A High-Tech Agenda for President Obama
On Tuesday Barack Obama becomes the 44th president of the United States, during a scary economic meltdown and at a time when America questions its role in the world and its self-perception as a champion for good and progress. Among his many agenda items as president is one for technology. His incoming administration has signaled it sees technology both as an enabler of the change it seeks for the U.S. and as a worthwhile investment in its own right.
The Obama administration's focus, at least publicly, has been on whether and how to extend the reach of Internet access via national broadband and net neutrality policies. That set of issues certainly should be on the national high-tech agenda, but Obama needs to take a wider view, as well as be wary of the high-tech industry's self-interests, many of which will not further the national interest. Here are my personal recommendations, based on 25 years' involvement in the tech industry as an observer, journalist, and occasional advocate.
Don't let the high-tech industry lead In economics and foreign policy, there are competing schools of thoughts that marshal think tanks, academics, activists, and business leaders to argue over and propose policy. The high-tech industry does not do this except in a very simplistic way. That fact is the biggest danger Obama faces in formulating and executing a high-tech agenda.
The tech industry is largely a libertarian one, which encourages creativity and respects differences among people but is perfectly happy with abusive monopolies, socio-economic imbalance, and profiteering. Its worst impulses are borne of meritocracy gone awry: The smartest succeed, and everyone else accepts what the meritocracy decides. Politically, this has kept Silicon Valley, Route 128, and other high-tech centers focused on keeping government out of the way so they can pursue economic and technology domination unfettered. Companies like Apple, AT&T, Comcast, Oracle, Microsoft, Sun, and Verizon are genuinely puzzled when people object to using their success in technology areas to lock out competitors and lock in customers to an ever-widening area of their own products. In the extreme version of this worldview, government is bad and/or incompetent, and customers are cattle.
But the high-tech industry also has a neo-socialist component, which distrusts both government and business. The positive aspect of this ideology engendered the open source movement and provided some balance to the techno-meritocracy in areas such as privacy and information access. But its extreme also promotes dubious ideas such as making software and Internet access free for all that are simply unworkable in the real world.
Both ideologies take for granted that technology is good, and if left unfettered goodness will prevail. There's an extreme naivet
Instead, start with the policy agenda -- what is good for the nation and people as a whole -- and put the high-tech industry in the position of having to deliver on that policy agenda. It will do better executing than leading. With that in mind, my recommended tech agenda starts with policy proposals, not with technologies per se.
Broadband and net neutrality The telecom industry and activist community have fought for years over national broadband and Internet access. The telecom industry has fumbled this technology, trying to game the system for years to prevent competition for both the actual pipes and for the information services that ride over them. The telecom industry has failed in providing compelling information services, and its pipes are overpriced. But the activists' proposals that the broadband infrastructure should essentially be a government-provided benefit for everyone is nonsensical; we don't do that for water or electricity, after all, which are more necessary for human survival.
Yes, every American should have access to broadband for Internet access. The Internet is as much a public utility as water, electricity, gas, and telephone. And it should be treated that way. The telecom industry has failed in making broadband near-universally available, favoring high-margin deployments in rich areas. The result is price gouging and limited availability. The Obama administration's public musings about encouraging broad availability is on the right track.
Infrastructure is expensive, and expecting for-profit companies to spend lots of money to build infrastructure in areas with low returns is foolish. It should be no surprise they avoid rural and poor areas, given their investor demand for high profits and margins. So it makes sense for the government to subsidize the infrastructure development in such areas, as well as set national minimum requirements for broadband bandwidth -- but not for the specifically technology used to deploy it.
In many cases, it will make sense to subsidize the incumbent providers rather than set up new quasi public agencies or rely on startup companies, but that decision should be made on a case by case basis. In some cases, it makes more sense to extend the incumbents' technology; in other cases, there is no capitalist rationale for providing service, so the government should step in and treat it as what it is: a public good that relies on public monies or tax transfers (such as is done with phone service access charges today).
