Observation, experimentation, data analysis. These are key tools of behavioral science, the study of what people do, and the elements that influence human choices and actions. Behavioral science isn’t new; it’s so well established that a journal devoted to its history has been around for fifty years. The familiar fields of psychology and sociology are forms of behavioral science. But a new move by President Obama is bringing fresh attention to the subject.
An executive order directing executive offices and agencies to use behavioral science has generated headlines ranging from Harvard Business Review’s “Why the U.S. Government Is Embracing Behavioral Science” to Daily Caller’s “President Obama Orders Behavioral Experiments On American Public”.
Despite the headline, the executive order does not directly mention experiments. What is does mention is recruiting behavioral scientists into government, and strengthening relationships with the research community. And it mentions testing and evaluation of insights gleaned from behavioral science, which could imply experimentation, though perhaps not the sort that makes for sensational headlines.
Controlled collection and analysis of data in American government goes back at least as far as the first US census in 1790, and it has been somewhat controversial all along. But how would we run an effective government without data, or without analysis? When the FDA decides whether to allow a new medical treatment on the market, it requires experimental data and proper statistical analysis to guide decisions. When Congress prepares a budget, it needs to know the number of Americans, where they live, their income and numerous other facts about the American people to make a reasonable plan for allocation of funds, and when the FBI conducts investigations, it needs fact-based information about criminal behavior to be effective and efficient.
Businesses, too, can and do make use of behavioral science. You may even use it yourself. If your business takes advantage of survey research, A/B testing (often used to compare effectiveness of alternative ads, web pages, or direct marketing materials) or any sort of market research, you’re using behavioral science.
A report by Obama’s Social and Behavioral Sciences
It’s nothing new for the federal government to use behavioral science. The FBI’s famed Behavioral Research and Instruction Unit is more than 40 years old. Other federal agencies, including the National Science Foundation and National Institutes of Health, also emphasize behavioral science, and have done so for decades. So, this executive order is less a change in policy than a reinforcement of a longstanding history of behavioral science for government decision making.
Likewise, use of behavioral science in business is not new. The testing methods used by Obama for America didn’t originate in government. They are classic direct marketing methods, which go back a century or so. Other forms of behavioral science also share a long history in the business world. The first degree in industrial psychology was a Ph.D. earned by Lillian Moller Gilbreth in 1915. So, we’ve had a little time to get used to the concept.
In government and in business, decisions are often made without obtaining good data, without rigorous analysis, and even without secondary research to learn what relevant information has been unearthed by others. Without data, management decisions are guesses at best.
Social science research can be poorly done or misused, so it's important to give thought to the meaning of research results, and use them appropriately. One popular example of behavioral science in action is the prodding of consumers with messages comparing energy use with neighbors. In the UK, these messages prompted reductions in energy use. Yet the simplistic comparisons of total energy consumption can be misleading.
My own electric company now sends similar messages with each bill, but the comparisons leave a lot to be desired. According to the bills, the comparison group is made up of nearby occupied residences; but the electric company doesn't really know which units are occupied, only which bills are paid. It also doesn't know the size, number of occupants of units, let alone who is (or is not) home working during the day. My home and my household are not very big, and we use less electricity than the US household average, but to read the electric bill, you'd think we were huge energy hogs. What's the problem? The "most efficient neighbors" in the comparison are primarily people who own two or more homes and don't really live here. The "comparable" households are simply not comparable.
Yet, when you take the trouble to obtain good data, analyze it properly and use the information appropriately, the outcomes can be remarkably valuable.
You may have heard of the Obama for America campaign’s extensive use of email subject line testing. If you were a donor, you probably saw a few emails from Barack Obama with the simple subject “Hey”. The campaign team discovered through tests (also known as “experiments”) that “Hey” worked. Those emails were opened more often, and led to more contributions than alternative subjects. These tests were key elements of the campaign’s fundraising success. So it’s not hard to understand why Obama would want to integrate similar analysis into government decision making.
Obama uses behavioral science. Shouldn't you?
