iiinova — where your digital self keeps evolving

eli knaap @knaaptime.com · 16/07/2026

scale matters a lot too. Perris CA has like 80k people, but its quite rural in many respects--even if people are commuting to the mecca that is riverside

040

eli knaap @knaaptime.com · 16/07/2026

thats what fascinates me. My midwestern memory thinks of places like Mahomet as being pretty "genuinely" rural even if they fall inside the *highly urbanized* champaign urbana metro :P

240

eli knaap @knaaptime.com · 16/07/2026

my frame of reference is, the reason SD is not part of SCAG is not enough people commute through Pendleton ...like i do 😬

010

eli knaap @knaaptime.com · 16/07/2026

like a top line of the FHFA report "Vermont is the most rural state, with 80.9 percent of the population living in rural areas." is what i mean. those are 'genuinely rural people' for whatever that means

120

eli knaap @knaaptime.com · 16/07/2026

i think there's also two pieces to this: who considers themseves rural (inclusive of exurbans), and who the academics consider rural based on criteria set (which i *think* tends to exclude the exurbans)

211

eli knaap @knaaptime.com · 16/07/2026

which is the foundation for the MSA definitions to begin with, yeah? could definitely take a good look with the commuting data. Worth doing if i ever get a free minute

100

eli knaap @knaaptime.com · 16/07/2026

im totally with 'the overwhelming majority of people are urban/suburban'. I'm curious about *of the rural share*, who's genuinely rural not an exurbanite consuming genuine metro resources

100

eli knaap @knaaptime.com · 16/07/2026

@danimmergluck.bsky.social do you have a take on this? rurality isnt a specialty of mine but i think housing scholars probably have a more nuanced view

000

eli knaap @knaaptime.com · 16/07/2026

its an interesting empirical question and i'd actually have a weak prior its wrong. Its true most people dont understand urban systems are metropolitan in scale. But the ~25% pop FHFA defines as rural are not really drive-till-you-qualify people, i dont think www.fhfa.gov/blog/insight...

fhfa.gov

Who Lives in Rural America? | FHFA

In this blog post, we share insights into the demographics and housing characteristics of rural America. Through our analysis of the Census Bureau’s American Community Survey (ACS), we find that, on a...

201

eli knaap @knaaptime.com · 02/07/2026

that whole thing was dramatically funnier than i anticipated

000

eli knaap @knaaptime.com · 02/07/2026

"its not too late to braise it" 😭😭😭

110

Reposted by eli knaap

Mark Copelovitch @mcopelov.bsky.social · 28/05/2026

This is the destruction of federal funding for scientific research & universities as we know it. A total, comprehensive politicization of the process & an end run around Congress.

13470219

Reposted by eli knaap

Paul Goldsmith-Pinkham @paulgp.com · 17/04/2026

Finally updated the search engine behind econlit-db to Meilisearch. Everything should be quite fast now -- under a second or two. Let me know if you have any bugs. paulgp.com/econlit-pipe...

paulgp.com

Economics Literature Search

Full-text search across 15,000+ papers from top economics journals and NBER working papers. Track how empirical methods have evolved over time.

4319

eli knaap @knaaptime.com · 15/04/2026

we're working on a huge documentation/example gallery, etc push that we're hoping will make the PySAL stack more discoverable and easy to work with for folks like you we might be able to suck into our orbit :P

020

eli knaap @knaaptime.com · 15/04/2026

i am a fan of the spatial panel (love a good elhorst)! I sent [this paper](papers.ssrn.com/sol3/papers....) to both JUE and RSUE and both editors were like 'interesting; needs reframe... everyone knows spatial effects matter...' i was like ...but do they??

papers.ssrn.com

Measuring the Purple Line's Effect on Residential Rents: Spatial Econometrics and Causal Inference in Urban Policy Analysis

In this paper, I leverage a unique dataset containing unit-level observations on rental prices for units in Montgomery County, Maryland to build a series of hed

110

eli knaap @knaaptime.com · 15/04/2026

another quirk in the econometrica paper is all spatial autocorrelation is viewed as misspecification and i cant understand how their approach works any time you have legitimate spillover in X or Y? arent they proposing griffith-style spatial filtering? onlinelibrary.wiley.com/doi/abs/10.1...

