4 ms·

Looks like the "credible" report is an approximation of a series of guesses and omissions. Measuring economic impact of almost anything is difficult but this on

Looks like the "credible" report is an approximation of a series of guesses and omissions. Measuring economic impact of almost anything is difficult but this one seems more vague than usual. If you want to discuss overall impact, you can't simply leave out things that are hard to measure but still wind up with a solid conclusion.

JumpCrisscross 9y ago
> If you want to discuss overall impact, you can't simply leave out things that are hard to measure They're not leaving out things that are "hard to measure." They're being honest about the limitations of their data: "Employers are required to report actual hours worked for employees whose hours are tracked (i.e. hourly workers), and report either actual hours worked or total number of hours assuming a 40 hour work week for employees whose hours are not tracked (i.e. salaried workers)" [1]. These data allow these researchers to make incredibly accurate measurements about the price and quantity demanded of labour. Because "the data identify business entities as [unemployment insurance] account holders...firms with multiple locations have the option of establishing a separate account for each location, or a common account." This means the researchers could only "identify business location only for single-site firms and those multi-site firms opting for separate accounts by location." Excluding multi-site single-account businesses from the analysis avoided muddying the precision of the hourly and quarterly wage data with geographic uncertainty. In any case, the study only purports to measure employment effects amongst "single-site businesses," which "as shown in Table 2, in Washington State as a whole, single-site businesses comprise 89% of firms and employ 62% of the entire workforce (which includes 2.7 million employees in an average quarter)" [2]. [1] http://www.nber.org/papers/w23532.pdf http://www.nber.org/papers/w23532.pdf page 13 [2] page 14

JumpCrisscross 9y ago

> If you want to discuss overall impact, you can't simply leave out things that are hard to measure They're not leaving out things that are "hard to measure." They're being honest about the limitations of their data: "Employers are required to report actual hours worked for employees whose hours are tracked (i.e. hourly workers), and report either actual hours worked or total number of hours assuming a 40 hour work week for employees whose hours are not tracked (i.e. salaried workers)" [1]. These data allow these researchers to make incredibly accurate measurements about the price and quantity demanded of labour. Because "the data identify business entities as [unemployment insurance] account holders...firms with multiple locations have the option of establishing a separate account for each location, or a common account." This means the researchers could only "identify business location only for single-site firms and those multi-site firms opting for separate accounts by location." Excluding multi-site single-account businesses from the analysis avoided muddying the precision of the hourly and quarterly wage data with geographic uncertainty. In any case, the study only purports to measure employment effects amongst "single-site businesses," which "as shown in Table 2, in Washington State as a whole, single-site businesses comprise 89% of firms and employ 62% of the entire workforce (which includes 2.7 million employees in an average quarter)" [2]. [1] http://www.nber.org/papers/w23532.pdf http://www.nber.org/papers/w23532.pdf page 13 [2] page 14