectives as part of a narrative that all land-use regulation is harmful.