Casey McQuillan
@caseymcquillan.bsky.social
PhD Candidate at Princeton interested in labor economics, public finance, and causal inference | GRFP Fellow at NSF | Former RA at NewYorkFed | AmherstCollege '18
Website: https://casey-mcquillan.github.io/
Website: https://casey-mcquillan.github.io/
Thank you for the kind words!
November 10, 2025 at 4:01 PM
Thank you for the kind words!
Another caveat: There are lots of workers in state and local government who were laid off as the direct result of DOGE cuts as well. I don't think people realize how much of the state and local government is actually federally funded
August 15, 2025 at 4:59 PM
Another caveat: There are lots of workers in state and local government who were laid off as the direct result of DOGE cuts as well. I don't think people realize how much of the state and local government is actually federally funded
Thanks Anna! I appreciate the kind words!
July 16, 2025 at 4:21 PM
Thanks Anna! I appreciate the kind words!
The full paper can be found here:
casey-mcquillan.github.io/files/McQuil...
casey-mcquillan.github.io/files/McQuil...
casey-mcquillan.github.io
July 16, 2025 at 4:12 PM
The full paper can be found here:
casey-mcquillan.github.io/files/McQuil...
casey-mcquillan.github.io/files/McQuil...
Lastly, we use the Marginal Value of Public Funds (MVPF) framework to compare the value of expanding UI eligibility against raising benefit levels or extending benefit duration.
We find lowering the eligibility threshold is the most cost-effective UI policy studied to date.
We find lowering the eligibility threshold is the most cost-effective UI policy studied to date.
July 16, 2025 at 4:12 PM
Lastly, we use the Marginal Value of Public Funds (MVPF) framework to compare the value of expanding UI eligibility against raising benefit levels or extending benefit duration.
We find lowering the eligibility threshold is the most cost-effective UI policy studied to date.
We find lowering the eligibility threshold is the most cost-effective UI policy studied to date.
Additionally, we show workers living near unemployment offices see larger gains from UI receipt. This suggests that access to re-employment services provided by these offices may explain why UI receipt led to a more productive job search.
July 16, 2025 at 4:12 PM
Additionally, we show workers living near unemployment offices see larger gains from UI receipt. This suggests that access to re-employment services provided by these offices may explain why UI receipt led to a more productive job search.
Three findings lead us to attribute the increases in hours/earnings to workers finding better jobs:
(i) gains are driven by the next employer after job loss
(ii) more stable employment (fewer employers overall)
(iii) higher hourly wages
(i) gains are driven by the next employer after job loss
(ii) more stable employment (fewer employers overall)
(iii) higher hourly wages
July 16, 2025 at 4:12 PM
Three findings lead us to attribute the increases in hours/earnings to workers finding better jobs:
(i) gains are driven by the next employer after job loss
(ii) more stable employment (fewer employers overall)
(iii) higher hourly wages
(i) gains are driven by the next employer after job loss
(ii) more stable employment (fewer employers overall)
(iii) higher hourly wages
When we look at earnings, we see a similar pattern. The effect on earnings is even more persistent in the quarters following job loss.
UI recipients earn $~15k more than non-recipients over the two years following job loss. (50%⬆️)
UI recipients earn $~15k more than non-recipients over the two years following job loss. (50%⬆️)
July 16, 2025 at 4:12 PM
When we look at earnings, we see a similar pattern. The effect on earnings is even more persistent in the quarters following job loss.
UI recipients earn $~15k more than non-recipients over the two years following job loss. (50%⬆️)
UI recipients earn $~15k more than non-recipients over the two years following job loss. (50%⬆️)
The delay in re-employment is only 20 hours of work in the quarter in which the job loss occurred.
Even more surprisingly, we find large increases in hours worked in the following quarters.
UI recipients work ~600 MORE hours over the two years following job loss. (37%⬆️)
Even more surprisingly, we find large increases in hours worked in the following quarters.
UI recipients work ~600 MORE hours over the two years following job loss. (37%⬆️)
July 16, 2025 at 4:12 PM
The delay in re-employment is only 20 hours of work in the quarter in which the job loss occurred.
Even more surprisingly, we find large increases in hours worked in the following quarters.
UI recipients work ~600 MORE hours over the two years following job loss. (37%⬆️)
Even more surprisingly, we find large increases in hours worked in the following quarters.
