commissioners. >> i'm sorry i thought miss -- wanted to say something. >> well i'm not necessarily going to tell the federal government what should they do but i will say you brought up the point of successful grant writers, i do think we do have a problem of a capacity at some institutions and capital in terms of being able to leverage the best grants and designs and so forth and so i think mabin vesting in institutional capacity to have stronger grant opportunities and more successful grant opportunities would be one way to think about where to spend additional funds and i do think even if we were going to redistributor -- between programming, i still think we need form of accountability that the money is being spent right and i think to dr. miner's point about not

offending the constitution, there is a way i think to be able to increase capacity of institutions with the lowest income students and still call for accountability. >> thank you. madam vice chair you'll be fold by commissioner narra sacky and heriot. >> thank you very much. and this question is to all of the panelists. as educate yorz and others have looked out and reviewed pathways to higher education for our poor, our first generation college, our underrepresented minority stumds, one of the fairly novel concepts that have been developed is that of the early college. and as i understand that program, it combines also and college. that by the time a student completes their high school requirements, they have also

completed two years of college. i was wondering if there is any data out there, and whether this is a trend that you see merit in or what do our statistics and our information tell us? >> well, what i would say is that in -- these are fairly new programs, not in all cases, but we haven't seen them as systematic programs. one of the chal efrmgs is that public education in our country belongs to the states. and a few places that i've lived i've had the pleasure of learning there were more school districts than countries which all have different calendars, different graduation requirements, different rules and regulations about how to account for courses. and i think it is challenging and i think theoretically and conceptually it is a wonderful idea. in two ways.

one is that students accumulate college credits which makes college for affordable but i think what is more important about that is that they actually understand themselves as clearly transitioning from high school to some post secondary institution so it is a way not normally -- formally but culturally but socially to get students in the mindset they are expected to transition from high school to some post secondary institution. so i think it is yerl by -- it is interesting i was in the state of florida just a few weeks ago and the legislature mandates they have four lab schools attached to the universities and one of them is fau, florida atlantic university which does earl yes college, i actually had an opportunity to meet a 17-year-old and a 19-year-old who were both on their way to graduate school

because they had accumulated so many credits during high school and in college at that time and we don't have models but systematic data of which models work best. >> is that something the department of education can understand how education is generally a state-run program but is there something that the state department of education can encourage folks to go and get additional information because you are right the kids are on a college campus in more -- more often than not and they begin to see themselves there. >> absolutely. it is one of the things that we expect to incentivize in some of our programs where it is appropriate. so we're very excited about the potential of early college. >> the national center for education statistics collects transcript data from high

schools and we're also beginning to collect data from middle schools as well because some of the kids are already involved in the programs. it is a new trend and it takes a while to sort of get this in the mold of data collection but we are on it and we understand it and even different models and types of program but it takes time to collect the data and get them into the pipeline. i should say though that one of the things that is going to facilitate this type of data collection, the digital approach to transcript data collection, currently what is done for most -- most schools and school districts is that we have to do it by hand which is labor intensive. the coining of the data is not standardized so there are some issues to work out but it will be available in the coming

years. >> i would add that i think the institute of education sciences started to fund a couple of researchers looking at the effect of dual enrollment. not necessarily early college high schools. but one of the things to note on the programs is what are we measuring. are we measuring students who would have gone to college any way and it is getting through the issue of selection bias and finding the benefit to students who may not have gone to college and i think that is one of the key things to disentangle out of this. but -- and forgive me for repeat this again but there are ways to begin to measure this and i think some of the state data bases like the one in texas would be able to give you some of the answers that you're looking at because we are seeing students from the rio grande valley from south texas, some of the poorest counties in the nation, graduating with associates degrees leaving high

school. we don't know what that means for long-term traj ktory but we do have evidence that completing the associates degree does leads to odds of completing a bachelor's degree. >> and do i have time for one other question? >> sure. >> as a former state trial and appellate judge i saw early on that indeed there was correlation between education and incarceration. in fact, it was often repeated that the number of students not reading at grade level by the third grade was one of the assessments that was used to project the number of prisons that were to be constructed and the number of prison beds that we would need as the states and the nation. can you comment on that? is there any truth dr. carr to

such a statistic being kept, and if you know, whether in fact, it is used for a projection as to the number of prisons and prison beds that we'll need? >> i can say that we certainly don't keep it: but i don't doubt that it doesn't exist or people aren't using it to make such projections. but i can say that the gaps between minority students and white students are large and persistent and they start early. and this is something that we really do need to take a look at. there are factors that are detrimental to the trajectory of the students and their academic pursuits. i think though we cannot lose sight that there has been significant progress and it is

