USAC annual report

So I was just looking at the annual report that USAC sent up to the FCC. Here are the factoids that jumped out at me:

• The E-Rate is costly to administer: While the E-Rate gets less than 25% of the total disbursements from the fund, it uses up over 75% of the fund's administrative costs.
• The cost of administering the program more than doubled between 2001 and 2006. But it's still only 1.35% of the total fund, which seems reasonable. The E-Rate administrative costs are 3% of the total, which is kind of high. Maybe we can simplify the program?
• Solix gets almost, but not quite, half the money that the SLD takes in. Two things about that:
1. It seems low: Solix has more employees and does more work than the SLD. But I guess most of the rest of the money pays for audits of one kind or another.
2. It's a little unsavory: The FCC gives money to a non-profit subsidiary of NECA (the phone companies' lobby group), which then turns a good deal of the money over to a for-profit subsidiary of NECA. On the other hand, I don't know who else could run the program, or would want to.
• I can't understanding something in the report. On page 12 there is a table that shows that 89.28% of funding went to applicants not classified as Urban or Rural, because "either FCC Forms 471 that did not include this information, or to FCC Forms 471 shared by both rural and urban entities (that therefore could not be classified)." Well, you can't file a Form 471 without including information on Urban or Rural; Item 3 in Block 4 is mandatory. So 89% of applicants include entities in both rural and urban counties? I don't think so.

Waive this! I dare you!

Man, I love three recent appeals to the FCC. I like them because they don't make excuses, they just ask for a waiver.

The first appeal asks to be allowed to purchase Internal Connections without a contract. Gutsy move. The SLD denied it, of course. But they've appealed to the FCC, saying that the FCC's rules don't actually require a contract for Internal Connections.

The second asks for an invoice deadline extension for a school that forgot to file for reimbursement for FY2004.

The third appeal is even better. Basically, it says, we violated a lot of rules, but it was just a bunch of mistakes. Best of all, it points out that the 30% rule comes down hardest on those who are trying to understand the rules and are honest.

I love these appeals because they're pushing the FCC to find out if there are limits to the FCC's willingness to waive rules. None of the appeals make any excuses about illness, military service, or anything else. They all just say, "We made an honest mistake. Please cut us some slack."

486 hell

I understand a new 486 is due this summer, but it can't come fast enough for me. Here's what we go through every time we want to file a 486.

First, we're working on reasonably peppy Windows machines with lots of memory, Internet Explorer 7 and Adobe Reader version 6. I know that the 486 works better with Reader v5.5, but v5.5 is no longer supported by Adobe, and has a critical security flaw which will not be patched. So I won't run that on any of my machines.

OK, so we go to the SLD site, click on "Apply Online," click on "486 PDF," click on "Continue," click on "Continue" and wait. Two minutes and 40 seconds later, the PDF appears. I have an assistant fill out the Forms 486, and apparently she can do several in one sitting just by saving and opening a new form.

I'm not so lucky. I open an existing 486 and submit, then E-Cert. If I then retrieve another existing 486, it opens, but the "Submit" button doesn't work. So I have to close all IE windows, then open them, go back through the SLD clicks, and wait another 2:40. Now the old 486 that I was working in comes in, sort of. So I have to Retrieve the form again, but Submit still doesn't work. But if I close IE, reopen, click, click, click, click, wait 2:40, then I can Submit and E-Cert. Just so we're clear, I've now had to open the 486 3 times, with a total waiting time of 8 minutes. Oh, and during those 8 minutes, Internet Explorer slows to a crawl for any other optn windows. (Here's a great expampl: I'm typing this as I wait for a 486 to load, and so I can type in a whole sentence before it appears on the screen.)

Eight minutes a form. And that's if all goes well. I've just had to use Task Manager to force Adobe Reader to quit in order to get control of my computer back.

And I can't contact Adobe support. They would tell me (quite correctly) that I should be on version 8. But if you go above 6, then the forms won't print out correctly.... And it doesn't seem to work better, anyway. Sorry, I just had to vent. Now I have to post this before I use Task Manager to force Internet Explorer to quit, and start all over. I'm about a half hour into this, and so far I've certified only 2 forms. I hate to create more paper, but it looks like I'm going to have to go back to certifying on paper.

FCC says no

Yesterday's FCC ruling on Sprint's appeal is noteworthy only because it harkens back to the bad old days.

Sprint was asking for an exemption for PBXes installed as part of a telecommunications service prior to 2003. When the FCC established the "on-premise Priority One equipment" loophole in the Tennessee Order, Sprint saw it as an opportunity to get Priority One funding for PBXes, and installed PBXes on client premises. Then came the 3rd Report and Order in December 2003, which explicitly said that PBXes were not eligible for Priority One funding.

