Monthly Archives: March 2015

Batch/cohort versus continuous training: A problem with no solution

Job training programs, education programs, and related programs can work in two basic modes: batch/cohort (we’ll call it “batch” for this purpose) and continuous. Batch training happens the way most conventional schools function: the academic year starts at a particular time—usually in September—and if you don’t show up by September 5, you have to wait until the next break in the academic calendar (which is usually around January). No matter how bad you want to start school, you have to wait until the next time you’re allowed to start.

The alternative is a continuous program, in which a given participant starts whenever she’s ready to start. Two people might start in September, five in October, another in November, and three in December. The person who starts in November probably can’t work or learn effectively with the two who start in September, however, because the two who start in September are too far ahead of the one who starts in November.

Neither of these approaches is necessarily right, and two federal programs illustrate the difference: YouthBuild versus Training to Work 2-Adult Reentry, both of which are conveniently funded by the Department of Labor.*

YouthBuild wants batch processing: usually one class starts every year, and training takes about nine months to complete. Training to Work 2, like many prisoner reentry and Workforce Investment Act (WIA) programs, wants continuous training: if an ex-offender is released in October, it’s important for reintegration purposes to start that person in October.

Batch processing is hard because people who think they want to participate in October lose interest by the time January rolls around. Continuous processing is hard because people tend not to have the sense of solidarity that comes with working in concert with others towards a specific goal.

One problem with many Workforce Investment Act (WIA) programs, going back to WIA’s inception in the mid-80s, is that they’re drop-in, drop-out programs; no one develops a sense of team. Following intake and assessment at a WIA American Jobs Center (AJC), the client is usually referred to to a vocational training vendor. The training usually is starts immediately but may not be continuous, and the trainee may not be part of a training batch.

It may seem to the trainee that they’re actually not building towards anything concrete, as she lacks a cohort to share training outcomes with. Cohort issues are powerful: Even with semesters at a university, for example, many people still find the university experience alienating, especially coming from relatively small high school communities. To some extent living in dorms provides community; so can sports, or the Greek system (despite the problems with the Greek system).

The military puts every recruit through basic training in a batch, in large part to build some sense of team identity or “unit cohesion,” as this often referred to in the military. The cliché goes that guys on the ground don’t charge the enemy for their country or glory or the girl back home; they charge for the guys around them. Building a cooperative unit out of individuals is inherently hard and the many federal job training efforts don’t always work to build cooperative units. Repeated interactions build knowledge and to some extent happiness. It also builds cooperation, as numerous iterations of prisoners dilemma, divide-the-money, and similar game-theory games.

In WIA-land, however almost all programs work on a continuous-entry, continuous-exit model in which any individual is on his (usually) own. Most WIA vendors are in the meantime operating off-the-shelf training programs. These programs can be better-run or worse-run, but they do get people started quickly. In this sense they aren’t doing as much cherry-picking as batch programs, which require more patience. But because they require more patience, they may get better outcomes due to selection biases.

When YouthBuild was first released, HUD (which ran YouthBuild at the time) didn’t require batch training. But HUD changed the second YouthBuild NOFA to reflect the very first YouthBuild proposal we’d ever written (I was about 10 at the time, so my contribution was limited), because Isaac had been involved in job training proposal writing for years and knew that batch training would be easier for YouthBuild trainees and grantees. HUD read our proposal—we wrote some of the fist funded YouthBuild grants—and realized that our approach was a winner. We like to think we had an important contribution to the way in which YouthBuild operates its training, though almost no one knows this.

We can’t tell you the right approach for your program. But we can tell you that you should be thinking about the trade-offs involved in either approach, and you should be closely reading RFPs so you can divine whether the funder already prefers one approach or the other.


* And Training to Work is on our mind because the Training to Work 3 – Adult Reentry FOA was just released.

Parsing the Department of Education’s “Developing Hispanic-Serving Institutions” (HSI) Program RFP–Which Colleges are Eligible?

As we’ve written before, parsing an RFP sometimes seems like deciphering the Talmud. The just-issued ED Developing Hispanic-Serving Institutions (HSI) RFP is a case in point.

HSI is a venerable program that provides grants to Institutions of Higher Education (ED-speak for “two- and four-year colleges and universities”) deemed to be “Hispanic-Serving Institutions.” But what is an HSI? To paraphrase President Clinton, it depends on what the meaning of “HSI” is? The RFP states:

In addition to basic eligibility requirements, an institution must have at least 25 percent enrollment of undergraduate full-time equivalent (FTE) Hispanic students at the end of the award year immediately preceding the date of application.

(Emphasis added.)

