X50S5

Last September 30th in Trenton, NJ, John Yeh and his brother, Joseph Yeh, were sentenced for their respective roles in perpetrating a $20+ million dollar fraud on the FCC. The Yehs, of Viable Communications, Inc. in Rockville, Md., along with other conspirators and participants, used video relay services (VRS) in illegally running up minutes for fraudulent reimbursement by the FCC.

John Yeh issued a ‘Statement to the Court” on YouTube. While it is signed in American Sign Language (ASL), it is captioned. I direct your attention to a key passage; “It was my intention to help the people of my community. It was my dream to create job opportunities for the Deaf and hard of hearing.” He misstates the true purpose behind Telecommunications Relay Services (TRS), a Congressional mandate. By his actions and others involved in the wholesale fraud of the TRS Fund, they truly have hurt the Deaf/HH community.

Congressional Rationale for TRS Services

Congress originally approved access to the nation’s telecommunications infrastructure for people with disabilities as a part of the Americans with Disabilities Act (ADA), via Title IV. At first, individual states were required to establish relay services for its citizens. Eventually, the regulatory framework encompassed interstate relay, and the TRS Fund currently supports a couple of services such as text relay and video relay.

TRS is intended to allow people with a hearing and/or speech disability to access the nation’s telecommunications networks in the daily course of their lives. That means Deaf people can call up their bank, set up doctor’s appointments, call their families, and to secure mainstream job opportunities.

The TRS program was never designed to create job opportunities for Deaf people in the relay industry. Rather, it was the existence of TRS that allowed mainstream employers to hire, retain, and/or promote Deaf people, knowing that they had access to the telephone. That said, Deaf people continue to work in the relay industry and are invaluable contributors to the engine of growth for the Deaf community in mainstream America.

More importantly, it was a Congressional mandate, a solemn public undertaking. It is not a program designed to ‘help the people of [a certain] community.” It is a service offered to the public at large; openly available for everyone, whether they have a hearing or speech disability, whether they are on a VoIP phone, videophone, computer, or just the plain old telephone. It is a service that connects the American public, cutting across the whole American fabric and way of daily life.

The Regulatory Side of TRS

Unfortunately, there was a period of laissez-faire regulation when it came to the early years of the VRS industry. Within a few years, various individuals and businesses in the VRS marketplace began to game the TRS Fund in lieu of serving their consumers. This led to an ‘arms race’, where companies acquire the Deaf ‘talent’ and the interpreter capital.

Some companies began to view their Deaf employees as profit centers, and competed with other VRS companies in acquiring and retaining interpreters as communications assistants. At the same time, there was a severe interpreter shortage, leading to a big rise in salaries and benefits for them.

Like dominoes, various participants exploited relay workers for their personal gain, and their activities toppled other competitors to do the same. The FCC caught on, and along with the FBI, arrested over two dozen individuals for their fraudulent activities in gaming the TRS Fund for illegal minutes. Millions of dollars were lost. Some were fined and sentenced. Some are still being investigated.

100,000 Stone Skippers in a Lake

The public knows what happened, and the FCC is attempting to rein in the fraud. But, what about the unknown damage that resulted from this illegal gaming of the TRS Fund? Were there any unintended consequences stemming from the actions made by these rogue participants? The answer is yes – allow me to illustrate this with a lake, if you will, with 100,000 ‘stone skippers’.

This lake is the sum total of mainstream American daily life, where the lives of ordinary Americans intersect, regardless of disability. People with disabilities are guaranteed access to most of this lake, thanks to the various state and federal laws governing accessibility to the mainstream societies they live in. However, guaranteeing access is not the same as ensuring that accessibility does indeed happen.

Case in point; the usual accommodation for a Deaf person is an ASL interpreter, regardless of what actual law is being invoked to allow access. Before VRS came into fruition, interpreters were in high demand, as they were needed to accommodate Deaf people in nearly all walks in life, from doctors, courts, their employment, and more… With a tight supply, the prices rose for such services. Despite this obstacle, Deaf/HH people made some inroads into mainstream society.

The 100,000’s of people who skipped stones into the lake were the ones who granted access for a person with a hearing and/or speech disability. By taking action in granting access, they threw and skipped a stone across this lake, leaving behind ripples. Due to the ongoing nature of the accommodation, the stone must ‘skip’ a number of times before settling into the lake.