Of course, the contracts Sprint signed way back when were long-term, so that the cost of the PBX was spread across many years. So when the 3rd Report and Order came along, Sprint's customers were stuck with long-term contracts that were suddenly not eligible for E-Rate.

Sprint's appeal was pretty reasonable, really: they asked for an exemption on contracts signed before the 3rd Report and Order came out.

And the FCC said no. I'm surprised at two things in the ruling. First, that the FCC said no. While there have been a few denials sprinkled in among all the appeal approvals in the past few months, for the past 15 months or so, the FCC has been very generous in granting appeals.

Second, the FCC treated the "no-PBX" rule in the 3rd Report and Order as a clarification of existing rules, applying it retroactively. This is the way the FCC operated in the past: we know that we didn't state the rule until 2003, but it was a rule before that, just not explicitly stated. In recent months, the FCC has not taken this position. In the case of vendor selection, for instance, the FCC has held that applications before the Ysleta Order don't have to follow the rules in that Order. I blogged about this a few months back.

So this decision took me back to the bad old days when the FCC denied almost all appeals, and treated every new rule as a "clarification" which applied retroactively. Let's hope it's not the pendulum swinging back.

AT&T and Orphan Annie's dog

What do they have in common? ARF!

The FCC has approved AT&T use of its ARF system. Three things bug me about it.

First, as described in an earlier post, AT&T led a delegation of the big telcos to the FCC arguing that clients don't have the resources to handle the reimbursement process, while the appeal that was just granted claims AT&T doesn't have the resources to handle the reimbursement process. (The ARF system forces the client to do all the work compiling info for a SPI filed by AT&T.) They want to have their cake and eat it, too.

Second, the AT&T appeal says applicants will receive payment in 6-10 days, while their Web site tells applicants to expect payment in 10-20 days. Which is it?

Third, AT&T said the reason it didn't want to offer discounting on bills is that it would cost $3 million. That seems like a lot of money, but the ARF system has costs, too. Let's say it only takes 10 minutes to get through the ARF process; that's pretty conservative, but possible. AT&T says that applicants can do it monthly, so that's 120 minutes (2 hours) per year. I did a quick search, and it looks like AT&T is the service provider on over 5,000 FRNs each year. That would mean applicants spend 10,000 hours each year ARFing. And thousands of checks have to processed by applicants each month. Now a one-time cost of $3 million doesn't seem so high.

Waives becoming a tsunami

Did you see the FCC's latest wholesale waiving of the rules? They've granted the appeals of 44 applicants who didn't get forms filed on time. Mostly, it's just a follow-up to Bishop Perry, but it pushes things a little further.

First the FCC laid down some guidelines: "Petitioners assert a waiver is appropriate for one of two reasons: (1) someone on the applicants’ staff failed to file on time due to misunderstanding or personal emergencies, or (2) the delay in the filing or the receipt by USAC of the FCC Form 471 was due to circumstances out of the applicants’ control."

OK, that's mighty lenient, but leniency is good.

However, in grouping the appeals, the FCC create a group of applicants who “indicate that relevant members inadvertently failed to file the application forms in a timely manner.” This seems to be saying that you don’t even have to say the dog ate your 471; just say “oops” and you get a waiver.

Check out the appeal from Easton Public School. It was filed 4 months after the window closed, and offers no reason why the form was not filed.

I’m all for leniency, but I’m starting to feel like a sucker for working so hard to get 471s in on time.

2-in-5 working?

The SLD has come out with their funding Demand Estimate for 2007-2008, and I'm trying to come to some conclusions, but not really succeeding.

I was really looking to see if I could detect the effect of the 2-in-5 rule on demand, but I couldn't really do it. Since the 2-in-5 rule took effect for Funding Year 2005, Priority 2 demand has been dropping slowly. There was a precipitous drop between FY 2004 (pre-2in5) and FY 2005, which could indicate that the 2-in-5 worked. But there was an almost equally precipitous drop between FY 2003 and FY 2004, when 2-in-5 was just a twinkle in the FCC's eye. I think the drop was caused by all the news stories about abuse, and people realizing that the E-Rate was not just a blank check. (In fact, the 2-in-5 rule came out late in the FY2004 application window, and some of my clients jacked their Priority Two requests because they saw 2-in-5 coming.)

But in the end, I think 2-in-5 has been a failure. The idea was to push Priority Two funding to a wider range of applicants. Unless I'm mistaken, in FY 2004, applicants down to 80% were funded for Priority Two. The Priority Two threshold hasn't gotten below that since, and I'm predicting it won't for FY 2007, either.