Now we have to determine what “award year” means. On page 19 of the 87-page RFP, we finally learn that award year “refers to the end of the fiscal year prior to the application due date.” Which raises the question, why doesn’t the RFP just consistently replace “award year,” which no one understands, with “end of the last federal year,” which anyone involved in federal grants knows is September 30?

This conundrum came up on Friday when I was talking about HSI with the internal grant writer for a community college we often work for. This guy is very knowledgeable about federal grants but thought the eligibility for HSI was that his college had to have at least 25% Hispanic students for one year before applying for a HSI grant. His college achieved that milestone at the start of the fall 2014 semester, or around September 1, so he didn’t think they were HSI eligible. A close reading of the RFP sections above shows that he was wrong: as long as the college met the 25% threshold by September 30, 2014, which in this case they did, the college is actually HSI-eligible.

It also turns out that ED does not certify or even maintain a list of HSIs. Instead, applicants self-certify eligibility by signing an assurance. How does a college know whether is has 25% FTE Hispanic students? The students themselves self-certify their “race and ethnicity” at the time of application and these data are aggregated by colleges.

This data gets really murky. Most Americans probably think “Hispanic” is a “race.” Not true, at least by some metrics. Those of us who work with Census data know that the Census definition considers “Hispanic” an ethnicity, not a race. From the Census website: “Hispanic origin can be viewed as the heritage, nationality, lineage, or country of birth of the person or the person’s parents or ancestors before arriving in the United States. People who identify as Hispanic, Latino, or Spanish may be any race.”

In other words, American college students self-certifying as “Hispanic” could have a partial family heritage anywhere from Spain to South America to the Philippines and many places in between. From a Census “race” standpoint, they could be otherwise black, white, Asian, Native American, or multiracial. Combined with immigration and intermarriage, this is why the population of some states, like California and Texas, either are or will be majority-Hispanic. As a practical matter, most IHEs in the southwest and south are likely HSI-eligible already; in a few more years, most IHEs across the country probably will be. This is great news for IHEs, Hispanic students and grant writers!

The above cautionary tale shows why it’s critical to closely read RFPs regarding applicant eligibility and other key factors. When I went through Air Force basic training over 45 years ago, the first class we took was “Rumors and Propaganda.” It taught us not to believe barracks scuttlebutt. The same is true in grant writing.

Some Positive Changes in Federal Grant Grants.Gov Submission Requirements Emerge

Faithful readers know that I’ve been writing federal grant proposals since the last ice age.* For most of the last four decades, federal grant writing has changed little, other than in obvious tech-related ways—computers, online databases, quick and reliable digital literature/data searches, easy access to applicant background info and so on. I recently realized that incremental changes, glacial in speed though they may be, have begun to have a cumulative impact on the way in which proposals, and especially federal grant proposals, are prepared.

The biggest change in federal proposal preparation was the switch from hard copy to digital submissions, starting around 12 years ago. While at first there were a number of portals developed by various federal agencies, over the years most, but not all, have switched to Grants.Gov. For the first five or so years, Grants.Gov was incredibly badly coded, and the upload process was often uncertain and dicey. In recent years, the reliability of Grants.Gov has improved dramatically and the required attachments mercifully streamlined.

In the bad old days of paper submissions, federal agencies usually required a zoo of attachments, like target area maps, evaluator CVs, key staff resumes, job descriptions, organization charts, evidence of 501(c)3) status, bylaws, financial statements, letters of support, MOUs, and the dreaded logic model. The narratives themselves were long, like attention spans back in those days, with maximums sometimes reaching 50 single-spaced pages. It was not uncommon to end up with a 150-page grant application, which usually had be submitted with a “wet-signed” original and up to ten copies. Sometimes the FedEx boxes we shipped to HUD or the Department of Education weighed over ten pounds.

In the early days of Grants.Gov, these attachment requirements continued, making the upload process very complicated (try uploading a 10 megabyte financial statement attached to a Grants.Gov kit file for example) and sometimes impossible, as there might not be an attachment slot for a given required attachment. As time passed, federal RFPs began to strip away attachments or even require only a couple of consolidated attached files. This is much simpler and makes the grants.gov kit file preparation easier and significantly more reliable.

Most federal agencies have also reduced the maximum length of the narrative and settled on a double-spaced, single-sided page formatting convention. For example, Department of Labor proposals now usually have 20 double-spaced page maximums for narratives. Before you say “hallelujah,” however, keep in mind that the RFPs themselves have not gotten any shorter–an RFP could easily be 150 pages, with the questions to be answered in the 20-page narrative actually being many pages longer than the maximum allowed response. It is often harder to write a shorter narrative of, say, 20 pages, than 40 pages, because the writer faces the “building-a-ship-in-a-bottle” problem. Furthermore, despite severe page limitations, all of the headers/sub-headers must be included to enable reviewers to easily find your responses.