The participants in the VRS industry were no different than their brethren in the government, public accommodations, and employment arenas. By enabling access to the telecommunications network for Deaf people, they also threw and skipped stones across the same lake.

Ripples Collide

You can easily visualize a picture where there’s a lake and there’s stones skipping, and ripples everywhere on the surface. Eventually, many of these ripples intersect and collide with each other. Some similar ripples (i.e., VRS participants) will join together, amplifying their power. The speed and frequency of the ripples (i.e., fraud) makes them stronger. The force of stronger ripples will eventually overwhelm the weaker ones, often dissipating them.

The nexus behind each ripple collision is the ongoing cost of the accommodation; the ASL interpreter. Already tight when VRS services came into being, the costs soared to unsustainable levels. The FCC, by way of the TRS Fund, were paying top dollar and steady benefits for the interpreter capital, at rates the private and public sectors could not, and were often unwilling to match.

In a span of few years, the lake was now full of ripples created by VRS industry participants; the private and public sectors largely retreated from the lake, and as a result, access to mainstream society for Deaf people was being severely impaired and/or curbed.

Collateral Damage Caused by Fraudulent VRS Activities

This is the irreparable damage caused by fraudulent actions against the TRS Fund by participants bent on personal greed. Who knows how many countless doctor’s appointments were cancelled because they could not accommodate the interpreter request for a Deaf patient? How many Deaf people have died or maimed for life because of unwise decisions made by medical professionals in eschewing ASL interpreters in their course of treatment?

How many Deaf people were denied jobs, promotions, or lost their jobs because their employers decided that the cost of the ongoing accommodations of ASL interpreters were too ‘burdensome’? How many Deaf students attended classes at their local schools with unqualified or no interpreters, because school districts could not or were unwilling to pay for such accommodations? How many Deaf people were denied justice when they appeared before the courts w/o interpreters?

The Unintended Consequence of VRS Fraud

The list goes on. The ripples still continue, albeit with less force, now that VRS reform initiatives by the FCC are catching on. However, the damage has been done. The vast majority of the people who retreated from the lake have not yet returned. Public confidence in the relay industry and the interpreting profession has been shattered. In addition to the fraud, that is the damage these perpetrators have done to the Deaf community; they have harmed them by restricting their access to their mainstream communities.

The VRS reform initiatives are still ongoing at the FCC in Docket #10-51. I invite you to comment on how to best reform the VRS industry for this decade and beyond. There is an important NPRM (11-184) that will profoundly reshape the VRS industry. While I caution the FCC not to have a heavy regulatory hand, they must craft final rule(s) that make way for practical realities imposed upon the American public by the various state and federal laws governing accessibility for Deaf people.

I also call upon all VRS Providers to pledge that they will only handle organic minutes and to serve the public, not gaming the TRS Fund. Public confidence is a fragile thing, and we need to restore our faith in relay services and the interpreting profession.

I truly believe the industry has already started on this path, and in partnership with stakeholders and the FCC, the Deaf community will get all the civil rights they need to access mainstream society in nearly all aspects of daily life. Only then, can this American lake can be full of ripples colliding in equal force, with the blissful harmony of overlapping concentric circles upon its surface.

If you haven’t read Part VII, please do so. This is the final entry in the multi-part series covering the proposed per active-user compensation scheme currently in discussions at FCC with VRS industry participants and stakeholders. To get all eight parts, click on the PAUCS link in the tag cloud. Thank you for reading!

Ostensibly, the FCC is contemplating a compensation paradigm shift from the wildly successful tiered per-minute rate reimbursement scheme to a per active-user compensation scheme. This is presumably done to enforce transparency and accountability. They want the VRS industry to be more consumer-oriented. Revenue influences business behavior in the regulated marketplace, and changes to the revenue levers can lead to unintended consequences or disastrous outcomes.

The Art of Creating Random Data

It is very difficult for humans to create truly random data. Individual(s) and VRS Provider(s) may try to ‘game’ the TRS Fund by engaging in waste and misuse. Some may even resort to outright fraud, but given the recent arrests and guilty pleas, this scenario is increasingly unlikely. The deterrent effect is very powerful in motivating the industry against outright fraud. That said, there will be some ‘grey areas’ where some participant(s) may dabble in, and run afoul of TRS rules.