More abstractly, as discussed in Conley (1999), “locations” might index
an economic characteristic and the “economic distance” between locations measures the
dissimilarity of the characteristic. We thus follow the geostatistical tradition of positing a
continuous parameter model of spatial variation, rather than, say, model dependence by
spatial autoregressive (SAR) models of Cliff and Ord (1974) and Anselin (1988).2

Spatial Dependence In many instances where regression unalysis is applied 10 cross-sectional data, various measurement problems, spatial externalities, and other spillover effects may lead to dependent error terms. In addition, in the emptrical implementation of models of spatial processes, spatial structure, or spatial interaction, the dependent variable at one point in space may be functionally related 10 i1s values at some or all other locations in the system. In other words, spatial dependence is reflected in regression analysis in two distinct forms, one as a misspecification and the other as a result of the explicit modeling of space. In more formal terms, these two basic representations of spatial dependence can be expressed as: y= f(y, X, beta, epsilon), Or epsilon= g(epsilon, lambda, xi)

010

eli knaap @knaaptime.com · 15/04/2026

*stares blankly in Regional Science* here's Luc and Dan, a mere forty years ago, discussing the inferential issues in spatial regression, which were already very well-known at the time

Do Spatial effects really matter in regression analysis? By Luc Anselin and Daniel Griffith, Papers in Regional Science 1988

The effects of spatial dependence in the error term are the same as for any nonspherical disturbance variance matrix. The regression coefficients will remain unbiased using OLS, but significance tests and measures of fit may be misleading. This result is due to the biased estimation of error variance, / test significance levels, and R2?.

C O N C L U D I N G R E M A R K S O n e o f t h e m a i n o b j e c t i v e s o f t h i s p a p e r w a s t o i l l u s t r a t e t h a l i g n o r i n g s p a t i a l c f l e c t s i n a p p l i e d r e g r e s s i o n a n a l y s i s m a y l e a d t o s e r i o u s e r r o r s i n t h e I n t e r p r e t a t i o n o f f a m i l i a r r e g r e s s i o n d i a g n o s t i c s a n d m i s s p e c i f i c a t i o n t e s t s . ‘ C h e r e s u l t s o f t h e s i m p l e n u m c r i c a l e x p e r i m e n t s r e p o r t e d o n i n t h i s p a p e r c l e a r l y d e m o n s t r a t e t h a t s p a t i a l c t l e c t s r e a l l y d o m u t t e r i n r e g r e s s i o n a n a l y s i s . A t t h t s p o i n t i n t i m e , a c o n s i d e r a b l e b o d y o f l i t e r a t u r e i s a v a i l a b l e t h a t d e s c r i b e s t e c h n i q u e s f o r h a n d l i n g s p a t i a l e f f e c t s i n a p p l i e d r e g r e s s i o n a n a l y s i s . H o w e v e r , t h e r e a p p c u r t o b e s e v e r a l m a j o r o b s t a c l e s h i n d e r i n g a m o r e e f f e c t i v e d i s s e m i n a t i o n o f t h e s e m e t h o d s t o t h e p r a c t i c e o f r e g i o n a l s c i e n c e a n d g e o g r a p h y , W e f e c l t h a t i n o r d e r t o a c h i e v e a n e f f e c t i v e a n d t h o r o u g h d i s s e m i n a t i o n a n u m b e r o f i m p o r t a n t r e s e a r c h i s s u e s n e e d t o b e a d d r e s s e d . A m o n g t h e s e w e b e l i e v e t h e f o l l o w i n g f i v e a r e m o s t c r u c i a l : ( 1 ) a m o r e e f f e c t i v e t e s t i n g f o r t h e P r e s e n c e o f s p a t i a l e l l e c t s i n ’ r e g r e s s i o n r e s i d u a l s , e i t h e r i n i s o l a t i o n o r i n c o m b i n a t i o n w i t h o t h e r s o u r c e s o f m i s s p e c i f i c …

230

eli knaap @knaaptime.com · 15/04/2026

that is precisely how i read it

020

eli knaap @knaaptime.com · 15/04/2026

on the one hand, it's dope to see spatial analysis win Econometrica's best paper (!) on the other, it's so deeply troubling that this statement is news to economists in 2026

excerpt from the paper's conclusion:

Applied researchers are well aware of the pitfalls of conducting inference with persistent time series data. Variables are routinely tested for extreme time series persistence
using unit root or related tests, and are often differenced to avoid spurious regression
effects.
This paper demonstrates that inference with highly persistent spatial data is equally
fraught: HAC or clustered standard errors fail in the presence of strong spatial correlation, leading to spurious significance between independent spatial variables. We have
provided tools to detect such strong spatial persistence, akin to time series unit root and
stationarity tests.
We have also suggested ways of restoring valid regression inference by suitably transforming the spatial variables, combined with spatial HAR corrections to accommodate
any residual weak correlations. The theory here is much less complete, however.