UI recipients work ~600 MORE hours over the two years following job loss. (37%⬆️)
However, we can more precisely measure this delay using quarterly hours worked. Re-employment as an indicator is coarsely measured, but hours worked offers more insight into what's going on. And when we do, the results are somewhat surprising...
July 16, 2025 at 4:12 PM
However, we can more precisely measure this delay using quarterly hours worked. Re-employment as an indicator is coarsely measured, but hours worked offers more insight into what's going on. And when we do, the results are somewhat surprising...
UI benefits delay immediate re-employment, but the effect is short-lived.
UI recipients are less likely to find re-employment in the same quarter as their job loss. By the next quarter, there is essentially no difference in the probability of re-employment.
UI recipients are less likely to find re-employment in the same quarter as their job loss. By the next quarter, there is essentially no difference in the probability of re-employment.
July 16, 2025 at 4:12 PM
UI benefits delay immediate re-employment, but the effect is short-lived.
UI recipients are less likely to find re-employment in the same quarter as their job loss. By the next quarter, there is essentially no difference in the probability of re-employment.
UI recipients are less likely to find re-employment in the same quarter as their job loss. By the next quarter, there is essentially no difference in the probability of re-employment.
And how exactly do we do this?
We exploit the fact that workers must record 680 hours of work in their base period to be eligible for benefits. Using a fuzzy RD, we can identify the effect of UI receipt on job search and re-employment outcomes.
We exploit the fact that workers must record 680 hours of work in their base period to be eligible for benefits. Using a fuzzy RD, we can identify the effect of UI receipt on job search and re-employment outcomes.
July 16, 2025 at 4:12 PM
And how exactly do we do this?
We exploit the fact that workers must record 680 hours of work in their base period to be eligible for benefits. Using a fuzzy RD, we can identify the effect of UI receipt on job search and re-employment outcomes.
We exploit the fact that workers must record 680 hours of work in their base period to be eligible for benefits. Using a fuzzy RD, we can identify the effect of UI receipt on job search and re-employment outcomes.
Second, the effects are local to a group of workers "marginally attached" to the labor force, who earned $11k on average in the last year. This may explain why our findings differ from previous work— liquidity and job search likely look very different for this population.
July 16, 2025 at 4:12 PM
Second, the effects are local to a group of workers "marginally attached" to the labor force, who earned $11k on average in the last year. This may explain why our findings differ from previous work— liquidity and job search likely look very different for this population.
First, we identify the effects of UI receipt rather than benefit generosity. This distinction matters because variation in receipt creates differences in weekly search requirements and access to re-employment services. Variation in generosity alone does not get at this.
July 16, 2025 at 4:12 PM
First, we identify the effects of UI receipt rather than benefit generosity. This distinction matters because variation in receipt creates differences in weekly search requirements and access to re-employment services. Variation in generosity alone does not get at this.
Our results are surprising because existing work has shown more generous UI delays job finding with little evidence that this leads to better jobs.
Two novel features of our setting may explain why we come to a very different conclusion.
Two novel features of our setting may explain why we come to a very different conclusion.
July 16, 2025 at 4:12 PM
Our results are surprising because existing work has shown more generous UI delays job finding with little evidence that this leads to better jobs.
Two novel features of our setting may explain why we come to a very different conclusion.
Two novel features of our setting may explain why we come to a very different conclusion.
Younger people may be able to tolerate this risk, but this is devastating for older workers trying to plan for retirement
April 11, 2025 at 6:04 PM
Younger people may be able to tolerate this risk, but this is devastating for older workers trying to plan for retirement
The dot com bust in '01 had a similar slow recovery, and it barely recovered by the time the Great Recession rolled around
April 11, 2025 at 6:04 PM
The dot com bust in '01 had a similar slow recovery, and it barely recovered by the time the Great Recession rolled around
If you had bought in just before The Great Recession, then it would have taken six years to return to that level. Even if you bought six months in when it seemed like it had stabilized, you would've waited four years
April 11, 2025 at 6:04 PM
If you had bought in just before The Great Recession, then it would have taken six years to return to that level. Even if you bought six months in when it seemed like it had stabilized, you would've waited four years
"Best week since '23" ?!?!
April 11, 2025 at 5:28 PM
"Best week since '23" ?!?!