Q&A: Social Security CIO On What's Next In Online Engagement
Social media, digital credentials, and behavior modeling are being employed to get more Americans to use the Web to get their retirement benefits, says IT chief Frank Baitman
InformationWeek: As you mentioned, there are lots of legacy systems here. I've heard that Social Security has more lines of Cobol than any other organization. How do you view things like mainframes and Cobol?
Baitman: That's probably true about Cobol. For the foreseeable future, though, it's not going away. It's remarkably efficient for certain things. We have business processes designed around it.
That said, we have business processes that are constrained because we're using Cobol and we have these legacy systems still in place. We have very limited dollars to invest, so you've got to invest where you're going to get the biggest bang. In some cases, that may mean replacing something, and in other cases, it may be running something new that's an add-on to an existing legacy system, which gets to one of my pet concerns, which is recruitment. We have lots of interesting, challenging jobs at Social Security for kids that are getting out of college or grad school right now in computer science. Social Security could be an incredible career step for them.
Baitman: They're not, so come here and learn something new. Cobol's not going away here or at Citibank. You can come here and do that and then do Java and Ruby as well. It will probably be an experience you couldn’t have gotten anywhere else.
InformationWeek: Clearing the benefits backlog has been a top priority for Social Security. What are you doing from an IT perspective to help with that?
Baitman: We're pleased that we've made some significant progress in reducing the backlog the last couple of years. That's been a combination of IT and human resources.
I'm keen on enabling people to do more business with us online, and to do that, we have to get authentication right. We have a project internally called ROME -- Recognition of Most Everyone. We're going to be using a combination of SSA data and we have a vendor on board who's going to share their data with us. Together, we've designed a process that allows NIST Level 3 credentials to be issued online.
The important thing is: it's optional. You can still go into an office, and you may not need a Level 3 credential to do what you're doing. That said, if we can get many Americans a Level 3 credential, they could do just about anything that they would need to do online. We could auto-populate applications. If you've already been authenticated, we could take your records and put them into the application. You could zip through the application and make sure the fields are populated with the right information. We hope to have the first version of ROME done mid-2011.
The other one at the top of the list is [health IT]. We're going to get a definition where doctors are finally going to be compensated for deploying HIT in their offices. We're going to begin to get information from practitioners and hospitals around the country as they begin to roll out HIT. We had a small amount of funding, at least 12 contracts, to give to caregivers so they could transfer health records to us. We're learning from those experiences. If we start getting structured data, instead of getting 100-page faxes, from hospitals, we're going to be able to steer someone's eyes to the information that they need to see and focus on those bits of information to quickly make a decision about someone's application for disability benefits.
InformationWeek: Are there potential business intelligence benefits in analyzing all that data at a larger scale in order to make Social Security more efficient and effective?
Baitman: There are international standards being developed right now, and as those standards are promulgated, hopefully clinicians and hospitals will begin to use those codes for actually marking up records. Using a functional spec, our understanding of that person's condition is greatly enhanced. It will have a huge impact on our ability to actually get to the information among all the extraneous stuff.
The timeframe is the issue. To get to the level of looking at the efficiencies of the system as a whole, we're talking at least 10 years. We don't know what adoption is going to happen once meaningful use is defined and doctors begin to adopt it. We don't want to get out in front of it, but we obviously want to influence the market because the way the information is ultimately shared helps us do our job more effectively.
InformationWeek: Social Security has a lot of data, much of which might be fascinating to people. What are your thoughts around open data and what you can release potentially?
Baitman: Under the open government initiative, we've released maybe 20 data sets. Some are getting more use than others. I'm a huge believer in open government and transparency. I think the American public should know what their government is doing on their behalf and how they're using their tax dollars. We made a real effort in putting our open government plan together to say, what is it that we have that you want to know. It's an investment that helps us to do our jobs more effectively.
One of the things that I think is really very interesting: We put an interactive map on our open government page of state furloughs. Social Security has agreements with every state and four territories to hire furloughed state employees -- we pay them -- to do the initial intake for disability determination. We call it Disability Determination Services, DDS. Some states have included their DDSs in those furloughs. It causes a bunch of negative things to happen, though. First, you have employees who have the opportunity to get paid by the federal government, and not state dollars, who are not getting those dollars. States are losing the tax revenue off that income by furloughing those employees as well. Perhaps most importantly, there are benefits that should be paid to disabled Americans that aren't being paid. So, we put an interactive map on our page that shows what your state is doing.
InformationWeek: Social Security has a new Office of Innovation and an Office of Vision and Strategy. What are some of the projects underway in those offices?
Baitman: The notices architecture is one of the things coming out of the Office of Vision and Strategy. We're doing our strategic planning right now. I'm co-chairing that with another deputy commissioner, Ron Raborg. That process is going to deliver a strategic plan in the spring of 2011. One of the things the SSA hasn’t done in the past is to actually have an architectural vision. The agency's strategic plan is going to give us that vision. Coming out of there, the Office of Vision and Strategy is going to build a blueprint, a technology roadmap for everything we want to do from voice over IP to servers to software.
Karen Palm leads the Office of Innovation. A project that exemplifies what they're doing over there is we're designing a mobile application. We may release our first mobile application in 2011. We're very keen about figuring out what information the public would benefit from. At this point, I don’t see people filing complex applications over mobile browsers, but there are other things that we could do in the mobile browser to enhance the agency's mission. The day the SSA puts out its database of baby names each year is the day our Web site gets the most hits every year. People love that database. We could embody that database in a mobile application so that expectant parents are sharing information, and sharing that with grandparents in such a way that we can let people know that you can actually file for benefits online, and we could even use baby names to drive people to our Web site.
Federal agencies must eliminate 800 data centers over the next five years. Find how they plan to do it in the new all-digital issue of InformationWeek Government. Download it now (registration required).
InformationWeek Tech Digest, Nov. 10, 2014Just 30% of respondents to our new survey say their companies are very or extremely effective at identifying critical data and analyzing it to make decisions, down from 42% in 2013. What gives?