Other interesting RFP changes involve the objective and evaluation sections, which are sometimes combined and always intertwined. Until recently, most RFPs let the grant writer essentially make up the objectives. Now, however, many ED programs like Student Supportive Services and HRSA programs like New Access Points provide more or less fill-in-the-blank objectives. I’m fairly sure this trend is to facilitate “apples-to-apples” comparisons by reviewers, but whatever the reason, it makes it easier to stay within the page limit. While evaluation section requirements used to be astoundingly complex, these days, RFP evaluation instructions tend to be much more straightforward and linked to specified objectives.

Now for the bad news. The budget and budget narratives sections have changed little. Grants.Gov kit files still use a variation of the venerable SF-424A budget form, which is actually a summary of federal object cost categories. To create the 424A, any sane person would use an Excel template. The only people in the US who do not seem to grasp the concept of a spreadsheet are federal RFP writers. There is still no federal Excel SF-424A template provided, although we use versions that we’ve developed over the years.** A well-laid-out Excel line-item budget not only displays each line item within each cost category, but it can also double as the budget narrative. See further in “Seliger’s Quick Guide to Developing Federal Grant Budgets.”

The budget narrative instructions in almost all federal RFPs are written as if the response is to be done on a typewriter, circa 1975. The budget narrative is also often excluded from the narrative page limit, with no page limit on the budget narrative. Although we would never do this, we’ve seen proposals from our clients in which the budget narrative is longer than the program narrative. Don’t do this—unless you think the tail wagging the dog is a good approach to life.


* Forty-four years to be exact, but who’s counting?

** We always provide clients with a draft budget in a handy reusable Excel template—which is one good reason to hire us!

The HRSA Health Infrastructure Investment Program (HIIP) Illustrates Why It’s Hard to Handicap Chances of Getting a Grant

Anyone who’s been to a race track or Vegas knows that the odds of a given race or sporting event are being constantly updated by pros who seem to know how to handicap future events. Prospective clients often ask me to handicap their chances of winning a grant competition (and we’ve written before about why grant writing is not like the Olympics). Trying to handicap a particular grant competition is like trying to handicap a horse race in which you don’t know the horses, riders, or venue until after the race is completed. If grant writing was really like a horse race, you’d just pick the cutest horse or jockey with the best colors and hope for the best.*

A prospective client raised the odds issue on Friday, regarding the recently issued Health Resources and Services Administration (HRSA) Health Infrastructure Investment Program (HIIP) FOA. HIIP has $150,000,000 available, with about 175 grants up to $1,000,000, for Federally Qualified Health Centers (FQHCs). FQHCs are sometimes called “Section 330 grantees” and provide primary health care to publicly (Medicaid) and uninsured patients. HIIP is a great opportunity for FQHCs: there’s a lot of money up for grabs, the grants are large, and the money is for facility improvements (facility improvements are always hard to fund).

Not surprisingly, we’ve received a number of inquiries from FQHCs. On Friday, a FQHC CEO in rural Montana called. I learned a bit about his agency and provided a fee quote. Then he popped the question: “So, what are my chances of being funded?” As I was starting my standard reply to this standard question, he interrupted. He said he didn’t think his chances were very good, because “thousands of FQHCs would apply.”

I said that’s not true, since there aren’t that many FQHCs. We got into a bit of a tiff over this, so I double checked after the call. The Henry J. Kaiser Family Foundation says there were only 1,202 FQHCs as of 2013. I would’ve guessed closer to 1,000, but the numbers are in the same ballpark. While new FQHCs are created every year, there are likely less than 1,300 today. Thousands of FQHCs can’t apply for HIIP because not that many exist. My caller was trying to talk himself out of applying.

Let’s try estimating the likely competition.

For various reasons, not every FQHC will want to apply for a HIIP grant. Some are already happy with their current facilities, while others are undergoing leadership changes. Let’s assume that 1,000 FQHCs want to apply and that HRSA will ultimately make about 175 grants. This would mean around a 20% chance of any given application being funded, which is pretty good odds in submitting a grant proposal or buying a lotto ticket.

But, of the hypothetical 1,000 or so applicants, many will not finish their applications, so perhaps 700 applications will actually be submitted. Of these, a fair number, say 100, will be technically incorrect and will not even be scored. Now the pool is down to 600. Many of these will be poorly written, fail to demonstrate need, etc., and will not score high enough to be funded. Let’s assume that 350 – 400 score high enough to be funded.

Now the odds are close to one in two!