151

Reposted by eli knaap

Adam Serwer @adamserwer.bsky.social · 09/04/2026

Wrote about how Hegseth’s effort to resegregate the upper ranks of the US military ties into the larger MAGA project to restore hierarchies of race and gender: www.theatlantic.com/ideas/2026/0...

472300831

Reposted by eli knaap

Don Moynihan @donmoyn.bsky.social · 09/04/2026

"To put it bluntly, an X post today receives less than 3% of the views a single tweet delivered seven years ago." If you think it is necessary to be on Twitter to communicate your fact-based worldview, the reality is that worldview is being smothered.

104107993240

Reposted by eli knaap

Jerry Chen @jcsalterego.bsky.social · 06/03/2025

starting to think podcasts are the problem

1553847238

eli knaap @knaaptime.com · 08/04/2026

at an analyst level i totally agree, but at a *software* level im curious what examples you have in mins that still don't do spatial indexing?

100

Reposted by eli knaap

Dan Garisto @dangaristo.bsky.social · 03/04/2026

BREAKING: In response to huge cuts in Trump's budget request, NSF is shuttering its SBE directorate. Staff will be transferred to other parts of the agency and "grants that align with Administration priorities" will be maintained. That & more w/ @maxkozlov.bsky.social & @edwrdchen.bsky.social

nature.com

Massive budget cuts for US science proposed again by Trump administration

Budget proposal would also curb federal payments for scientific publishing.

26407315

Reposted by eli knaap

Eileen Clancy 🧿 @clancyny.bsky.social · 02/02/2026

Notepad++ "Chinese-linked cyberespionage group with a long history hijacked the update process for the popular code editing platform Notepad++ to deliver a custom backdoor and other malware."

04228

eli knaap @knaaptime.com · 29/01/2026

(That’s real data from the paper :P)

000

eli knaap @knaaptime.com · 29/01/2026

if foot traffic flows into a place like downtown LA looked something like this before the ICE raids, then, say, like this afterward... well, there will be economic consequences, on top of everything else socialecology.uci.edu/news/oc-busi...

a racial dot density map of the LA regions. Colored lines of varying widths show the level of flow from different communities (before the June ICE raids)

a racial dot density map of the LA regions. Colored lines of varying widths show the level of flow from different communities (after the June ICE raids). There are substantially fewer lines and smaller widths indicating fewer people traveling downtown from a smaller set of origins. Many of the reduced flows come from predominantly nonwhite communities, suggesting a chilling effect across the diverse region

110

Reposted by eli knaap

eli knaap @knaaptime.com · 20/06/2025

DID NO ONE SEE WATCHMEN???? bsky.app/profile/scot... it. um. won 11 emmys and explores the inherent racism underlying masked law enforcement

141

Reposted by eli knaap

Jeff Rueter @jeffrueter.bsky.social · 25/01/2026

Kare 11 local news just read, in full, this statement from Michael and Susan Pretti, the parents of Alex Pretti. "Please get the truth out about our son."

Statement from Michael and Susan Pretti
Parents of Alex Jeffrey Pretti

“We are heartbroken but also very angry.

Alex was a kindhearted soul who cared deeply for his family and friends and also the American veterans whom he cared for as an ICU nurse at the Minneapolis VA hospital. Alex wanted to make a difference in this world. Unfortunately, he will not be with us to see his impact.

I do not throw around the ‘hero’ term lightly. However, his last thought and act was to protect a woman. The sickening lies told about our son by the administration are reprehensible and disgusting. Alex is clearly not holding a gun when attacked by Trump’s murdering and cowardly ICE thugs. He had his phone in his right hand and his empty left hand is raised above his head while trying to protect the woman ICE just pushed down, all while being pepper sprayed.

Please get the truth out about our son. He was a good man. Thank you.

1962470512668

Reposted by eli knaap

Jens von Bergmann @jensvb.bsky.social · 23/01/2026

Global datasets are really cool. Lots of fun to be had, I used this for looking at city evolution over time, sprawl, building heights, … Just really cool when one can do that globally with the click of a button. E.g.:

doodles.mountainmath.ca

Density timelines – Mountain Doodles

Global city density patterns across time.