not all doomsday. it looks bad. i realize. but the data suggests that all students regardless of race or ethnicity, are improving, but although the gaps are still there. and the only reason the gaps are pharaohing, even as mall as they are, is because the bottom of the distribution is coming up quicker and that means that minority of black and hispanic students are making significant improvements. >> dr. flores any comment? >> i would concur with peggy. i don't doubt that the statistic exists. it is not something that the department of education maintains. >> i would just add that there is evidence out of economics that shows that increased education, especially the completion of the high school degree reduces crime. >> thank you, commissioner nagasaki followed by commissioner heriot. >> thank you. dr. minor you made a comment

that there were kpleerly -- clearly many more students eligible to be served who are probably not being served because of the limitations on resources. do you have an estimate of about how many we are talking about? >> i think it depends on state. but in most programs, let me say it this way, we probably could double the number of students that are dg served by the program -- that are being served by the programs that are currently being funded. >> so some of the witnesses that are testified over the two days of hearings have proposals of either they feel that there is insufficient data to show that trio and the other programs have been sufficiently successful so we should just eliminate funding for that or some of them have been successful perhaps it would be better to roll it all

into one big general grant program that was more flexible. i'm wondering what your take on -- in terms of the data, how could we improve the data collected. dr. miner you noted the department has been doing more requirerous data base collection and i'm wondering what you've learned and i'm wondering if any of you have the response to the issue of how the programs could be improved? >> well, thank you. i appreciate you highlighting the point. there is no doubt about it that we need to have better evaluation and data attached to this kind of data annually. i make no bones about that. in terms of what to propose in place of or in stead of is a good question. because as durable as these programs have been i don't think there is consensus in the field about how to replace or do the

work better. i think the one thing we are clear about is that there are many factors that contribute to a young person being successful in the education system and so there is some need for a diversity of efforts. but one of the things that i've been very clear about and i think the dep is very clear about is increasing the rigor of the evaluations attached to the program. some of the programs were started 50 years ago and rigorous evaluation about the effectiveness was not a part of the legislative record at that time. but i think now, as we move forward, i think we do -- we are significantly more sophisticated in terms of the social science. we still have serious data problems to fix. but i can guarantee it is not just the department that the grantee communities and the constituents are very cooperative and willing to learn how to more effectively serve

students. i met with the group just two weeks ago and one of the things that i try and communicate to them these are not federally funded programs to build roads or to build bridges, these are young people. and i take seriously the issue that we could be spending taxpayer dollars in programs that don't effectively help students be successful in educational systems so it is something that we're very serious about and i expect that to become a much more significant factor going forward. >> anybody else? >> has congress been providing sufficient funding to do the kind of research that i everybody agrees would be i deal? >> the answer is no. what is interesting, when we raise this to the grantees dr. flores mentioned the kind of expertise and the kind of data

collection and capacity required to do the kind of evaluation has not sort of been banked into the budget and so one message there the grantees is we are hard to serve the students and now you want to lay on this exquisite elaborate evaluation without resources is problematic and so i think that is something that we have to take up if in fact we're going to ask individuals who have been awarded grants to do additional work, to be responsible for rigorous evaluation we have to be responsible about providing that kind of support. >> commissioner heriot? >> i think dr. carr was next. >> no it is okay. >> i actually have one more question. >> go ahead. >> so it has been my experience that the cost of attending college is not just the tuition and fees, the challenge it seems, and a lot of the reading

that we have, is that not surprisingly, if you come from a poor, low income family, you are trying to work full time, or a lot, and that contributes potentially to not being able to finish on time. and so i'm wondering how much research if any, has been done on the efficacy of providing a stipend so students -- so they can spend more time being able to study and take a full load than having to have the stress of working full time as well as trying to carry a full load? >> let me just say quickly i'm very proud of one of the programs that is run by the department of education, it the not a tree or a greer-up program but refer to it as childcare and access and it provides childcare access for students who have children. and so i think it is -- it is a

critically important factor. one of the things that i want to make clear and i don't know that this data point has come up in the day and a half that you all have heard testimony is that we often talk about college students as 18-year-olds who jest lust high school. when, in fact, that is not true. the mean age of students has gone up over the years. right now in this country there are more individuals between the ages of 25 and 64, individuals we expect to be in the workplace, that have some college but no degree. meaning that they started college somewhere and they fell out. there are 36 million individuals in that age group and only 33 million individuals in that age group who actually have a bachelors degrees. what that tells me is not only do we have to provide traditional opportunities for individuals to earn a post secondary or individual degree, we also have to provide less or