Of course, the incumbents should compete for the subsidies like anyone else. And there should be strong strings attached to any subsidies, such as maximum pricing for the basic service tier, a prohibition against tying that service and its price to use of other offerings (as the feds forced AT&T to accept as a condition of approving its BellSouth purchase, but as they have not forced the cable companies to do). The government may also need to create a subsidy system for the poor to pay for access, as it has long done for other utilities with the concept of lifeline pricing for qualified citizens.
The other half of this issue is net neutrality. Activists rightly object to carriers charging carriage costs based on content, knowing full well the carriers will put competitors for services such as streaming media and VoIP at a disadvantage vis
At the same time, the minimum service level the government should mandate must be high enough so poor users aren't frozen out of "normal" services, including access to streaming media. People should be able to download music and movies, as well as have VoIP service and regular Internet access for news, commerce, and so on. A minimum speed of 1.5Mbps up and 768Kbps down and a minimum allotment of 250GB per month should suffice for that. This minimum level should be raised over time as the infrastructure capacity grows, of course.
Carriers that offer information services should be required to spin out subsidiaries -- as many have already done -- so there's no ability to use carriage fees to subsidize those services at the expense of competitors. If AT&T charges a VoIP provider 10 cents per megabyte for carriage, it must charge its VoIP subsidiary no less than the same. In other words, equal access and equal rates per usage unit must be enforced.
High-tech jobs and education The U.S. has severely depleting its manufacturing skill base and capacity in a wrong-headed belief that it could rely on the rest of the world to make its goods and have the population rely solely on white collar and service jobs. Silicon Valley has been making the same argument for moving white-collar tech jobs go elsewhere as well.
This train must be stopped now. Service jobs are under siege from automation and low pay, and the white-collar information-worker jobs that were supposed to replace skilled labor jobs are also now moving offshore. Plus there's the risk of being held hostage by providers of key goods and services that can no longer be made here. We make a lot of noise about creating energy independence, well information technology independence is just as critical.
This issue is a murky one, full of misleading arguments. The biggest one is that there is a shortage of qualified U.S. workers, which requires companies to hire overseas. The result is that older tech workers are often replaced by younger, cheaper, often imported or outsourced workers. Ostensibly it's because their technical skills haven't kept up with fast-moving technologies, but we all know that much of it is because older workers are costlier and less flexible in terms of working hours and locations. The high-tech industry prefers to throw these workers away and import cheaper ones, rather than help figure out how to keep their skills current so their wisdom and experience can be brought to bear. After all, wisdom and experience often lead to faster, better results than the brute force of the eager but inexperienced all-nighter-inclined young. Sure, there are some areas where the U.S. doesn't have the talent available, and industry should be able to hire overseas when it can't hire here. But I believe this situation is much less common than the tech industry would like us to believe.
Not so shockingly, as a result of the offshoring moves by industry and the ever-increasing pressure on domestic IT workers, the proportion of Americans going for high-tech education is declining. Why get a degree for a job that will be shipped overseas or pay minimum wage (when you factor in the unpaid overtime) to remain competitive with India or China? At this rate, the U.S. education system will be exporting high-tech know-how overseas, ironically using foreign teachers to do so (since American high-tech educators are also in decline). This is not a strategy for remaining a high-tech leader or having the ability to use technology to drive the economy forward.
Addressing the problem is not easy: A protectionist approach will drive the many foreigners who contribute to the U.S. knowledge economy away, and could actually accelerate the transfer of tech jobs overseas. The goal should be to keep a healthy IT sector in the U.S., so we retain a viable long-term talent pool and IT industry.
Through policies that encourage the use of U.S. information workers and discourage the use of overseas resources for purely economic reasons, the government can create a sufficient homegrown demand and thus maintain a viable talent pool. When foreigners are hired because allegedly no Americans are able to do the job, the government must enforce that the salaries paid are in fact the same as for American workers; the truth is that the foreign workers are typically paid yes. If the pay is truly the same, that will reduce the number of workers imported under false pretences.
