Earlier this month, the Obama Administration published a report on Big Data: Seizing Opportunities, Preserving Values (see also the summary on the White House blog). Much of the focus of attention surrounding the report has been on the positive potential of big data and the concerns over privacy. However, the report also addressed another issue: the potential for discrimination. As The Economist (“Regulating big data: Rules for the new tools“) noted about the report:
But the most novel examination is the way it looks beyond privacy to consider how technology can discriminate in subtle ways. For example, some online retailers use “predictive pricing” algorithms that charge different prices to customers based on a myriad factors, such as where they live, or even whether they use a Mac or a PC. Though there may be innocuous reasons for the price discrimination, there are few safeguards to ensure that the technology does not perpetuate unfair approaches.
It is important to “examine how algorithmically-driven decisions might exacerbate existing socio-economic disparities beyond the pricing of goods and services, including in education and workforce settings,” it states. “The increasing use of algorithms to make eligibility decisions must be carefully monitored for potential discriminatory outcomes for disadvantaged groups, even absent discriminatory intent.”
This discrimination manifests in different ways. For instance, an app called Street Bump, released by the city of Boston, uses a smartphone’s sensors to identify whenever a car is jolted, relaying that information to the city’s road-maintenance team. But the system has an inherent bias: richer and younger people are more inclined to have a smartphone and download the app, so bad roads in less tech-savvy places might not get fixed so readily because there’s less reporting. The bias was accidental but its effects are real. In this case, however, the designers realised the potential shortcoming at the outset and corrected for it by giving more weight to bumps from less posh areas.
As the report states:
Just as neighborhoods can serve as a proxy for racial or ethnic identity, there are new worries that big data technologies could be used to “digitally redline” unwanted groups, either as customers, employees, tenants, or recipients of credit. A significant finding of this report is that big data could enable new forms of discrimination and predatory practices.
Interestingly, the report recommends using the technology to monitor the technology:
The same big data technologies that enable discrimination can also help groups enforce their rights. Applying correlative and data mining capabilities can identify and empirically confirm instances of discrimination and characterize the harms they caused. The federal government’s civil rights offices, together with the civil rights community, should employ the new and powerful tools of big data to ensure that our most vulnerable communities are treated fairly.
To build public awareness, the federal government’s consumer protection and technology agencies should convene public workshops and issue reports over the next year on the potential for discriminatory practices in light of these new technologies; differential pricing practices; and the use of proxy scoring to replicate regulated scoring practices in credit, employment, education, housing, and health care.
– – –
The dangers that the White House reports raise are not new – but deserve the renewed attention they are receiving. This attention is part of a new look at the issue of the “digital divide.” For example, last November, the Washington Post held a forum on the topic (see earlier posting).
Almost a decade and half ago, Athena Alliance held a conference on Inclusion in the Information Age: Reframing the Debate. That was followed by our report Extending the Information Revolution: A White Paper on Policies for Prosperity and Security. As we argued back then, the major issues of the digital age are not just access to technology (aka computers and broadband). The issues of the “digital divide” go deeper to access to and use of information. As we pointed out in the key points that came out of our 2000 conference:
Point one: Focus on the transformation, not the technology.
The issue of concern is the transformation to the Information Age. It is not simply a question of technological deployment. The end purpose is not to narrow some gap, but to ensure that everyone has access to the expanded opportunities. Our framework should be one of inclusion for all in the broader activities that make up society and the economy.
Point two: Review and coordinate efforts.
The problem has aspects of telecommunications policy, such as infrastructure and standards and elements of technology policy, such as research and development and technology deployment. But it also has aspects of policies on training and workforce development, education, economic development, housing and community development, human services and trade. Reaching our goal requires a coordinated approach — in the private, public and non-governmental sectors – that combines the various elements of providing opportunity and inclusion in the information age. To coordinate policy, the focus of governmental digital opportunity efforts should be the White House, not in any one department or agency.
It is also time to take a new look at some policy areas. For example, a comprehensive approach is needed toward all parts of managing the information commons: privacy, intellectual property rights, “right-to-know” policies and other related areas.
Point three: Work to ensure that everyone has access to the technological infrastructure.