However, by ‘gaming’ the TRS Fund with waste and misuse, they will attempt to create their dubious call data and mask it with their organic call data, giving it an element of randomness. However, it is very difficult for human beings to engage in dubious activities and blend them into real-time organic activities, as they invariably and unsubconsciously resort to internal patterns. The opposite is also true, humans will also try to discern patterns out of truly random data. There is an excellent book on this subject; Fooled by Randomness, by Nassim Nicholas Taleb.

Machine Learning for the VRS Industry

Enter computers. Computers can test VRS call data for anomalies and abnormalities. More importantly, with intelligently designed algorithms, they can even evolve as they collect more VRS call data and even detect more suspicious data. This is called machine learning, where computers use previous data to analyze current and fresher data. Some human guidance may be needed to set up the initial algorithms and to fine-tune them periodically.

Lately, I’ve been reading some books on this subject. Computer programming is a hobby of mine. Even so, I found these books quite difficult. However, they are now revolutionizing the Internet. If you have used email before, then you’ve seen machine learning at work, where bayesian algorithms filters unwanted emails (spam) away from your inbox. Here are two books I’ve been reading; Machine Learning In Action and Programming Collective Intelligence.

Various algorithms, properly designed, can detect inorganic use in VRS call data, or at least flag them for non-payment until proven legitimate. Some algorithms can even be customized for specific tiers in the VRS industry, or even be provider-spedific. One neat thing about the algorithms is its adaptability; it uses machine learning to continually learn from VRS call data, and to discern ever-evolving and creatively designed inorganic elements out of a huge organic stream of call data.

One Ridiculously Simple Tip in Rightsizing the VRS Industry

Sorenson Communications came up with an elegant solution in policing the industry’s behavior in ‘grey areas’, that are prone to waste and misuse. Over two years ago, Sorenson petitioned the FCC,

“In addition, an eighth new standard would require Internet-based relay providers to use computerized algorithms on a monthly basis to detect anomalous calling patterns that might be the result of minute-pumping schemes or other illicit practices designed to inflate the provider’s minutes. Under this standard, each provider periodically would submit its algorithms confidentially to the FCC and the Fund administrator, thereby giving those entities a broad menu of diagnostic tools to review any provider’s submitted minutes. Since no provider would know what algorithms its competitors had submitted, any provider inclined to submit non-compensable minutes would face a heightened deterrent to doing so.” (Boldface emphasis mine)

I have no idea why the FCC had not adopted such a recommendation and codified it into a regulatory rule. However, something tells me that the FCC is already using computerized algorithms in its enforcement arsenal in scrutinizing VRS provider’s monthly call data. I assume VRS providers are already submitting their monthly call data electronically to the TRS Fund for reimbursement. If that is the case, then the FCC, all in likelihood, is already using machine learning/computer algorithms in scrutinizing such data.

If that is indeed the case, then the FCC knows exactly who are the good, bad, and just plain ugly providers in the VRS industry. Now that they have a new TRS certification regime in place, they will use this information and some VRS Provider(s) may not be able to obtain certification. I strongly suspect they are very data-driven, and are relying on provider submitted call data in determining regulatory issues affecting the TRS/VRS industry.

Mathematical Models or Human Intuition and Wisdom?

Human behavior can be irrational and unpredictable, and artificial intelligence algorithms have been trying to make sense of the human psyche ever since the microprocessor came into fruition. I suspect that is what the FCC is trying to do, making sense out of provider submitted monthly call data and making regulatory decisions.

How else they would float a per active-user compensation scheme and discuss it with industry participants and stakeholders? I wouldn’t be too surprised to see some concrete threshold numbers to come out in the NPRM and their data-driven rationale for an alternative compensation scheme. By using machine learning algorithms, they may have lost invaluable insight into human behavior, which will simply adapt and game the new alternative compensation scheme should it become regulatory law.

If the FCC indeed has become very data-driven in its regulatory decisions, they have failed the Turing Test. Sometimes, there is just no substitute for human intuition and sound judgement in regulating an industry and enforcing its existing rules. While I will try to keep an open mind about their proposed alternative compensation scheme, I am not optimistic about its chances of becoming regulatory law for the VRS industry.

Too much is at stake. The TRS industry is one enduring success story in the regulatory history of the U.S. A key component of this success story is the FCC’s wisdom in implementing a tiered per-minute reimbursement rate mechanism for the VRS industry. Abandoning this touchstone will only introduce instability in the TRS industry, undermine public confidence in the VRS sector, and compromise functional equivalency for roughly 500,000 Deaf/HH Americans who rely on ASL.

