Research

When a county grants a tax abatement to a data center, the fiscal consequences extend beyond the government that approved it. School districts, municipalities, and other overlapping jurisdictions share the same tax base but typically have no role in granting the incentive. We study how those costs are distributed and how affected governments respond. Using a panel of 3,080 U.S. counties from 2000 to 2023, we link data center entry to the finances and municipal debt issuance of local governments within each county. Instrumenting data center entry using state incentive adoption interacted with preexisting county fiber and power infrastructure, we find that host counties collect no more revenue after a facility arrives, while school districts offset part of the loss in locally raised revenue through higher intergovernmental transfers and municipalities receive no comparable compensation. At the state level, increased support for affected school districts is accompanied by lower per-pupil aid more broadly. Data center entry also changes the composition of county borrowing: counties become less likely to issue debt requiring voter approval, but only in states where such approval is mandated by law. Together, the results show that the fiscal cost of a local tax incentive is shifted across overlapping governments and that voter-approval rules shape how those governments absorb and respond to that cost.

How does AI manage household stock portfolios? We collect a prospective, daily time-series of stock recommendations using several LLM’s and study AI’s investment style. AI recommends undiversified portfolios that positively load on momentum, large companies, and low book-to-market firms. AI primarily recommends stocks based on how much media attention firms receive. Using multiple versions of queries and requests, we find that buy-and-hold and actively managed AI portfolios do not appear to earn statistically-significant abnormal returns based on the methodology of Daniel et al. (1997). 

Immigration stimulates economic growth but also increases demand for local public resources. We examine the net effect of immigration on local governments’ fiscal health using a Bartik shift-share design built from U.S. Census records since 1880 and ad￾ministrative data on unauthorized immigration. We find that immigrant inflows lower municipal bond yields, improving access to finance. Yield reductions occur for both legal and unauthorized immigration and are larger in counties with tighter labor markets and greater financial slack. Over longer horizons, immigration increases employment and improves operating margins as growth in sales tax revenue and state intergovernmental transfers outpaces expenditures.

We use minute-by-minute Bloomberg online status data to characterize two important dimensions of sell-side equity analysts’ work habits: we estimate the average workday length (AWL) to proxy for analysts’ general effort provision, and we use the percentage of away days (PAD) to proxy for their private information collection. Both AWL and PAD improve forecast precision, a causal result that we confirm using the COVID lockdown as an instrument. Frequent traveling not directly related to firm events is positively correlated with the likelihood of becoming an All-Star analyst, suggesting that institutional investors also value private information transmitted through in-person interactions.

We examine whether wildfire smoke affects sell-side financial analysts’ earnings forecast pessimism. Using plausibly exogenous county-level variation in satellite-based smoke plume coverage matched to analysts’ office locations, we find that each additional smoke day around an earnings announcement lowers an analyst’s forecast relative to peer consensus by 0.083 percentage points—about 29% of the interquartile range of relative forecast optimism. Cross-sectional tests point to a salience mechanism: smoke-exposed analysts ask more climate-related questions on conference calls, and the effect is stronger for firms with greater climate-risk disclosure and recent environmental violations. Analyst workload does not moderate the effect, and controlling for air quality does not attenuate it, inconsistent with a physiological-impairment channel. The smoke effect diminishes with prior exposure, consistent with adaptation, and is robust to alternative smoke windows and outcome measures.