nontraditional ways for students who may have started tree years -- three years ago, who stopped out to work and raise children and have a family and do those kind of things and we have to provide degree opportunities and pathways for those individuals to return. >> i think i would add that the common student is no longer the 18-24-year-old without work responsibilities or family responsibilities. so this idea of a stipe epd would be a great experiment to implement. would it work. you still have to fill out the fapsa and figure out how to comply with federal regulations and at the end of the day for poor students they never get near filling out the application. so there is significant applications to fill out for a stipe end and so we come back to the simplification to make

yourself known as a student in need and the easier way out so to speak is to just pay as you go in commune colleges. so i think it is a great idea and a great experiment but it is going to require additional staff -- scaffolding. >> thank you. >> commissioner herry ot. >> thank you. i just wanted to go back to the point that the chairman started with and point out that it is a complicated world for all races and we talked about disaggregating data for asians and hispanics but this is going to make things look different for blacks and whites as well i believe. for instance my understanding that caribbean blacks tend to do better in the higher education setting than noncaribbean blacks and that among whites you get

big differences as well. some ethnic groups do better. jewish students have been extraordinarily successful in the higher education and scots-irish have been less successful in that setting and have not done nearly as well. this is not to say that these groups don't excel in other areas but in the area of higher education, there are big differences among sub groups within blacks and within whites. has anyone selected any data on that? is there any plan to collect data on that kind of issue? i guess this is for you dr. carr the most, but anybody else who would like to jump in there? >> well it is a very complex set of questions you start asking people those sorts of things about their religion, even sometimes their origin and the country of origin. so we have to be very careful.

we work closely with omb regarding how we can ask these questions and how we can report out on these questions just because they state -- the state collects the data, doesn't mean omb will support us collecting that data that way. but i think do there is a wealth of data, through other means not just through the data for collection of statistics reflecting that blacks, blacks from the african nation, for example, tend to score higher and the caribbean blacks as well. so there is a lot of information that tells us that we need to pay attention to the differentiations but we have to be careful about how we ask these questions. >> i appreciate your question. i think it is very important in terms of -- when the question to me makes me think about studies of immigrant students and

generational status and the census has many data sets where you can begin to segregate between among whites blacks and biogenerational data sets an i would be happy to refer you to those groups and the answer is not to desegregate and to spend the money it is important to know where the gaps are. and i'll stop there. >> and i would add sort of one technical problem with the desegregation sort of pathway and this is a statistical one and once you start desegregating at a certain level, you won't have enough size or power to detect patterns that are reliable and dependable over time. so in many instances you can't go down as far as you would like or to cross those sub groups

with gender for example. pacific islanders, there is a good case there are very few and they are located in certain states, only about five states to be specific. >> the thing that worries me is i think a lot of americans get the idea blacks and whites as a group are monolithic and neither are monolithic. they are very complicated groups and i -- i take your point on the difficulty of collecting the data and the sensitivity of the issue but it is -- it's important to me that people understand these are not monolithic groups. >> is it a quick kme, commissioner? >> i hope so. thank you to the panel and thanks, mr. chair. dr. carr what factors contribute to the determination of what -- first of all

socioeconomic status i think we have a general understanding that it has to do with primarily income or are there other factors that contribute to someone's socioeconomic status? >> there are three factors and in the literature that are typically used to determine socioeconomic status. holing head 1964 for example identifies income, parental education and occupation is the three key factors but having done research in that area myself i could say even within those key factors there is differentiation about what they mean based upon the cultural and racial makeup of the family. so income of -- $400,000 of income for a black family might mean something different for a white family or having a

four-year degree for a black family, a family with a parent with a four year degree may be something different from a family with a different sort of access to a different type of a four-year institution. so it varys and we have to be -- varies and we have to be very careful. and so the department has depended most notably on data from the free and reduced price lunches that i mentioned earlier but we're having problems now with the reliability of that data and clerkting those actual -- collecting those actual income from the parents and that is also a bit of a herring because parents don't want to tell you how much they make even when you give them ranges. >> so income, parental education and occupation. i noted from your graph, asians even from low s.e.s.

dramatically outperform, not just other groups from low s.e.s., but groups from high s.e.s. do you have any analysis or data on why low s.e.s. asians outperform just about everybody else. >> well the asians are not disproportionately located in the lower s.e.s. compared to blacks and hispanics. unless you separate the asia-pacific islanders out they are very poor, then you don't see the indications we see today. >> and a question for dr. minor, and you mentioned the gear-up and the campus program and do you have an understanding of how much the programs and total expenditures and all of the programs has been level has been flat and increased from

