A few weeks ago we noticed that “HRSA made it harder for NAP applicants to shoot themselves in the foot;” now it appears that DOL is getting in the game. In this year’s YouthBuild SGA, DOL includes a form called “WORKSHEET_weighted_average.xlsx,” which models what previous YouthBuild SGAs have only instructed applicants to do regarding unemployment rates. Years ago applicants could do pretty much whatever they wanted regarding unemployment rates, using any data sources, but over time DOL has gotten more and more specific, presumably so that they’re comparing homogeneous numbers.
Today, calculating weighted average unemployment rates isn’t hard, exactly, but we’d bet that DOL got all kinds of interesting, incompatible responses to these instructions, from the 2015 YouthBuild FOA:
The applicant must provide weighted average unemployment rate (rounded to one decimal place) of the combined cities or towns identified as part of the target community(ies) compared to the national unemployment rate as of the latest available comparable data. This data is broken into two youth age subsets: 16 – 19 and 20 – 24. Applicants will have to average the unemployment rate for these two age groups by adding the populations together and then dividing by the total population.
We know how to model this in Excel, but we shouldn’t have had to: DOL should’ve included a template long ago. Last year we wrote a post about how “Funders Could Provide Proposal Templates in Word,” and doing so would likely raise the quality of the average proposal submitted while simultaneously reducing the busy work of applicants. Funders aren’t incentivized to do this, save by the knowledge of what they’ll get if they don’t provide templates, and consequently they don’t.*
Still, there are downsides to the the DOL approach. Applicants must now collect and aggregate specific data points for all the zip codes they’re serving, rather than choosing a different geographical unit, like a city or county, that ordinary humans understand. Few people say, “I really love living in zip code 66666.” But they might say, “Austin is great!”
Those of us who’ve done data work on large numbers of zip codes know how irritating that can be. I’m thinking of a particular project I worked on a couple months ago that had dozens of zip codes in the target area, and I never could figure out how to really expedite the process via the Census’s powerful, yet maddeningly Byzantine, website. There was (and is) probably an efficient way of doing what I was doing, but I never figured it out. The Census website is hardly the first piece of software with fantastically sophisticated abilities that most users never learn because the learning curve itself is so steep.
Overall, though, the simple, included form in this year’s YouthBuild SGA will probably lead to better proposals. We’re a little sad to see it, though, because conforming to the form makes it harder for crafty grant writers like us to weave threads of cherry picked and obfuscated data into an elegant, but sometimes specious, needs assessment tapestry that is coin of our realm.
* Given the unstated role of signaling in proposals, which we write about at the link, funders might be incentivized to make the grant process harder, not easier.