Meanwhile, the 2-in-5 rule increases the complexity of the program.

I'd rather see the FCC adopt the recommendation of the Taskforce on Waste, Fraud and Abuse: Cut the top discount rate for Priority Two funding to 80%. I noticed that the discount cap was not in the FCC's big NPRM (Notice of Proposed RuleMaking, a request for public comment on ideas for new rules) 18 months ago. Apparently the large 90% districts have killed it.

Rant time: 90% is too much. I used to work as a tech director in a 90% district. If I went to the Supt. and asked for $1,000 to replace a dying server, I had to show an urgent educational need. But if I went in and said I could $9,000 in funding if the district ponied up $1,000, I got the green light with no questions about urgent need.

As a consultant, when I talk to tech directors about their plans for internal connections, they talk about a switch here, a server there, replacing a broken component on the PBX. When I tell them they can get a 90% discount, their eyes widen and they start talking about replacing their entire voice and data infrastructure and going wireless to boot. (By the way, I don't see the same behavior with telecommunications: no one adds more phone lines just because they're cheap.)

In general, that's what the E-Rate is supposed to do: tilt the decision-making balance away from crushing budget realities and toward the technological needs of the districts. But 90% is too close to free.

In my opinion, 80% would be a big improvement, but is still a little too close to free. I'd like to see equipment discounts capped at 70%.

FCC filings from the Bizarro World

So here's a little slugfest that's right out of the Bizarro world. (For those of you who don't read Superman (or watch Seinfeld), the Bizarro world is an alternate dimension where everything is the opposite: Superman is an ugly bad guy, etc.)

First, the ALA (American Library Asscociation) says that applicants should file invoices, not service providers, and get paid directly from USAC. (Item 13 (page 24) in their NPRM comments.)

I'm completely with them on the "get paid directly" part; if I choose to fill out a BEAR, why do I need to have service provider sign off on it and have the check go through them? When service providers submit an invoice, they don't have to get the applicant to sign off, and checks come directly to them. The rules seem to reflect an attitude that service providers can be trusted, but applicants can't. And to become an applicant, you have to prove that you are a school or library recognized by the state, while to become a service provider, you just have to file a form. (While I'm on this rant, here are some telling numbers: certifications required on a BEAR: 4 applicant, 2 service provider; on the SPI: 0 applicant, 0 service provider.)

But why does the ALA want to do away with discounts on bills? I think we should leave the choice with the applicants. Discounts are a great choice for lots of applicants.

Then (16 months later) AT&T leads a pack of large phone companies into the offices of the FCC to oppose the ALA's proposal to have applicants fill out the invoices, because "the large service providers do not believe that most applicants have the staff necessary to prepare and submit invoices and respond to SLD invoice reviews in a timely fashion." That's bizarre. I mean, these service providers sign off on thousands of BEARs every year. They know damned well that some applicants can handle the work.

So in this bizarro world, the ALA is saying, "make the libraries do all the invoicing work," and the big phone companies are saying, "let us do all the invoicing work."

To make it more bizarro: these are the same service providers who for the first six or seven years of the program claimed they were unable to discount bills. The whole BEAR process only came into being because the service providers said they couldn't manage the invoicing.

The bizarro crown: AT&T to this day requires applicants to manage the invoicing process. When applicants were given the right to choose discounted bills or reimbursements, AT&T filed an appeal, which I think is still pending, saying that it couldn't discount. AT&T does have the ARF system, which will generate a SPI, but the applicants have to do all the work of calculating and entering all the information that will go on the SPI after having paid the bill. As far as I know, AT&T still doesn't discount bills. (How do they get away with that, anyway? The SPI directions make it clear that you can only submit a SPI if you have sent out a discounted invoice. AT&T is using the SPI to allow applicants to apply for reimbursement.)

How did we get in this bizarro world, and why is AT&T raising a stink now, so long after the ALA first proposed the changes? Well, I noticed that some time in Dec. or Jan., AT&T removed the signed BEAR page 4 from their Web site, and started requiring applicants to fax in BEARs for signature. It was a sad day for applicants; up until that time, you could fill out a BEAR, print a copy of the page 4, date it, and send it in. It saved the hassle of faxing back and forth to get the service provider signature. I wonder if someone from the SLD or FCC twisted AT&T's arm, or if some AT&T lawyers noticed that they were certifying forms without looking at them and had a cow.

Since they now handle BEARs like everyone else, AT&T has realized that for recurring services, BEARs are a larger pain than SPIs for responsible service providers. Especially if you set up a system like AT&T's ARF where you have your clients fill out the SPI for you. So now they want the BEAR process to end.