1990 to the present do you have any data related to that? >> yes. we have very specific data for all of the programs in terms of the appropriation levels from year to year. i would say over the last decade there have been very small incremental increases subject to the budget, but fairly flat compared to lots of other indicators. and the big question again is wlorntd the investment or the 2% increase or 39% increase, whether or not that is sufficient to see the movement we need to see across the country. but in the last just several years it has been relatively flat with small relatively incremental increases. >> and when do the four grams -- for example, when did the bulk of the programs have thein sip yhency. was it recently or can you take it back to 1970s, 1980s, 1990s? >> some of the programs that we spoke of earlier.

the suite of trios talent search e.o.c., they are about 50 years old. and they were part of the legislation -- the great society that sought to end poverty in the 1960s. some of them, gear up as we mentioned, came online in 1998. some of them, first in the world, as recently last year 2014 was the first year of the grant program. so a majority of them, there was a bundle that came online about 50 years ago. some mid to early 90s some are extensions of other programs and some of them are new. you heard me mention earlier the president's goal to be first in the world. that has been complemented by the establishment of a grant program to spur innovation into degree completion in post secondary education. so that program this year is only two years old.

>> i'm going to exercise chair's prerogative and wrap up. we are over time. and i did want to ask one quick thing before we close. dr. flores, you mentioned one thing and dr. miner concurred that starting an associates -- of course, in the community college, makes it less likely that you'll obtain your bachelor's degree, is that correct correct? >> students who start -- yeah. >> so yesterday, dr. flores, william flores, president of the university of houston downtown indicated that one of their success factors is that those students who enroll in a community college and transfer to their school they actually have them go back and complete the associates degree and then graduate -- go through a graduation ceremony and that increases the likelihood of completing the bachelor. i don't think that is necessarily inconsistent with

what you are saying but could you address that, if you are familiar with that latter issue? >> i think my light is off so i'll speak loudly. so the evidence i was speaking about didn't account for that potential innovations. >> turn yours off. >> okay. i don't think those are necessarily inconsistent stories, i think we're talking about an additional intervention. so the university of houston downtown study started this intervention of taking students back, right. the other studies i'm talking about didn't account for that intervention. they are not inconsistent. that could be an additional way, right. the students transfer and that says a lot about the student because most students never even transfer. >> mr. chairman, can you just answer this question. >> is that what you want to do? go ahead. >> yeah. so for the asia-american community, again the

demographics are shaped by how immigration has shaped the community here and the biggest predictor of poverty in the asia american community is the english proficiency and many asian languages aren't basin in latin so it is difficult to -- much more difficult to learn english. so you have a situation where a lot of parents for example from korea, and other countries, may be highly educated and may have college and advanced degrees but can't automatically turn their professional licenses here turn into their professional career and they end up owning grocery stores and low income work so they are high educated as parents which is a best predictor for their kids but their income is low. >> my understanding is that is

education is a factor. >> yes. but it needs to be a single construct. >> well i'm going to wrap this panel up. it is fascinating and we could talk up and we have another panel and we want to be respectful of that time. it is helpful. as i bid farewell. you can stick around for the balance of the day. i ask the other panelists to begin to move federal and our staff to change the name plants so we can get starts on our next panel. thank you. while congress is on break this week, we're showing you events from american history tv normally seen only on weekends. tonight on american history tv prime time visit historic sites in key west florida tulsa oklahoma, st. augustine florida, and topeka, kansas. join us at 8:00 p.m. eastern. on c-span tonight, the

interviews with freshman members of congress, starting at 8:00 p.m. eastern, it is elise stefanik of new york, talking about being a new member of congress. >> i pinch myself the first state of the union is a historic event to be sitting on the house floor when prime minister netanyahu delivered the joint session, i looked around and once again there was a moment i can't believe i'm here. in february i participated on a coddel to a congressional delegation to afghanistan, iraq, kuwait and jordan and for me i was able to visit with soldiers deployed to afghanistan but one of the visits we had was with president ghani, the newly elected president of afghanistan and with members of congress with the president and i pinched myself at that moment as well. so yes it is a very awe

inspiring experience but you understand that the reason why you are put in office is to represent the people your friends and neighbors from back at home. but when i -- when i stop pinching myself i think it is time to go. >> our interview with republican elise stefanik from new york is one of four professional profiles from tonight. and you'll see brad ashford and done beyer of virginia. the personal stories and reflections on being a member of the house of representatives tonight at 8:00 p.m. eastern on c-span. the c-span cities tour is partnering with our cable affiliates. join us this weekend as we learn about the history of omaha nebraska where the depores club was the first group fighting for racial equality.