Unemployment insurance rates should be higher for companies that cut U.S. staff and then move or outsource jobs overseas. And given the savings that U.S. companies get in terms of health ands retirement benefits, perhaps they should also continue to pay Medicare and payroll taxes for those jobs moved overseas, as part of leveling the playing field so that overseas hires are truly based on who's best for the job. Or perhaps companies that exceed a certain portion of nondomestic labor would be ineligible for the many tax breaks the federal government gives for research and other investments -- after all, the reason for the taxpayer subsidy is that the benefit accrues to the nation.
Tax breaks for the retraining -- and hiring -- of older U.S. workers will also help. The hiring part is key: Any displaced older worker knows how hard it is to find a new job in a new skill area, given they are competing with cheaper, younger candidates. The retraining per se isn't enough.
I have no illusions that this is social policy, where the government designates information technology as a strategic national capacity. But we all know it is, so let's act accordingly. We can do so without resorting to Ugly Americanism.
National health records system Despite its title, the 1996 Health Insurance Privacy and Portability Act does nothing to make health records portable. It does establish standards for the handling of medical records, which created a huge compliance effort for many businesses. But the real problem remains unresolved: a national standard for medical data that everyone would have to use, allowing health records to flow easily among doctors, hospitals, insurers, and so on.
The benefits are simple: a standard health record will reduce bureaucracy costs, reduce the chances of errors, make it harder for insurers to play coverage and reimbursement games, and to simplify the delivery of patient care.
Yet 12 years after HIPAA was enacted, we don't have the instrument needed at its core. So break the impasse and make it happen. It should be a straightforward effort for the American Medical Association and insurance industry to propose a health record standard and for the federal government to mandate its use, with a clear deadline for execution. So get to it. And forget relying on the unwieldy system of regional health information organizations (RHIOs) to lead the health records effort, which were meant to avoid a federal mandate and let regional associations essentially compete to develop the best approaches. RHIOs can be effective in leading the implementation of national standards, and extending them for regional purposes (the California RHIO is a good example of that).
A national identity and privacy system The U.S. today has a fiction that there is no national identity card, which somehow is a good thing. We do have one (the Social Security number), but because we pretend we don't, we have a very inefficient system for maintaining and accessing information on ourselves. Big businesses, the government itself, and spy agencies around the world already have the resources to connect the information stored on each of us for data mining, surveillance, and other purposes -- but individual citizens must maneuver through an impossible maze to get and correct financial, employment, health, credit, and other such information. The lunacy of this is the fact that the major credit reporting agencies make us pay for access to the financial information about ourselves, while they are free to sell it to pretty much anyone they want and have few burdens of ensuring its accuracy.
There should be a national identity system, preferably one that uses some sort of secure card so it can be used across various business and government systems. Passports, drivers' licenses, and so on could be replaced with a single card, one that likely would use good old two-factor authentication -- biometric data and a password -- to maintain its security both at the user end and at the transaction end.
Such a card could even be used as a rechargeable cash card, saving the government much of the expense of printing, storing, and transporting cash and saving merchants the expenses of handling cash or the high fees for debit card transactions. Leave space on such cards for personal and business information, and you could use it in lieu of credit cards and insurance cards. Make card readers available for personal computers, and you can secure online transactions, from voting to e-commerce to tax preparation (imagine being able to access all the relevant tax data and have it downloaded into one place, rather than have to rekey it from paper forms into your tax software).
Yes, this would require a large card-management system to allow password changes and so on, but the financial industry has long used such systems, so the issue is an implementation one, not a fundamental technology one. And it would end the waste in having so many separate systems for drivers' licenses, passports, and so on.