Barriers to access to the infrastructure are many. Ways of overcoming those barriers are also varied, including public access facilities that can combine access with training and other activities, as well as home access. With respect to access in the home, we must return to the question of universal access. We also need to address the development of broadband capabilities – both at home and at work. Both home use and public access points are important. Multiple access public points are needed, such as existing public facilities, training centers, libraries, and after-school centers. For these facilities, sustainability is the key. But, it is not enough to simply provide access. We must work to weave information technology into the operations of community groups in a way that will both help individuals use the technology and will make those groups more efficient and effective in their core mission.
Some of the barriers to digital inclusion are physical: the usability of the technology. This is not, as commonly thought of, an issue only for those with disabilities. The problems of usability and human-machine interfaces affect all of us and research on ways to increase access for those with disabilities will pay off in increased usability for all.
Point four: Encourage and facilitate participation and involvement by all in the digital economy and information society.
To foster participation and involvement, the technology must meet people’s needs – not define those needs. Information technology can help people in their day-to-day lives if it is designed and structured in such a way that it helps answer their questions and solve their problems. Otherwise it becomes a barrier and a source of frustration. This is the danger of what some refer to as the “over-wired” world.
It is important to understand that individuals have different needs. A one-size-fits-all may help some – and increase their participation and involvement – but will block others. By focusing on “demand-pull,” rather than “technology-push,” we can better tailor the technology to meet individual needs.
Development of meaningful content, including more locally-based content, is one of the ways to increase the level of participation. E-government is one important form of meaningful content. But, we must also insure that those who are not on-line are not left behind. No services or information should be removed or dramatically cut back from traditional means of dissemination in favor of electronic dissemination until and unless all members of the community have access to that electronic means as easily as they have to the traditional means.
Point five: Focus economic development on the Information Economy, not the Internet Economy.
The information age will require a new approach to economic development. Key to the process is using and developing assets: financial, social, skill-based, and information assets. We must focus on building the local economy’s vitality and ability to compete in the age of globalization and help people make the switch to the new economy.
Our priorities should include:
• development of processes for identifying and assessing local assets,
• revitalizing programs for training the existing workforce,
• helping small and medium size enterprises make use of IT, and
• fostering entrepreneurship at all levels and in all sectors.
We must also develop and utilize new mechanisms for financing the transformation, including Individual Development Accounts and new ways of financing intangible assets.
Collaborative learning and sharing of information is also important in the larger process of economic development. There are a number of examples of information assets being applied within businesses and with local economies. We need to utilize new knowledge management techniques and old-fashioned communications techniques to collect, disseminate and better utilize that information.
Point six: We need a better understanding of what is going on.
We need to re-look at the data needed for economic development in the information economy. The problem of data extends beyond the scope of local economic data. We need both better data and expanded analysis of the socioeconomic aspects of the information technology. That research must be translated into policy relevant terms. For this reason, Congress should seriously consider re-establishing the Office of Technology Assessment (OTA).
Point seven: The decision making process must be open.
True inclusion and opportunity can only occur if the process of decision making is open and transparent. Information technology has a tremendous potential for opening and maintaining channels for general input and advocacy. However, decisions made about the technology can have the effect of closing off the process rather than opening it up. We must insure that all parties are at the table when decisions, including issues such as standard setting, are made.
Point eight: Innovate and experiment.
We are in a time of transformation and change. The speed of that change and the pace of economic activity will vary. Yet the change is real and will continue. In such a time, we must often invent new ways of coping with our problems and new policies for guiding our economy and society. Such experimentation will require great policy discipline, however. It requires a strong, unbiased means of evaluating programs and policies – and the political discipline to follow the guidance of that evaluation. We must also find means to ensure that the evaluations are timely for the fast moving policy arena. The goal in evaluation is not simply proving the effectiveness of an action – it is to facilitate learning. Learning is the hallmark of the Information Age. Our public policy process must embrace that concept as tightly as the rest of our economy and society already have.
– – –
The policies outlined in the White House report on Big Data are good steps forward. As we move ahead to confront the challenges of the era of “Big Data” let us keep in mind that last point on the need for policy experimentation and innovation. And use big data to monitor how we are doing.