>> omaha had a reputation in omaha and the united states as a city that when you came in, if you were black you needed to keep your head down and be aware that you weren't going to be served in restaurants and stay in hotels and when the depores club began their operation the idea of the -- and the term civil rights -- they used the term social justice because it wasn't part of the lexicon and the idea of civil rights was so far removes from the idea of the greater community of omaha or the united states that they were kind of operating in a vacuum. i always like to say they were operating without a net. there were not the support groups, there were not the prior experiences of other groups to challenge racial discrimination and segregation. >> we look back to the union pacific and how the construction of union station helped omaha's economy. >> the union pacific is one of the premier railroad companies

of america. it was founded in 1862 with the pacific railway act signed into law by lincoln and it combined railroad companies to make union pacific and then they were charged with building the transcontinental railroad to connect the east and west coast. so they started here and was moving west and central pacific started in -- on the west coast and moving east and they met up in utah. and that is really what propelling us even farther and we become that point of moving west, the gateways to the west. >> see all of the programs from aum uh-huh on saturday at noon eastern on c-span 2's book tv and on sunday on american history tv on c-span 3. did quantitative easing work. during the financial crisis the federal reserve purchased

trillions of dollars in bonds. the brookings institution hosted a discussion on monetary policy. financial experts and federal reserve staff were among those taking part in the 90-minute discussion. good morning. i'm david west hall director of the bookings institution here on the monetary policy. our mission is to improve fiscal and monetary policy and the public understanding of it. hopefully we'll not be judged by performance. but it seemed to us that one of the most vexing questions that coming up with monetary policy is did the feds very aggressive

asset buying of bonds known as quantity tasteive easing contribute to the easing of in equality and if so, how much. and this is a topic on which there is strong opinions and there have been some analysis and our in tent here is to build on that analysis and to think to the extent of to which it is true and if so, how did it work and if not, why is it so many people think that inequality did result from the feds quantitative easing. we have three papers to present this morning. all of them are posted on our website. three different cuts at the issue. and our plan this morning is to have each of the papers presented and have a discussion and take some questions for a few minutes after each paper and at the end we'll end with a panel discussion that includes the presenters and my colleague

david cone and eric hurst now at stanford. as i was reading the papers, some of which are more economic based and some are not. there are two important things to keep in mind. the first is if all that the federal reserve did was increase stock prices then there won't be much to talk about. obviously stocks are more wildly held by people at the top than in the middle and the bottom and if stock prices go up they benefit more. but as josh bivens points out in the paper he'll deliver in a moment, house prices are a big factor and nor the middle class house prices are a large part of the asset. to the extent the house prices go up it muddies the story of just looking at what happened to stock prices. but the other thing that is important to keep in mind is one that economists do very well and few other people do which is to keep in mind the counter

factual. what would have happened had the fed not done quantitative easing. compared to the hypothetical is really importantment because you can't assume that something would have happened or you have to examine the alternative and that is something we'll try to do here this morning. before i introduce josh, i want to point out that ben bernanke isn't with us today, traveling in asia but is interested and weighed in on his blog this morning and no surprise bottom line, he rejects the view that the feds monetary policy has hurt the poor and middle class relative to the rich and can you see that on his blog. we're joined this morning by web cameras from brookings and c-span and so this is live so don't say anything you don't want the rest of the world to hear. and i'll start by introducing josh bivens from the economic policy institute in washington. he'll look at the channels

through which monetary policy work and discuss the distributional consequences of each and do that important compared to what question, what would have happened if the fed had not done quantitative easing. and susan lund of the mckenzie institute which has done work on the subject will respond and the two of them will join me up here for a discussion and before we go to the next paper which i'll discuss when that time comes. so josh. >> good morning. first of all, thank you to the hutchings center for the invitation to participate in this conference and david wessel and lisa shiner and who talked to me about the paper and made it better than otherwise. i'm from the economic policy institute. if you don't know, epi has been for years has been as exercised

as anybody about the rise in the economy over the past generation. and besides being exercised by it, we think it has roots in policy decisions so when i was asked if i want to write something about the intersections, i was intrigued. my priors going in were definitely that successful macro economic stabilization is strongly aggressive, not just useful but provides bigger benefits to the lower and moderate income households. so the great recession should that be different. well we know the monetary policy since the great recession has been very different in how it is implemented and the tools being use sod this is worth looking into it. and if my lookeds anything to this debate and who knows, it might not, it is what david talked about in his introduction you want to specify a baseline on which you are judging the monetary policy and the asset purchases in general over the past six or

seven years. so the two brace lines, the first one, i'm going to compare it to a fiscal policy stimulus that yielded equivalent impacts on economic output and so that is one. and what is the distributional consequences of shifting from one to other. and the other is no change in macro economic stimulus at all. if the fed decided not to do the asset purchases in 2009 and nothing else happened and what is the kwonts kwens of that. the first baseline of fiscal policy stimulus that fails to yield limit and it is a blogger rick rowe he writes at a place called worthwhile canadian initiative and worth reading and he frames the question like this and what he thinks the right question is if we use fiscal policy rather than monetary