Why do the other service providers want to kill the BEAR? Well, there is the work issue: I think the BEAR process is a little more work than the SPI process, once you have your accounting system set up to discount bills.

But the real issue is the float. It used to be that it tooks months for USAC to process invoices, which meant that service providers who discounted were essentially floating a loan to USAC for the time between the discounted bill and the payment of the SPI. Whereas with the BEAR, the applicant was floating the loan from the time the undiscounted bill was paid until USAC paid the BEAR. (As an added bonus, at least one of the large service providers often takes six weeks to turn around checks that it receives from USAC, which means that USAC is floating a loan to the service provider, which is money floated to USAC from the applicant.) So when it took USAC 3 months or so to process SPIs, service providers who discounted were floating a lot of money.

Now USAC's average invoice processing is less than a month. On-Tech's clients are schools and libraries, and let me tell you, it is very rare that our invoices get paid within a month: first payment of my invoice has to be approved at a board meeting, then the check has to be approved at another meeting. I see a lot of applicant phone bills, and it is routine for them to take over 30 days to pay. I assume that service providers have realized that USAC is actually paying more quickly than schools and libraries. So naturally they would prefer to invoice USAC directly.

BEN hurts

Could we stop using the word "entity"and the acronym "BEN"? It all started when the SLD assigned Billed Entity Numbers to buildings. Wrong. School and library buildings are not billed entities, and billed entities are not buildings. Billed entities are usually associated with an address (the central office), but any billed entity can have multiple buildings.

It's like treating "company" and "office" as synonyms. IBM is a company with many offices. The individual offices are not companies. And IBM is not an office, it is a company. USAC is a company with one office. But USAC is not an office, and "2000 L Street, N.W., Suite 200" is not a company.

Now there seems to be some sort of kludgy distinction between "entity" and "billed entity," but it causes confusion. Here's how to clear it up.

First, don't call buildings "entities." Since two buildings on the same campus can be counted as one location, let's use the term "location" to describe buildings. Every location gets a Location Code. The simplest thing would be to use the current BEN. (The database administrator in me would prefer that all the codes be the same length, but applicants are familiar with their BENs.)

Second, don't call organizations "billed entities." Call them organizations. Every organization gets an Organization Code. You could use the BEN, but I think it would be better to use the FCCRN, because right now BENs can denote both a location and an organization (there is often not a separate BEN for the main branch of a library system).

To clear things up, the SLD should delete all the duplicate FCCRNs that were created when the SLD told us all we had to get an FCCRN for each school, which is not really possible, because FCCRNs can only be assigned to organizations. So what really happened was that applicants created duplicate FCCRNs with different addresses. By the way, the FCCRN mess was created because when the FCC said that all "entities" must have FCCRNs, the SLD misunderstood and thought that all locations (which are still called "entities" and are assigned a Billed Entity Number) needed FCCRNs.

Then when you apply, the FCCRN is used everywhere except Block 4 of the 471, and the Location Code is used in Block 4. Clear. Terminology matters.

Admittedly, it's an inconvenience for applicants with only one location, because they would have two IDs to keep track of. But with a little improvement to the online 471, the Location Code would autofill for such applicants. [Say, why doesn't Block 4 autofill now? Yeah, we might have to do some editing, but for 90% of applications, the work would be done.]

It's a major overhaul, but if they started now....

2-in-5 rule hell

First off, I'm surprised at how applicant-friendly the SLD's new 2-in-5 rule changes are. It creates a whole bunch of new procedures for the SLD, and it seems to me they've gone beyond what the FCC ordered. But maybe the SLD figured that the FCC was going to make them do it, anyway, and this way they could avoid appeals.

There was one unfriendly wrinkle: The 30% rule will be combined with the 2-in-5 rule, so if you put a building in Block 4 that's used up its 2 years along with two buildings that have not, then the 30% rule means the whole FRN is denied. Harsh. (OK, let's see, 2-in-5 could be written 2/5, so it's the 30%/2/5 rule, which your calculator would tell you is the 3% rule, but if you first convert 30% to .3 and 2/5 to .4, then you have the .3/.4 rule, which would be the .75 rule. Dang, even the nickname gives me a headache.)

The ability to go back and remove buildings from Block 4 of previous applications is really applicant-friendly Applicant-friendly, yes. Simple, no. It's going to be way beyond what most applicants will want to do. Which, of course, is good news for consultants like me.

The SLD said that if removing a building would lower the discount level below the funding threshold for that funding year, then the whole FRN gets denied. But it doesn't say what happens if the removal causes the discount to go up....

I would say that we can expect a lot of appeals because of the 30%/2-in-5 combo and the retroactive threshold denial, but this process is so beyond most applicants, I don't think many applicants will be trying it.