If you haven’t read Part VI, please do so.

A Modest Proposal

If the FCC really wants to eliminate 99% of the waste, fraud, and misuse within some participants in the VRS industry, all it has to do is to lower the per-minute reimbursement rate down to a dollar-something a minute. Problem solved! Oh- The VRS industry will simply collapse. So much for my modest proposal, eh? But, it underscores the regulator’s job in maintaining a healthy, robust, and competitive marketplace.

As unbelievable it may sound, the TRS industry is not over-regulated. The banking and insurance industries are examples of extensively regulated industries. The regulations that form the framework of the TRS industry looks like it’s made out entirely of popsicle sticks and Elmer’s Glue as compared to the banking and insurance industries. What they all do have in common, though, is that they maintain public confidence in the regulated industries, ensure that they are available for consumers, and that market forces shape both consumer and business behavior.

Every Person is a TRS Consumer

With apologies to Huey Long’s catchphrase, I suggest that both FCC and the VRS industry is looking at outreach/marketing in somewhat a myopic fashion. To date, industry outreach and marketing efforts tended to target the Deaf/HH community. Nothing wrong with that. In fact, the FCC wants the industry to capture and serve new consumers and expand the VRS market.

What the FCC and the VRS industry needs to be reminded of is… Every single person in the U.S., regardless of their hearing and speech status, are TRS consumers. If the VRS industry even does a modicum of outreach and marketing towards the general public, they promote acceptance and understanding of the VRS market. And in turn, this helps expand the VRS industry.

I loved what Convo Communications did with their billboard in Austin, TX, and their floor ad at the New York Ave. Metro Station in D.C. I love what Purple Communications are doing with employers, educating them about the benefits of utilizing VRS at the workplace. (Admittedly, I learned about this on Twitter. Why didn’t they issue a blog entry or press release about this?) Even if these outreach/marketing efforts do not gain a single Deaf VRS consumer, they have gained mainstream acceptance and understanding about VRS by 1000′s of people.

Unfortunately, acceptance and understanding of VRS does not necessarily translate into actual VRS usage and dollars flowing to VRS industry coffers. Admittedly, the ROI on outreach/marketing efforts on the behalf of VRS companies for the general public is quite poor and hard to quantify. Such efforts expand the potential VRS market, but do not necessarily expand the use of VRS services. I call upon the FCC to spur more outreach efforts on the behalf of the TRS industry, either by themselves or by compensating TRS companies directly for branded outreach/marketing activities aimed at the general public. (Obviously, this direct subsidy to the industry would be capped.)

Finally, the series will come to an end… I will talk about one ridiculously simple idea to rightsize the VRS industry.

If you haven’t read Part V, please do so.

A Proposed Solution – The Power of the Monthly Nut

The VRS industry is an intensely competitive field dominated by one provider. The lesser providers have enormous pressure to make their ‘monthly nut’, i.e., to cover their expenses, and hopefully make a profit. Some individual(s) and participants may resort to unseemly means, i.e. waste, misuse, or plain outright fraud to make their monthly nut.

One proposed solution is to make that monthly nut for certified TRS providers operating in the VRS sphere. Look up comparable prices the mainstream public is paying for VoIP services and have the TRS Fund pay that monthly nut to VRS providers. Allow me to illustrate with an example: VoIP services cost $20 a month.

A VRS Provider with 10,000 consumers would receive $20/month for each consumer for a total of $200,000 a month from the TRS Fund. This VRS provider, knowing that there is a monthly nut, will face lessened competitive pressures. And with it, lessened incentives to commit waste, misuse, or even resorting to fraud. Gaming of the TRS Fund will still continue, but on a lessened basis. Even with the $20/month subsidy per VRS consumer, the savings to the TRS Fund could still run into the million(s).

The rationale for this monthly nut is that VRS consumers utilize end-point hardware and/or software to make VP calls. They have to be assigned a 10-digit phone number. There has to be server infrastructure to facilitate VP/VRS calls. There has to be storage costs for video mails, archiving, contact lists, call history, and more. These fixed costs are amortizable to a large population of users, and a monthly subsidy can be an appropriate substitute to cover these costs from the TRS Fund. It also transforms the provider/consumer relationship to a provider/CUSTOMER relationship, an important distinction, as Deaf/HH consumers currently do not pay for the service.