policy would that switch from one to another and have distributional policies and that is what i'm looking at. and there is no such thing as a generic fiscal stimulus. can you do a lot of different things. just think about tax cuts. the implications can be different. we have the 2001 and 2003 tax cuts which were justified some on supply side long run growth grounds but they were justified on chancy an aggregate grounds and not progressive but in 2008 the stimulus act signed by george w. bush and much more progressive and within tax cuts it is hard to disguise the implications. and transfers the way they are in the u.s. economy, they tend to focus the benefits on the bottom two fifths of the distribution and so increase in transfers are going to be strongly progressive. and then really important issue concerns the benefits of direct

government spending in investment increases. theoretical models and in the research of the policy effective of the economy stuck at the zero bound of the interest rates and government spending comes in strong and so we want to know what the distributional implications of that spending is. the congressional budget office suggests allocating it in two ways. allocate the direct spending on a per capita basis a lump sum spread across the population or allocate it to the distribution of income. those yield two different results. the second one by definition is neutral if you allocate it proportionally to income. the first way is pretty progressive, if it is a lump sum spending on a per capita basis. and then of course it is probably different depending on what kind of spending we're talking about. the defense spending are different than grants to provide bus service in urban areas or

school construction in poorer school districts. so the punch line of all of this is it is hard to think of a generic fiscal appointsy of which to judge since the federal reserve began and so i decided to try to find a particular fiscal policy intervention that is about equal to the estimated impact of the l saps over the last six or seven years on economic activity and it turns out there is a decent one the stimulus part of the 2010 fiscal deal that pushed back when the bush tax cuts on the top 2% would be rolled back. and you can see, i would just sort of look at the two columns on the right the gdp effect and the unemployment effect. basically the stimulus portions of the 2010 fiscal deal which were a 2% payroll tax cut on the employee side, expanded unemployment insurance and extension of the refundable tax results part of the american

recovery and reinvestment act and the same effect on gdp as the impact of the l sap and the asset purchases there is a lot of vary is there, it is not tightly estimates so rough orders of magnitude. the first question might be why didn't you compare it to the american recory and reinvestment act. that is the most policy of stimulus and the answer it was larger than the asset purchases. and i'm not really willing to say that the impact of the asset purchases are linear that you can scale them up and say we've done two of them and that would equal ara and i don't think it equals that and why not push back the high end tax bush cuts because they are really small. the full range of tax cuts if they expired in 2010 might have been big but that was never going to happen. every single politic in town

wanted to preserve everything except the 2% and below. so this is sort of the specific fiscal intervention i'm going to compare to the l sap. and the distribution of the 2010 fiscal deal, you can use basically the congressional budget office data on household income to get a sense of where that money went. so that first column the payroll tax cut, that tells you the share of total payroll taxes paid by different income groupings. so the bottom fifth of the income distribution pays 5.6% of payroll taxes. the sort of 96-99% pays 11% of all payroll taxes an then compare that to the column at the left that is the over all income and their share in over all income. and so payroll tax rut is pieldly progressive. the euc extension, the unemployment insurance extension, that column is measuring the share of nonsocial security cash transfers that go

to different parts of the income distribution and i'll say that is a pretty good proxy, probably for where the unemployment insurance extensions went and you can see those are strongly progressive and in the middle, say three fifths of the income distribution and the refundable tax credits are strongly important for the bottom two fifths and basically i allocated. and those are the portions of the income distribution that have negative federal income tax rates in the cbo data and that is why i'm allocating those. and so the combined impact of the fiscal provisions of the 2010 fiscal deal are pretty progressive, particularly the fundable credits an the payroll tax a little less. and now we can get to the large squail asset purchases did to equality. and it gets to asset income

generation is at the top. and that is true. this is just a measure of share of total capitol incomes claimed by various income percent isles and the top claims 30% of asset holding, by 2007 they were claiming 57% of income generated by assets and the bottom percent has seen their share shrink. and this is the root. and when i look at the measurement of the l saps and i think the distribution effects are smaller than are often characterized in the popular press, some people have even called them the fed doing a reverse robin hood. that strikes me as wrong and there are three reasons why. timing, the fact they are not that different from conventional monetary policy stimulus and then the impact of housing as

