jim sullivan and i believe that the income data were not capturing people's well-being quite a while ago, though he did not have what you might call the smoking gun, where we could look and see if we had benefits to people's income who did not report them, then you really can see that people at the very bottom had much higher income then it seemed. >> there is one other thing, the grant was -- block $16 billion in 1996, and it is still $16 billion now, roughly. for people like that on the left, that is just like the biggest crime of all because of the major social program has not expanded and has not continued to grow. i think that there is a desire to find a way, find an argument for spending more through the tanf program and that might have motivated some of them. >> next question. >> following up on this conversation, myers wrote a nice article about a year ago regarding the surveys and the response, having guided the income surveys but relying largely on the expenditure surveys. can you tell us why they think the expenditure surveys are so good since you told us that the upper en

jim sullivan and i believe that the income data were not capturing people's well-being quite a while ago, though he did not have what you might call the smoking gun, where we could look and see if when you add benefits to people's income who did not report them, then you really can see that people at the very bottom had much higher income then it seemed. >> there is one other thing, the $16 billion in 1996 and it is still roughly now and for people $16 billion like that on the left, that is just like the biggest crime of all because of the major social program has not expanded and has not continued to grow. i think that there is a desire to find a way, find an argument for spending more through the tanf program and that might have motivated some of them. >> next question. >> following up on this conversation, myers wrote a nice article about one year ago with response, having guided the income surveys but relying largely on the expenditure surveys. can you tell us why they think the surveys are so good that you told us the responses awful? >> that is a fair question. so when you look

jim sullivan and i believed that the income data were not capturing people's well-being quite a while ago, but we didn't have what you might call the smoking gun, where we could look and see if when you add the benefits to people's income who didn't report them, then you really can see that a big share of people at the very bottom had much higher income than it had first seemed. mr. doar: there's one other thing. the block grant, the $16 billion in 1996, and it's still $16 billion roughly now. and i actually like that. but for people on the left, that is just like the biggest crime at all because a major social program has not expanded, has not continued to grow. and i think that there's a desire to find a way, to find an argument for spending more through the tanf program, and that might have motivated some of the findings. ms. marshall: next question. yes, sir. >> following up on this conversation. professor meyers wrote a nice article about a year ago on what's wrong with surveys nowadays. it was a response having gutted the income surveys but relied largely on the expenditure surv

in work i have done with jim sullivan, we find strong evidence that low consumption is more closely associated with deprivation and low income. looking at consumption gives you a better idea of people's living standards. what i am going to show you is what we call well measured consumption, rent, utility, gasoline, percentage of the value of cars and rental of homes. rent and food at home, that they consume it home, are the ball and those are fairly easy to report. people know the rent. people know about what they spend each month on groceries bills. here is the pattern of spending by single mothers. we focus on single mothers since they are most affected by welfare reform. here i am reporting the spending by households adjusted for inflation and family size. this figure shows the changes since 1990, when welfare reform was just getting started at the state level. what you see is that consumption at various percentiles, the top two percentiles here are the fifth and 10th percentile. consumption at each of the percentiles went up steadily. at the bottom, it went up by 50%. leveled off around 20