Some Reservations…

There are caveats, though. Only certified VRS Providers would be eligible for this monthly subsidy. Only VRS Providers that provide end-point infrastructure are eligible. Subcontractors and certified VRS call centers are ineligible. Only one default 10-digit phone number is eligible; if a Deaf/HH person obtains a second 10-digit phone number, there is no monthly subsidy. The monthly subsidy per consumer must be offset by a comparable decrease (10-cents? 15?) in the tiered per-minute rate reimbursement scheme.

VRS Providers must have a credit card number on file. In lieu of a credit card number, then the VRS Provider would need to verify a Deaf/HH consumer’s identity via some other means; i.e., Driver’s License, SSN #. No phantom or duplicate Deaf/HH consumers. Perhaps in 5-10 years, the FCC could phase in a requirement that Deaf/HH consumers pay for this monthly nut, with the USF covering the monthly cost for indigent consumers. I know this last point is a sore point, but it bears repeating… Deaf/HH consumers need to be CUSTOMERS for a vibrant, robust, and responsive VRS industry.

There’s some nagging issues with this monthly subsidy. It does nothing to help VRS providers that are already in Tier III. They already are making their monthly nut and profiting, to boot. It is essentially a naked subsidy to these providers. Also, what of business 10-digit numbers? Do they pay the monthly fee, or does the TRS Fund pick up the tab? Undoubtedly, there’ll be other issues that will crop up with this proposed compensation scheme.

A Bounty on Your Hands!

One thing I do know is that the FCC is keenly interested in expanding the VRS user base. They’ve prodded the industry participants on this issue, questioning them on their outreach/marketing plans on acquiring new TRS consumers. Their big broadband initiative push can only help bring in a new class of VRS users, all of whom would be needing VP/VRS services. And mobile VP/VRS calling is on an upward trajectory, leaving its landline cousins in the dust.

The FCC could propose a bounty ($100? $200?) for each and every new VRS consumer a provider signs up for a new 10-digit phone number. Again, this is limited to certified VRS providers with end-point infrastructure. Some form of verification would be needed, i.e., a credit card, credit check, driver’s license… This would be offset by a corresponding decrease in the per-minute tiered reimbursement rate that reflects marketing/outreach efforts. This way, the VRS industry, instead of poaching each other’s customers, would tailor their marketing/outreach efforts in securing new consumers.

This bounty proposal could be a palatable one for the FCC, and may even cause long-term savings to the TRS Fund. This is because, the more customers a VRS provider has, the more organic their call volume they will be handling, and have lessened incentives to resort to waste, misuse, or fraud. It also boosts functional equivalency, as a whole new class of ASL users are unshackled from their keyboards and are using their natural ASL language to converse with people over the telecommunications network.

Next, I offer a modest proposal and discuss how VRS providers can conduct outreach/marketing campaigns in a new direction.

If you haven’t read Part IV, please do so first.

Ever hear of ‘click fraud’? No, not the kind where friends, content publishers, followers, etc. click on the Google AdSense links to garner some revenue. The kind where competitors bid on Google keywords and then clicking on their competitor’s search results to bankrupt their competition. This is because certain search keywords have high values due to intense bidding, and competitors can click on those links to deplete their competition’s ad budgets or even bankrupt them.

(On a related note… This blog is ad-free. I just don’t want to use computer-generated ads, as VRS providers do utilize such networks.)

The same kind of fraud could potentially happen here. Consumers and competitors that favor certain VRS provider(s) could choose to use their VRS services less than the per active-user compensation threshold. Then, they could utilize VRS services from a competing VRS provider(s) A LOT more than the per active-user compensation threshold, potentially bankrupting them in the process. Admittedly, this is an unlikely scenario, as VRS provider(s) will cut off high volume users. Still, it forces VRS providers to compromise functional equivalency for the sake of their bottom line.

Undoubtedly, this blog entry could be a byproduct of my overactive imagination. Now imagine the intense competition in the VRS industry and the endless creativity and havoc they can wreak upon the FCC’s proposed per active-user compensation scheme. In considering an alternative compensation scheme for the VRS industry, they are potentially exchanging a whole new set of benefits and problems completely different than what they were accustomed to in the current tiered per-minute rate reimbursement scheme. Are the rewards in switching worth it for the FCC, the consumers, and the industry?

Afterword

Next, I talk about proposed solutions the FCC and VRS industry can consider for the proposed per active-user compensation scheme.