William Nordhaus, 1941-2026

I was saddened to hear of the passing of William Nordhaus last Friday. Nordhaus won the Nobel Prize in 2018 "for integrating climate change into long-run macroeconomic analysis", and it is his work on climate change and economics that he might be best known for. However, that was far from his only contribution to economics. My ECONS102 students might recognise Nordhaus as one of the first economists to formally analyse the tradeoff between strong intellectual property rights, which provide incentives for creation, but lead to under-consumption, and weak intellectual property rights, which can lead to optimal consumption, but lead to under-investment in idea creation (for more on that tradeoff, see this post).

My favourite Nordhaus contribution was actually some work he did on the change in the cost of light over time. The original research paper is available here, but the story is best told in this article from the BBC:

Back in the mid-1990s, an economist called William Nordhaus conducted a series of simple experiments with light.

First, he used a prehistoric technology: he lit a wood fire.

But Prof Nordhaus also had a piece of hi-tech equipment with him - a Minolta light meter.

He burned 20lb (9kg) of wood, kept track of how long it burned for and carefully recorded the dim, flickering firelight with his meter.

Next, he bought a Roman oil lamp, fitted it with a wick, and filled it with cold-pressed sesame oil.

He lit the lamp and watched the oil burn down, again using the light meter to measure its soft, even glow.

Bill Nordhaus's open wood fire had burned for just three hours on 9kg of wood.

But a mere eggcup of oil burned all day, and more brightly and controllably.

Why did he do this?

He wanted to understand the economic significance of the light bulb.

His research showed just how dramatic the fall in the cost of light was. Drawing on his work, the same BBC article notes that 60 hours of labour:

...had once produced the equivalent of 54 minutes of quality light [but] now produced 52 years.

That fact demonstrates just how powerful technological progress and economic growth have been over the very long run. But Nordhaus's deeper insight was that simply looking at prices and economic growth alone can lead us to miss the enormous improvements in living standards. The significance of that cannot be overstated.

I'm surprised that there doesn't appear to have been much coverage of Nordhaus's passing. Nevertheless, Yale News has a good obituary. And while searching for some others, I ran across this podcast from earlier last month, that neatly covers his life in the way a good obituary would.

Nordhaus's work extended far beyond the climate economics he is most known for. His research on the cost of light is a reminder that seemingly ordinary things can nevertheless tell important stories. He will be missed.

Is it better to publish more papers, or longer papers?

Academics get rewarded for their research output. More papers, published in higher-ranked journals, is a large part of the metric against which academics are judged. Notice that there are two aspects to that metric though - quantity (more papers) and quality (higher-ranked journals). Since it usually requires more effort to successfully publish a paper in a higher-ranked journal, is there a trade-off between quantity and quality of publications?

This 2014 article by my colleague John Gibson, published in the journal Economics Letters (sorry, I don't see an ungated version online), asks a related question about how we measure the quantity of research output - is it better to publish one long paper, or several papers that in total have the same number of pages? To answer that question, Gibson links the 2010 salaries of 223 economists from the University of California system to their publication histories (in terms of the number of publications, the quality of journals they were published in, and the number of publication pages), while controlling for seniority, experience, gender, campus, and whether the economist held a named chair or was a Nobel Prize winner.

Gibson notes that there is a wide range of journal ranking and weighting schemes, so he applies nine different ones. He finds that:

Six of the nine equations show a significant positive effect on salary from publishing more articles, conditional on the total number of quality-adjusted pages. The size of the effect is not trivial; on average, a standard deviation increase in the count of articles holding pages constant increases salary by 15 log points.

Ok, so publishing more articles is associated with a higher salary. And the difference is not small - 15 log points is equivalent to a salary increase of about 16 percent. What about more pages? On this, Gibson finds that:

...if an econometric ‘horse race’ is run between total pages and counts of articles, Vuong non-nested tests favor the model using total pages as being closer to the truth, even in the regressions where counts and pages are both statistically significant.

In other words, the total number of pages published seems to be a better predictor of academic salary than the total number of articles published, even though the number of articles also matters when holding the number of pages constant. How much does this difference matter? Gibson does some back-of-the-envelope calculations, finding that:

The predicted salary with a [Journal of the European Economic Association] article added to lifetime output is $195,980. Adding just as many pages to the literature with three Economics Letters articles increases predicted salary to $196,970—an annual gain over the single-article strategy of $990. This average economist is 16 years into their career, so with a 40-year career and five percent discount rate, the net present value of publishing three Economics Letters rather than one JEEA is $13,700.

