BOOK DESCRIPTION
Improving public policies, creating the next generation of AI systems, reducing crime, making hospitals more efficient, addressing climate change, controlling pandemics, and reducing disruption in supply chains are all problems where big picture ideas from analytics science have had large-scale impact. What are those ideas? Who came up with them? Will insights from analytics science help solve even more daunting societal challenges? This book takes readers on an engaging tour of the evolution of analytics science and how it brought together ideas and tools from many different fields – AI, machine learning, data science, OR, optimization, statistics, economics, and more – to make the world a better place. Using these ideas and tools, big picture insights emerge from simplified settings that get at the essence of a problem, leading to superior approaches to complex societal issues. A fascinating read for anyone interested in how problems can be solved by leveraging analytics.
ABOUT THE AUTHOR
Dr. Soroush Saghafian (Wikipedia) is interested in using and developing operations research, management science, and operations management techniques that can have significant public benefits. He is the founder and director of the Public Impact Analytics Science Lab (PIAS-Lab) at Harvard, which is devoted to advancing and applying the science of analytics for solving societal problems that can have public impact. His current teaching focuses on Machine Learning and Big Data Analytics tools for solving societal problems. He has been collaborating with a variety of hospitals to improve their operational efficiency, patient flow, medical decision-making, and more broadly, healthcare delivery policies. He also serves as a core faculty member for the Harvard Center for Health Decision Science, a faculty affiliate for the Harvard Ph.D. Program in Health Policy, the Harvard Mossavar-Rahmani Center for Business and Government, the Harvard Data Science Initiative, the Harvard Belfer Center for Science and International Affairs, the Harvard Center for Public Leadership, is an associate faculty member at the Harvard Ariadne Labs (Health Systems Innovation), and a holds a collaboration appointment at Massachusetts General Hospital (MGH).
[Alessandra Seiter]
We are living in an age of unparalleled complexity. Climate change. Global health crises. Economic inequality. These are all what policymakers call “wicked problems,” because they're interdependent and have no single solution. Traditional approaches to problem solving tend to rely on intuition, historical precedent, or political considerations which aren't well equipped to handle the complexities that we face. At least, that's according to Harvard Kennedy School professor Soroush Saghafian, who argues that we need a fundamentally different approach to solving the world's most pressing problems.
In his new book, Professor Saghafian outlines his approach. It's called analytic science, and it's a systematic way of using data and analytical thinking to understand complex systems and develop effective solutions. On this episode of Behind the Book, we speak with Professor Saghafian about his latest book Insight Driven Problem Solving: Analytic Science to Improve the World.
Professor Saghafian defines analytics science as the discipline of using data to build models that help governments, businesses, and nonprofits make better decisions.
[Soroush Saghafian]
If you want to understand the world around you, you should make models and you should use them to solve problems. So a lot of analytics science is about making the right models that describe the situation that you're trying to solve or understand, and use that model—which are essentially a set of assumptions—to go from whatever evidence that is available to you to solutions.
[Seiter]
Take the 2008 financial crisis when it started unfolding. The full scope was overwhelming: failing banks, a collapsing housing market, rising unemployment. So policymakers approached the crisis analytically. They identified key drivers of the crisis, including how mortgage-backed securities distributed risk throughout the banking system. For each of these drivers, policymakers designed targeted responses like new regulatory frameworks to address systemic risks, liquidity programs to restore credit flow, and stress tests to assess bank stability through insight-driven problem solving.
Policymakers were able to get the global economy back on track by creating elegance out of complexity. This example represents what Professor Saghafian sees as the most valuable promise of analytic science: the ability to leverage the vast troves of information available to us into actionable insights for social good.
Professor Saghafian's own research focuses on using analytics science to improve various aspects of health care.
[Saghafian]
In my lab, we are using, for instance, AI techniques to find better solutions for treating cancer patients. How do you use the large amount of data that is out there in EHR systems of the hospital, for instance, to create AI tools that can recommend treatments for each specific patient?
[Seiter]
Analytics science is not value neutral. Professor Saghafian urges us to remember that while data can reveal patterns and relationships, humans must decide how to act on those insights and shape the kind of society we want to live in.
[Saghafian]
One of the issues that I discuss in the book is how do you prevent using data to go to wrong insights? Although during that path you are doing everything correctly, you still come up with wrong insights because data is not enough. I use this quote that says, “You're smarter than your data.” So don't think that just because you have data, you can just provide the solution. There are steps that need to be taken.
[Seiter]
Professor Saghafian reminds us of several ethical dimensions to keep in mind when pursuing analytics science. The first is the question of equity. If analytics science remains concentrated in wealthy organizations or countries, it could exacerbate existing inequities rather than addressing them.
The second is the question of privacy. The same tools that can optimize public services can also enable large-scale surveillance.
The third is the question of fairness. Analytic systems can amplify existing biases in historical data, leading to discriminatory outcomes.
[Saghafian]
In medicine, for instance, the idea has been that we treat an average patient. They say, oh, for an average patient that is female and is between 40 and 60, the best thing to do is this. But look, the average person doesn't exist. We have lots of data about each individual. We can completely personalize treatments for patients.
[Seiter]
And finally, there's the question of transparency. When important decisions are made using analytical models, how do we ensure that those affected can understand and challenge those decisions, if necessary?
To address these ethical implications, Professor Saghafian calls for a deployment of analytics science that considers human welfare and social good from the start. He emphasizes the need to challenge assumptions that data-driven decisions, or the data themselves, are fair or even objective, and he thinks it's critical that the tools and methods of analytics science are made available to as many people as possible, including through his new book.
The book is Insight-Driven Problem Solving: Analytics Science to Improve the World. It's written by Soroush Saghafian, an associate professor of public policy at Harvard Kennedy School. It's published by Cambridge University Press.
This has been Behind the Book, a production of Library and Research Services at Harvard Kennedy School. Find past and future episodes Behind the Book by subscribing to Harvard Kennedy School on YouTube, subscribing to our newsletter, and visiting our website.