iiinova — where your digital self keeps evolving

Will Marble @wpmarble.bsky.social · 12/03/2026

nice point. I also think people just got used to the very low (sometimes near 0) inflation we had for a decade post-recession, making the 2022 inflation seem even more anomalous. Groceries (like gas prices) are a salient heuristic

030

Will Marble @wpmarble.bsky.social · 08/01/2026

The package supports both full Bayesian calibration (i.e. calibrating each draw from the posterior) as well as a plug-in estimator which works with posterior summaries.

Picture of R package documentation for argument called "method":

Calibration method, either "plugin" or "bayes". Plug-in estimates use posterior means of predictions and correlations across outcomes to compute logit shifts for calibration. Bayesian estimates compute the logit shifts separately for each posterior draw, which are then summarized. Defaults to "plugin".

110

Will Marble @wpmarble.bsky.social · 08/01/2026

calibratedMRP provides a high-level interface for doing this type of calibration. First, specify a `brms` multilevel regression model. Second, call the `calibrate_mrp()` function to calibrate cell-level estimates to known geographic margins. Third, poststratify to the target groups.

Example code that implements a brms multilevel model with multivariate outcomes.

example of syntax for calibrate_mrp() function

110

Will Marble @wpmarble.bsky.social · 17/07/2025

There are also differences in contemporary public opinion: people living near colleges are more liberal on a range of issues. These attitudinal differences are not solely driven by the presence of students nor by differences in the average educational attainment in the community.

110

Will Marble @wpmarble.bsky.social · 17/07/2025

From Reconstruction through the end of the 20th century, there were minimal differences in voting patterns between places with a university and "runner-up" locations. Since 2000, though, a gap has emerged, with college counties becoming significantly more Democratic. In 2024 this gap was ~10pp.

110

Will Marble @wpmarble.bsky.social · 17/07/2025

The establishment of a college also leads to a county casting significantly more votes in presidential elections—an effect that's explained by a population growth channel rather than a turnout rate channel.

110

Will Marble @wpmarble.bsky.social · 17/07/2025

We find that places with colleges have significantly higher levels of social capital and trust today, relative to "runner-up" locations that were considered but not chosen.

140

Will Marble @wpmarble.bsky.social · 17/07/2025

How do universities shape the surrounding community? Building on meticulous archival work by my co-author Mike Andrews, we answer this question by focusing on cases where multiple locations were considered for a major university and the winning location was chosen for idiosyncratic reasons.

To further illustrate this methodology, we describe several of the as-good-as-random site selection experiments here. In some cases, which of the finalist locations won the college was literally random, with the winning location drawn by lots (this occurred for the University of North Dakota and North Dakota State University). In other cases, such as University of Illinois, Purdue University, and the University of Florida, multiple towns submitted similar bids to receive the college, with only one ultimately chosen. In still other cases, it took multiple rounds of balloting to find a winner (7 rounds of balloting for University of Mississippi, 8 rounds for Southern Arkansas University, 24 rounds for the Georgia Institute of Technology, and a whopping 111 for what would become University of Nebraska at Kearney). The fact that finalist locations were tied for multiple rounds of voting suggests that both the winning and runner-up locations were similar in terms of political influence and enjoyed comparable popular support. Site selection experiments occur as early as 1839 and as late as 1954, though the majority of experiments are concentrated in the 1880s and 1890s.

120

Will Marble @wpmarble.bsky.social · 17/07/2025

Universities often serve as "anchor institutions" that deeply affect the character of their communities. In a new paper, we estimate how (and when) the establishment of a college influences local political and civic life. 🧵 osf.io/preprints/so...

Screenshot of title page.

"The Long-Run Effects of Colleges on Civic and Political Life" by Michael J. Andrews, William Marble, and Lauren Russell.

Abstract: Social theorists and education advocates have long argued for the civic benefits of education. As large, durable institutions, universities are especially likely to affect the civic life of their communities. We investigate how the establishment of a university alters the civic and political trajectory of the surrounding area. For identification, we leverage historical site selection processes in which multiple locations were considered for new colleges. We bring together data on social capital, political preferences, and elections to assess the long-run impacts of college establishment. Communities with colleges exhibit higher levels of civic engagement and greater social trust today, relative to “runner-up” locations without colleges. These counties are also more politically liberal — a gap that has grown substantially since 2000. Our findings suggest understanding universities as place-based policies that shape the long-run civic and political development of their communities. They also shed light on current political battles over higher education policy.

13010

Will Marble @wpmarble.bsky.social · 10/06/2025

The claim that Trump is just "enforcing immigration law" is ridiculous given that the administration has repeatedly *violated immigration law* by detaining/deporting people without due process.

Screenshot from WSJ op-ed by Tom Cotton: 'Meanwhile, incredibly, Democrats and the liberal media have again called this outbreak of violence “mostly peaceful protests,” while in the same breath blaming the riots, arson, and looting on President Trump for enforcing immigration law. The New York Times described “largely peaceful” riots with fireworks fired at police, cars set on fire, and more than 150 arrests.'