111

eli knaap @knaaptime.com · 23/01/2026

that looks.. awesome? I'd cooked up an example with copernicus data once github.com/pysal/tobler... but im unfamiliar with GHSL. Time to read up. It's a bit like NLCD where they go from remote-sensed imagery into a model of population density?

github.com

tobler/notebooks/extract_urban_areas.ipynb at main · pysal/tobler

Areal & Dasymetric Interpolation; Change of Support - pysal/tobler

110

eli knaap @knaaptime.com · 23/01/2026

fwiw, serge has a new paper in the works based on that bipartite idea that's just about the econometrics of two interacting spatial systems, not necessarily schools and neighborhoods (its a tribute to harry kelejian), but its baaaad ass

020

eli knaap @knaaptime.com · 23/01/2026

not sure which one you meant, so heres both www.dropbox.com/scl/fi/jg7lz... www.dropbox.com/scl/fi/24lvp...

120

eli knaap @knaaptime.com · 23/01/2026

we're on a roll here... school attendance boundaries are another boondoggle i find extraordinarily interesting :) link.springer.com/epdf/10.1007... doi.org/10.1007/s110...

link.springer.com

Measuring Spatial Congruence in the School-Neighborhood Nexus

130

eli knaap @knaaptime.com · 23/01/2026

hm interesting. If you actually want to *test* for the presence of school aged children, though, you'd definitely need census data center access, because the chidots data is still a noisy estimate? especially since we're talking about a very specific population, not necessarily uniformly distributed

110

eli knaap @knaaptime.com · 23/01/2026

💯 yeah, for sure. even without access to a census data center, there's a paper in here if i (we?) ever had time, looking at these different methods and the accuracy going from, e.g. tract and downscaling to block using different dasymetric layers and different point process DGPs

120

eli knaap @knaaptime.com · 23/01/2026

ah snap. Thats nice. You still end up with the volumetric problem, but having real parcel data at your disposal is a massive boon

010

eli knaap @knaaptime.com · 23/01/2026

i grew up in urbana... im like 80% sure you can see my childhood home from the hancock observation deck

010

eli knaap @knaaptime.com · 23/01/2026

lololol i meant terrain *over here* 🤣

110

eli knaap @knaaptime.com · 23/01/2026

(aside, as a native illinoisan, one of the things that entertains me about becoming a southern californian is how much quality improvement you get in this kind of approach *just* by considering the natural environment. Because, you know, it has terrain :P)

110

eli knaap @knaaptime.com · 23/01/2026

i like this workflow though because you can choose your dasymetric layer. Maybe I'll play with a building version just for kicks

210

eli knaap @knaaptime.com · 23/01/2026

i can imagine. I considered doing a building version, but then you get the opposite problem where you allocate people to industrial and office buildings. Hard to parse *just* residential. And also you have no height info, so cant densify taller buildings the way you want.

210

eli knaap @knaaptime.com · 23/01/2026

ha, neat! at the moment, this is all the code it takes to do it from scratch (windy city or anywhere else!). All you'd need to do is swap out the NLCD raster for a building layer or similar (though i dont have building footprint/non-building logic as i think you linked)

$python code to create dot density map using the geosnap and tobler packages import geosnap as gsp from tobler.dasymetric import masked_dot_density raster = "s3://spatial-ucr/nlcd/landcover/nlcd_landcover_2021.tif" cols = [ "n_nonhisp_white_persons", "n_nonhisp_black_persons", "n_hispanic_persons", "n_asian_persons", ] chi = gsp.io.get_acs(gsp.DataStore(), msa_fips="16980", years=2021) chi_race_density = masked_dot_density( chi, raster=raster, categories=cols, scale=0.4, pixel_values=[22, 23, 24] ) chi_race_density.sample(frac=1).lb.explore("category", cmap="rainbow")$ racial dot density map of chicagoland

110

eli knaap @knaaptime.com · 23/01/2026

(you can tell i was thinking a bit about @kylewalker.bsky.social 's LEHD maps this week... :P)

110

eli knaap @knaaptime.com · 23/01/2026

the quality of the allocation is really just a function of the dasymetric mask, so this feels like a pretty useful concept i'll stick in `tobler`. This also lets you play with DGPs other than uniform (e.g. clustered poisson), which can get interesting)

100

eli knaap @knaaptime.com · 23/01/2026

here im using NLCD to allocate CA's population to only "developed" areas (pixels 22,23,24), which isnt perfect but works pretty well @kylebarron.dev 's `lonboard` continues to impress. This is 15M points (40% of CA's total pop) and embedding the data in standalone html ends up ~450mb w/ no delay

proportional dot density map showing racial segregation in southern california

proportional dot density map showing racial segregation in the bay area

210