iiinova — where your digital self keeps evolving

For those interested, here are the inflation-adjusted costs (tuition+fees) of 4-year public, 4-year private, and 2-year colleges over the last few decades. From: research.collegeboard.org/media/pdf/Tr...

0122

That doesn't change the basic story of declining/plateauing prices:

100

Public colleges have shown steeper net price drops:

200

It's gratifying to see our work featured on the @nber.org home page. www.nber.org/papers/w35670

0121

This researcher has hit the last name jackpot. Nearly everyone cites her work.

4576

Thanks to @bostonu.bsky.social alum Reed Romanko for sending this my way. To see our pathbreaking but decade-old work in this area, check out: drive.google.com/file/d/1bd-V...

140

Our "A Few Goodmen" paper now has some serious competition, coming from the many Katherines of the world. arxiv.org/pdf/2404.00732

8658

Joshua Goodman @joshua-goodman.com · 23/09/2026

Had a blast visiting @utknoxville.bsky.social to present our work on the academic impacts of private tutoring chains like Kumon and Sylvan. Great to see @ckcarruthers.bsky.social, @mikekofoed.bsky.social, @robertkelchen.com, and others in truly impressive econ and education departments. Go Vols!

082

Published paper: www.sciencedirect.com/science/arti... Ethan Lewis' webpage: sites.google.com/view/ethannl... Full abstract:

020

Joshua Goodman @joshua-goodman.com · 10/09/2026

I aspire to write papers that get heavily cited years before they are even written.

1130

Joshua Goodman @joshua-goodman.com · 04/09/2026

My business model is being threatened.

090

We argue that this work speaks 1) Narrowly to prior lit on tutoring centers; 2) More broadly to prior lit on effective approaches to improving student achievement; 3) Most broadly to growing inequality in child development driven by unequal family investment.

1160

Consistent with this divergence, chain entry widens within-district test score gaps, increasing the White-Black math gap by 0.037 SDs and the White-Hispanic gap by 0.020 SDs.

1120

This overall effect masks substantial racial heterogeneity and is driven largely by White and Asian students. Math scores for White & Asian students rise by 0.040 SDs, while effects for Black, Hispanic, and Native American students are statistically indistinguishable from zero.

1151

Chain tutoring centers have a strong, positive, and persistent impact on student achievement. Following the opening of a district’s first chain center, overall math test scores rise by 0.029 SDs and reading scores by 0.016 SDs. These effects show no pre-trends, are robust to controls, etc.

1172

Tutoring usage is sharply stratified by income & race. Asian HHs are far more likely to use tutoring (3X the national rate). Nearly half of users come from those in the top 10% of income. Conditional on income, White, Black & Hispanic HHs have similar usage patterns while Asian HHs spend more.

1162

Chain tutoring centers have grown substantially since the late 1990s, expanding from roughly 500 locations in 1997 to a peak of over 4,000 by the mid-2010s. Math appears to dominate the business model, accounting for roughly 2/3 of usage (with reading representing the other 1/3).

1131