18420

Will Marble @wpmarble.bsky.social · 05/10/2023

Work on conjoint survey experiments in political science is relevant here--the classic cite is Hainmueller et al 2014, who are clear about the estimand in the "short" model averaging over the interactions. The notation is a bit gnarly in that paper though... www.cambridge.org/core/journal...

020

Will Marble @wpmarble.bsky.social · 25/09/2023

Why does economic decline seem to benefit right-wing parties in the U.S. and Europe? Junghyun Lim and I propose an explanation rooted in the interplay between social arrangements and economic geography. osf.io/preprints/so...

Title page for a paper called "Migration, Social Ties, and Political Preferences" by William Marble and Junghyun Lim. Abstract: Why have economically declining regions turned toward right-wing parties? To explain this puzzling phenomenon, we develop a theory linking internal migration, localized social institutions (e.g., family and friend networks), and voters’ preferences for social insurance. We start with the observation that social ties provide insurance against adverse life events, such as job loss, and highlight two implications. First, those with strong social networks prefer lower spending on social insurance, because they have access to informal insurance that acts as a substitute for public programs. Second, social ties discourage people from moving, even when better economic opportunities are available in other regions. Combining these mechanisms, we argue that the effect of economic shocks on a region’s politics depends on the strength of social ties.[continued]

0183

Will Marble @wpmarble.bsky.social · 18/09/2023

The framework I use here enables a principled interpretation of the correlation between vote choice and attitudes. There are issues of causality that I discuss in the paper. It's not perfect, but IMO this framework provides important insights into the nature of voting decisions.

100

Will Marble @wpmarble.bsky.social · 18/09/2023

3) The relative weight that whites w/o college degrees placed on cultural issues was low until the mid-2000s—limiting the returns Republicans could reap from the WWC's conservative cultural attitudes. Now, the weight placed on cultural issues is identical for the two groups.

100

Will Marble @wpmarble.bsky.social · 18/09/2023

The upshot is that there is now consistent polarization across issue domains between whites with and without a college degree.

100

Will Marble @wpmarble.bsky.social · 18/09/2023

2) The white working class has long been more conservative than college-educated voters on cultural issues. But in the past ~20 years, college-educated voters have become significantly more liberal on economic issues as well.

100

Will Marble @wpmarble.bsky.social · 18/09/2023

I present sever findings: 1) Ideological constraint has increased dramatically among both college-educated and non-college-educated voters, expanding the possibilities for group-based polarization.

100

Will Marble @wpmarble.bsky.social · 18/09/2023

I use survey data to generate issue-specific ideal point estimates for Economics, Moral/Social Issues, Race, and Foreign Policy. I introduce a framework that enables me to decompose differences in vote choice into different preference distributions and different issue weights.

100

Will Marble @wpmarble.bsky.social · 18/09/2023

I study this transformation over ~40 years, asking which issues have contributed most to this realignment.

100

Will Marble @wpmarble.bsky.social · 18/09/2023

First, the well-known realignment: In 1992, GHWB won 45% of white voters w/o a college degree and 52% of those with a college degree. In 2020, Trump won 65% of non-college whites and just 42% of whites with a degree. The college divide is widening.

100

Will Marble @wpmarble.bsky.social · 18/09/2023

The white working class has steadily become part of the Republican base while whites w/ college degrees have become Democratic. In a new paper, I study the issue basis of this realignment, showing that both economic and cultural issues have contributed. williammarble.co/docs/EducPol...

Title page for the paper "What Explains Educational Polarization Among White Voters? by William Marble (September 2023). Abstract: Over the past 40 years of American politics, college-educated white voters have defected from the Republican Party, while the white working class has become a reliable source of Republican support. I study the issue basis of this realignment. To do so, I generate over-time estimates of public opinion on four broad issue domains from 1984 to 2020 and develop a theoretical framework to understand how issue attitudes translate into electoral coalitions. Using this framework, I find that both economic and cultural issues have contributed to the observed realignment. College-educated white voters have become increasingly liberal on economic issues since the mid-2000s; college-educated voters now express more liberal views than working class voters on every issue domain. Over the same time period, cultural issues have become more important for the
voting decision

153

Will Marble @wpmarble.bsky.social · 19/08/2023

No help on the other site, trying here: What is the easiest way to step through/debug a function using tidyeval in R? With base R I can just run this code line-by-line interactively after assigning the args in the global environment. But AFAIK the {{ }} tidyeval syntax only works inside a function.

000

Will Marble @wpmarble.bsky.social · 12/08/2023

Small model detail: w/ brms term (1 | id | group), the random intercept for each individual j is b_j ~ MVN(0, Sigma). The correlation matrix for Sigma has LKJ(1) prior by default. (This comes from looking at generated stan code, so could be wrong. I couldn't find clear documentation.)

110

Will Marble @wpmarble.bsky.social · 09/08/2023

Glad the new proposal raises the arena floor to accommodate Jefferson Station, but it's hard to tell whether it's enough to save the natural lighting that Inga Saffron has highlighted in the Inquirer. AFAIK still no word on whether the new proposal would require MSG-style pillars thru the station.

000