Of course, these results are correlations rather than causal estimates, so we can't conclude that splitting a long paper into several shorter ones would necessarily increase an economist's salary. Nevertheless, the results imply that not only does quantity of research seem to matter, but how that quantity is broken up matters too. So, having lots of small articles published may lead to a higher salary for academic economists. That conclusion aside, one of the best aspects of this paper, you may have noticed, is that Gibson published it in... Economics Letters.

This week in research #150

Here's what caught my eye in research over the past week:

• Antler, Bird, and Fershtman (with ungated earlier version here) develop a theoretical model of dating and marriage
• Bowden, Rodriguez, and Weingarten (with ungated earlier version here) find that academic leniency increases high school GPA, but significantly reduces effort among lower ability students, based on lower attendance and lower numerical grades

PhD students' growing use of generative AI in STEM subjects

I see a large number of initial research proposals, sent to me as part of a cold-emailing strategy by prospective PhD students. The vast majority of those proposals are now obviously written by generative AI. For the most part, that makes the proposal largely worthless in terms of judging the quality of potential PhD students. If anything, it is a negative signal about their ability to conduct primary research. If they can't write a short initial proposal without significant help from generative AI, how will they cope with writing research papers and a thesis?

No doubt I am being too harsh in my approach. However, when you receive three or four cold emails per day from prospective PhD students, as I do about this time of year, you have to be a bit ruthless. I know some colleagues that delete those emails unread. I at least want to give these prospective students a chance to capture my attention.

My own experience is based on a sample of one. What does the evidence tell us about the growing use of generative AI by PhD students more generally? We receive an answer to that question in the context of STEM PhD students in this new NBER Working Paper (sorry, I don't see an ungated version online) by Daniel Gross (Duke University), Dror Shvadron (University of Toronto), and Hansen Zhang (Duke University). They collected data on US-trained PhD students from 2019 to May 2026, ran their dissertation abstracts through the AI-detection tool Pangram, and then collected data from each student's LinkedIn profile to determine their post-graduation outcomes. The growth in generative AI writing in dissertation abstracts is striking, as shown in Figure 1 from the paper:

Prior to 2023, almost no dissertation abstracts are flagged as having AI content. By the first five months of 2026, 29.4 percent of abstracts contain AI writing (at least, according to Pangram). Looking across fields, AI adoption has been most rapid in civil engineering and computer science (both around 50 percent in 2026), and lowest in mathematics and statistics (17 percent in 2026).

Gross et al. then dig into the characteristics of the PhD students using generative AI in their dissertation abstracts, finding that:

Non-native English-speaking students use AI far more intensively than native speakers... Among 2026 graduates, 41% of dissertations by students educated in non-English-speaking countries have AI-generated abstracts, against 23% for students from English-speaking countries...

AI use also increases at institutions with lower research intensity and programs of lower rank. In the 2026 cohort, 28% of dissertations at R1 universities have AI writing, compared with 31% at R2 universities and 37% at other doctoral institutions, and this gradient is particularly strong among US-origin students (foreign-origin students write with AI at high rates at all institutions...).

Gross et al. then look at where students go after receiving their PhD, finding that:

Among US-origin graduates, those with AI-written dissertations are 7.4 p.p. less likely to be employed in academia in 2026 (17% over the 44% base rate) and 8.0 p.p. more likely to be employed in industry (16% over a 50% base), with no difference in government employment. These differences attenuate substantially for likely foreign-origin graduates. We also find differences within academia: conditional on academic employment, graduates with AI-written dissertations are 5.4 p.p. more likely to hold non-tenure-track positions (on a base of 12%), with no significant difference in the likelihood of tenure-track or postdoctoral positions.

Gross et al. explicitly caution against taking a causal interpretation of the effect of generative AI use on student outcomes. I didn't interpret the results that way, instead seeing them as perhaps indicating that students with dissertation abstracts written by generative AI may be lower-quality students, or those with less of an academic focus for their future employment. Alternatively, some students may simply be using AI to overcome language barriers, or to polish their writing, rather than to substitute for their own research skills. However, the results don't definitively establish either of those interpretations either. 

What the results do show is more banal - PhD students are increasingly using generative AI. Of course, it is important to keep in mind that these results tell us about the prevalence of AI-generated writing in dissertation abstracts, rather than the dissertation more generally. Since generative AI is quite good at summarising, and that is essentially the purpose of an abstract, I would expect AI-generated writing to be more prevalent in abstracts than in the main text of the dissertations.

Finally, while these results apply to STEM subjects, it seems likely that this is indicative of a more general trend across all fields. If this trend continues, it will not be long before AI-assisted dissertation writing becomes the norm rather than the exception. What we should be wondering now is whether that trend will hamper PhD students in their development as researchers, or not.

[HT: Marginal Revolution]

Read more:

• Lessons from Joshua Gans on AI for economics research
• This is not how generative AI should be used in research