generational and income lines, not only support social security, but also agree on specific changes for the future. and with that i'll turn it over to matt greenwald who will tell us more about the trade-off analysis part of the study. thank you. [ applause ] >> thanks elisa and good morning. my job is the statistical technique. let me just say by background, most of my work is in the private sector. and a lot of my clients have this fixation about offering products people will buy in a competitive world, that people will prefer over their competitors.

they don't care that much if they like the car, they want to like it better than what their competitors are offering. in 1978 a statistical techwas d pennsylvania often called trade off analysis. since 1978 it's been used more than numerous times because a lot of private sector companies don't want to build a product and see how it works. they want to do research first before they put the car into the factory. this trade-off analysis has the benefits of identifying the preferred package, including costs, as someone once told me if you're doing analysis, the one is the least attractive. and identifies the impacted of each feature on preference. so, if i can give you the example of a car. a car can have numerous features. it can have engines of different size, transmission -- transmissions of different quality, a backup camera or not, different types of sound system, leather seats or cloth seats. each feature has a cost. and there could be easily 300

combinations. which is the best? they want to know the best. but you can't ask an individual 300 questions and expect quality answers. and the genius of trade-off analysis, if i can use that word, is by properly aligning the choices, you can ask eight to 12 questions and estimate how people would respond to each of the 300 using basian math, basically. also understand the impact of each one. so, the objectives we had in working with nasi were to understand the preferred package of social security benefits where there are a couple hundred different possibilities. make sure that each of the costs was savings of each feature is

taking into account when determining a preference. of course, people want more -- higher benefits, but are they willing to pay for it? and if you're interested in closing the financing gap, which survey indicates people are, how do you go about doing it? also if you're looking at different features and trying to make the trade-off decisions, what is the value of each feature? some features have the benefit, quote/unquote, of closing the financing gap. others increase the financing gap. what is the reaction to each one? understanding that the goal is to get the best package. that's where we're interested, the best package at the best price. what do americans prefer? this is what we did. elisa reviewed it. i want to say it again. we tested 12 specific changes or

features. some cost more. some cost less. each respondent saw ten questions. each question they were given four choices. every choice had a cost in terms of what it did to the financing gap. three new options and one is to keep it the same as it is now where you don't impact the financing gap at all, the choices just continue on. and so built into each feature was the revenue impact of that feature. and what this did is it put costs into the equation. if you want a benefit increase, it had to be paid for. if there was a benefit reduction, it closed the financing gap. and allowed for us to understand -- not just if people like a feature or not, but which

overall package, which impact on the financing gap people preferred. this wasn't a wish list. this was what americans -- how americans prefer to solve the problem. and one thing we saw, if i can just get one -- is they do want to close the financing gap. and this was evident in the responses. that was evident in the package they preferred the most. so this is the methodology used very often in the private sector. we've used it tons of times. don't think it's been applied in the public policy, or i'm not aware at all, but i think it really gave us new and important insights. with that i turn over the clicker. [ applause ] >> we are hoping jason furman will join us. is jason here yet? on his way, okay. i guess we can have questions until jason arrives. over there. just state your name because c-span is -- there is. the question is, where is it? who has the mic?