david talked about before. and on the timing the asset purchases will boost prices now but at some point in the future they will probably almost surely be unwound and that will put downwarddownward pressure on prices. you held it over a long period of time you'd see your fund go up and then go down in terms of the marginal impact of lsaps on than stocks incredibly concentrated. when you push up stock prices you're helping today's stock owners maybe at the extension of tomorrow's but tomorrow's stock owners are not poor people they're pretty rich people as well. on the second point, it's not that different what the lsaps are doing. conventional monetary policy pulls down short-term interest rates and then hope through arbitrage the long-term rates that matter for consumption and investment decision fall in sympathy, and that's how it's supposed to work. lsaps go past that intermediating link because the short-term rates are buried at

zero. the goal of both is to push down long-term rates. the long-term rates cannot go down if you're not pushing up asset prices at the same time. it's not that radically different a thing than conventional monetary policy. some of the estimates in this say, the impact of given decline in long-term rates as a result of the lsaps is less on stock prices than conventional monetary policy. i think that's an interesting thing to examine but that was replicated in a couple of different studies. part of it may be measuring effect of monetary policy at zero lower bound you're measuring its impact in a pretty fragile and weak economy. but it's interesting to thing about. last one that's important is, housing is an asset price -- sorry, housing is an asset and its price can be affected by the lsaps and it's an asset held that's very important in the portfolio of the middle clasp that thing i've highlighted that's the broad middle class, middle 60% of the wealth

distribution. housing accounts for 63% of their total wealth hoedings. so the impact on home prices stemming from the lsaps is comparable at all to the estimated impact on stock prices, then you're going to see, you know a neutral effect here. and in fact what i found is most of the estimates or the implied estimates of the effect of lsaps on housing prices are probably greater than its impact on stock prices. what i found on stock prices, may be 5% to 10% increase in equity prices stemming from lsaps. the broad distribution of housing wealth is another thing that makes these impacts smaller than you might think. then i think the really important thing, that's sort of fiscal versus monetary. but we think about the real world. is it really true that if the fed decided to not do much in the way of asset purchases fiscal policymakers say everybody, jump into the gap, fill that hole in aggregate demand. this is what actually happened to government spending over this

recession and recovery. that line on the bottom 2009 is what's happened since the trough of the great recession. you can see the red line, it slopes steeply up before the frof and that's the american recovery and reinvestment act that we did do real fiscal stimulus to fight the recession. starting late 2010, early 2011, fiscal policy's been an outright drag on growth and that's historically astone, stonstonishastonishing. look at real government spending in this recovery versus othered, fiscal policy has been a very large drag on growth. if it is true output stabilization is progressive at all, then this baseline becomes really important. and if it's true the lsaps manage to keep unemployment lower than it would have been, it's going to have strongly progressive effects. we find that low unemployment and that sort of macro economic

stabilization is progressive. these are coefficients on a phillips wage regression which is the chang in nominal wages on the left-hand side, lag consumer prices, lag productivity growth some time period dummies and the unemployment rate on the right-hand side, and you can see that -- we've got wages at 10th, 50s, 95 percentile we've done it separately for men and women, the wages of the 10th and 50th are -- the 95th percentile male wage, it's not significant, it's transparent. so there's a strong equalizing effect of output stabilization and low unemployment rates. that has to factor really largely. if you think the lsaps help stablize the economy at all, this channel says they were progressive. just to end, what this says is that since the great recession,

as the fed has been pursuing. it's clear that's a progressive intervention. i think when trying to choose fiscal monetary policy i'd prefer fiscal policy carry a heavier load in stabilization but not because of distribution but bought is reliable and effective and distributional outcomes are second order. i think we should -- this fact that lower unemployment rates have really strong distributional consequences, it it does mean we can look at history of federal reserve actions and what is going to happen observe the next couple of years. that should affect us strongly. what this says to me if we go into contraction area monetary policy in the coming years that will have reg implications but what happened over the six or seven years it's been strongly progressive. and i'm out of time. thank you.

>> thank you, david for inviting me today. good to see all of you here. first, i want to congratulate josh on what is very accessible and interesting paper on the impact of quantitative easing and lsaps in particular inequality. he takes us through a lot of the theory and empirical evidence in a very accessible way and presents two very interesting counterfactuals, which is what would have happened if we had fiscal policy and what would have happened had the fed done nothing. so overall, i find his arguments and argumentation in the paper very interesting helpful for anyone to read. so i would congratulate josh. i want to expand on his overall contention that in fact quantitative easing and the unconventional monetary policies of last several years have not significantly worsened in equality in either terms of