Still, grant handicapping is more complex than this simple analysis. Of my theoretical 400 potential grantees, some will be urban, some rural, some will serve special populations (e.g., homeless, Native Americans, etc.). Some will serve African Americans, some Hispanics and so on. Since, like all governmental funders, HRSA is a semi-political entity, the organization wants to spread the sugar. Even if the top 200 applications, based on points alone, were somehow clustered in the Northeast, applicants in other areas would still be funded.

My 400 possible grantees are actually competing against similar applicants, rather than all applicants, because not all applicants are equal in the eyes of HRSA administrators. If your FQHC is the only highly scored applicant that serves rural Native Americans, your chances of being funded could be 100%. If your FQHC serves a general population in a large city like New York or LA, you might be one of ten possible grantees in that city. HRSA will likely make multiple awards in a given big city, but not ten. Now your odds could be one in three. This particular exercise can be played ad infinitum, but it doesn’t mean much because no one outside of HRSA knows the organization’s subjective priorities in advance and because you don’t know who else is going to apply.

Not knowing who else is going to apply really counts. If four other FQHCs similar to yours operate in a given region, they may all say they’re going to apply—just to scare you, or intimidate you, or impress you, or for any number of other reasons. Will they? Maybe, maybe not. You can’t control them, and we recommend that you not be dissuaded by their rhetoric. They may claim to have juice with power players in Washington, or any number of other advantages. You don’t know and can’t know if they’re telling the truth.

My advice to all callers is the same: if your agency is eligible and you want to provide the service, you should disregard real or imagined odds and apply. The logic is similar to seeking a new job. In most cases, you don’t know the other job applicants. Most people apply for jobs they want to do in places they want to live. Say you’re a highly qualified lion tamer and there is a great job open at a circus in Seattle. You should only apply if you like rain, coffee, and tech / nerd culture. If you like sunshine, Cubano sandwiches, and salsa dancing instead, wait for a circus opening in Miami.

The same is true for HIIP: FQHCs who need facility improvements should complete technically correct and compelling proposals that are submitted on time. Worrying about the odds is an interesting but pointless enterprise.


* This is actually the way I bet at horse races, which is why I’m not much of a gambling man.

One Way You Know a Grant Maximum is Too Low: HRSA and “‘Now is the Time’ Project AWARE-Community”

In 2014 HRSA released a program melodiously called “‘Now is the Time’ Project AWARE-Community,” and the program had almost 100 awards available for an eight-figure pot of money—but the individual maximum grant was only $50,000. Last week, HRSA released the same RFP, but with different funding parameters: 70 awards are available with a maximum grant size of $125,000—or 150% more than last year’s award.

We’re guessing that the maximum award changed because $50,000 was just too little money to get most organizations interested in the program, which is designed “to train teachers and other school personnel to detect and respond to mental illness.” Fifty thousand dollars, once overhead and administration is accounted for, won’t even yield a full-time trainer. The current maximum grant, $125,000, will. The program just got a lot more compelling for both nonprofits and school districts. HRSA is also signaling to applicants that they know the last funding round didn’t offer large enough grants to be interesting.

Links: Bread Bags and Poverty, the App World, College, and Jobs, Jobs, Jobs

* When Bread Bags Weren’t Funny, or, we are now spectacularly rich in ways that rarely make the news.

* “Is Snapchat Really Confusing, or Am I Just Old? A 32-year-old’s hopeless quest to understand America’s fastest-growing social app.” This describes me, and I too remember old people telling me when I was younger about life before computers and so on, “What’s the point?”

* Dubious, polemical, yet: “Today’s Apps Are Turning Us Into Sociopaths.” See also Facebook and cellphones might be really bad for relationships.”

* “Why college isn’t always worth it: A new study suggests the economic return on a college degree may be a lot more modest than you think.” This better matches anecdotal yet seemingly universal observation, and it better matches work like that in Paying for the Party. The more I learn about college and about pre-school education the more skeptical I am of either as panaceas.

* What life is like for non-sports fans; a shockingly good metaphor.

* “American Schools Are Training Kids for a World That Doesn’t Exist.”

* College students use social media to be anti-social.

* Cops murder a guy on camera.

* “Orchestra in Los Angeles gives disadvantaged youth a lifeline through music.” Never before has such a project been tried!

* “Meet the [Washington State] Sex Workers Who Lawmakers Don’t Believe Exist,” from The Stranger and probably SFW.

* Employers want better technical writers but aren’t getting them.

* “Why GM Hired 8,000 Programmers.”

* “Lesbian” takes testosterone, sees personality and ideology change. This is not the piece’s actual title.

* Robots aren’t yet taking all our jobs because there aren’t enough smart human engineers to operate them. Which is too bad: the future in which we have all our material needs met and can spend all our time making art.

* “Scientists know there are more giant craters in Siberia, but are nervous to even study them,” which may be the most important article you’re going to skip.