At the same time, we need a national privacy regime. Citizens should own the information about them, so those who gather and trade in such information must follow national standards on privacy protection, such as citizen opt-outs for reselling the information, and provide online access through an at least partially standardized user interface so citizens can view the information, correct it, and monitor who is using it -- at no cost to the citizen. Yes, that means we'll all get fewer preapproved credit offers in the mail, but that's a good thing.
Improved e-government Where the tech industry can make a big difference is to reduce the friction of citizen interaction with government, which would save the government, businesses, and individuals time and money. Paying taxes, complying with various disclosure rules, and so on should be easy to do, while being secure. Activities such as promoting XBRL for public companies' financial disclosure and requiring electronic submission of tax forms are good examples of what should be encouraged, but the feds could do better in making the information systems easier to connect to and use.
Here's just one example: The Social Security Administration has a powerful retirement calculator for Windows that is impossible for a lay person to use, and a simple Web application that doesn't really help in retirement planning. Why not redo the sophisticated application as a Web app, in a way that most retirees could use? That would be more efficient for them, and require less effort in Social Security offices to do the same calculations for those who make an appointment.
The trick is in the systems design and delivery. Governments have had spectacular tech implementation failures, such as the ongoing saga at the FBI. And even successful systems, such as the Veterans Administration's Vista health management system, have been stung by data corruption issues. Part of the blame lies at contractors who overpromise and underdeliver -- a hallmark of the defense industry, and an outcome the Obama administration needs to prevent in its tech spending.
But the government itself also deserves blame. Often, regulators overcomplicate such systems, making them impossible to create or at least use. The feds must not repeat the mistake it made with the Electronic Federal Tax Payment System, which makes something as simple as sending a check for your taxes into an amazingly bureaucratic, slow process for electronic funds transfer. In the case of EFTPS, the issue is validating payment account information, which the feds do by making you want several weeks to get a paper confirmation with new access code -- an excessively inefficient approach compared to how banks and investment firms let people validate payment account information. It's easier to simply mail in a check, which means that EFTPS is a failure.
Obama has hired a national process officer to help streamline government processes. Taking that approach on federal technology initiatives is key to success. Perhaps Apple could lend some help, and companies such as Google and 3M that let employees use work time to explore personal projects could ask employees to use that time to help the feds for the national good.
And the federal government should consider replacing some of its legacy systems to get rid of the integration burden that they cause. The upfront cost is worth the long-term integration savings and transaction easing.
Research priorities The high-tech industry has been good about focusing research dollars on key problems. Apple, AT&T, Hewlett-Packard, IBM, Intel, and Microsoft, among others, are well-known examples. Government already knows this, and many agencies work with such corporations and various universities for the joint good of business and the public. Keep it up.
The Obama administration should beware the research priorities it hears from the venture capital community. The VC business is about exploiting tech-based business bandwagons as they start to move -- not about creating new technologies, as popular mythology has it -- and so VCs reward the likely winners after someone else has done the hard work. VCs follow, not lead, and when they appear to lead, it's typically in service of their own investments (example: the several efforts by VC Vinod Khosla to get voter approval of ballot initiatives that would essentially have the state fund his environmental tech startups' products).
The near-term research priorities where government funding is useful are fairly obvious: wide-area network design (especially wireless), sensor technology, analytics technology, materials reuse and reduction technology (especially for the often toxic materials used in high-tech products), and the tri of battery, alternative energy, and energy-efficiency technologies. The industry will take care of the core product-oriented technology areas such as storage, semiconductors, user interfaces, speech and voice recognition, network management, and software design.
Finally, the federal government should further invest in basic research programs such as DARPA that have kept U.S. computer science advanced for decades and led to the Internet and other now-common innovations. DARPA, the Defense Advanced Projects Research Agency, is obviously skewed to military research, so I urge the Obama administration to consider a civilian counterpart that would be as effective in nonmilitary research as DARPA has been in its sphere. The whole point of basic research is long term investigation, not for specific problems, so a tech agenda that focuses only on specific problems starves the resources for addressing the opportunities and problems we don't yet know about. So carve out a bigger space for the unknown.