wealth or income inequality. i want to begin with the caveat, no one's disputing the fact we've seen increases in both income and wealth inequality. in the united states and other countries not only since 2007 but predating that as a long-term trend and it's a concerning trend to an economist in terms of aggregate consumption, particularly for an economy like the united states, 70% of gdp growth comes from consumer spending. so the statement wages and incomes of a large segment of the population is indeed a concern. that said, i would agree very much with josh, there's very little evidence that the federal reserve's monetary policies have contributed significantly to this. there's been three reasons why this is so. first, there's been most of the attention has been paid to the impact of low interest rates and qe on asset prices. we did some work in a report a year and a half ago looking at this question and found, in

fact, for equity prices, there's very little evidence that ultralow interest rates have been responsible for the run-up in the u.s. stock market. first point is that many people tend to forget although i'm not one of them looking at my personal financial statements, that at the end of 2008 u.s. equities lost 35% of their value. global equities lost 37%. so the increases we've seen in the stock market, first of all, let's put it in context, we're climbing out of a very deep hole. secondly, you need to look at theory of through what mechanism should low interest rates boost equity prices. well the most direct mechanism would be through a dividend pricing model. we discount future cash flows and the lower the interest rate the lower the discount rate. that should boost any valuation of the net present value of future corporate income streams so you think all right there's a very direct mechanism. however, both empirically and

these rhettically there's problems with the argumentation. first, if you're a rational expectations investor realize that quantitative easing and ultralow interest rates are a temporary policy. so you should not be val ewing corporate prospects ten years out based on whatever the current short-term or long-term rates are today. secondly sophisticated investors should be using an overall cost of equity, not only the risk-free rate the interest rate, but an equity market premium. and we've done modeling at mckin zi estimating what is the cost of equity. we find it is a very stable figure over many decades ranging between 7% and 8% for u.s. equities for instance. and this holds true since the recession. so, even if investors believe the risk rate has gone down, apparently increasing the equity market risk premium with

expectations rates increase. there's no real evidence there. the other argumentation is that, well, there's a substitution effect. you can't earn anything in bond funds so we switch to equity funds. look at behavior, there's little evidence either retail investors or certainly not money managers see equities as an alternative asset class to bond fixed income funds. we conclude that qe has boosted u.s. equity prices by maybe 5% at most, compared to where they would have been. so then the question is why have they increased? well, corporate profits are at all-time historic highs companies are sitting on over a trillion dollars of cash. there are good fundamental reasons to believe equity prices have gone up when you look at one-year forward looking price earnings ratio, u.s. equities stay or valued slightly over

what a long lfr-term average would have put them at. they may have gotten some slight booth to the value of the wealth holding. the more important impact of ultralow interest rates has been on housing prices. it's a direct channel because people buy housing with mortgages and so as the cost of mortgages goes down, it should have supported housing prices, which again, let's be clear, they have fallen dramatically fell 30% or more across the u.s. on average. but the evidence is that they would have fallen even more and been slower to recover without low interest rates. it's a mortgage direct effect in economies where mortgages are variable rate, like the united kingdom, the bank of england has said that low interest rates may have supported uk housing prices by 15% to 20%. and the u.s. is somewhat less direct effect because most people have a fixed rate mortgage.

you needed a good credit score to refinance to take advantage of loetthe low interest rates. but as josh's paper points out, we believe housing prices have been supported, that would affect the broad middle class and not the wealthy where for the top 10% of the income distribution housing represents only 10% of their wealth portfolio compared to the broad middle class where it's the mane financial asset. now the more direct impact of quantitative easing on household financial well-being that we look at in our report is on interest earnings and interest payments. so here it's a very direct effect. we don't need to go to theory. on one hand anyone with debt lass a lower debt service ratio. u.s. household overall today's debt service ratio includes cost of interest payments and principal payments is at a level last sense in 1990s, 20 years ago. this reflects both reduction in mortgage debt through the process of deleveraging, but

also the very low interest rate. so that's a huge help for households with debt. offsetting that, of course, anyone with deposits in banks and certificates of deposits or in fixed income funds have seen very low returns on those types of savings. so, we look at a figure called the net interest earnings of american households, which nets out the lower debt service payments from the lower income earned on cash deposits and fixed income. and what you see is that there is a clear effect across the age distribution. so basically households headed by people under age 54 have benefited because they have more debt and it's almost a linear effect households under age 35 on average have benefited by 1500 per year, so almost 3% of their disposable income. the group, 35 to 44 benefiting by 1700 per annum, 2% of their income.

to the other end of the age spectrum households over 75 who tend not to have debt are losing 2,700 per year through lower interest earnings. households between 65 and 75 are losing $1,900 per year, 2% of their income. there has been a generational impact for sure. again, overall, though hard to see how this worsens inequality. i want to end with an open-ended question meant to spark discussion. in the research we did we looked at impact of ultralow interest rates compared to what they had been. but in my mind, there's a very important question, to what extent are ultralow interest rates even the product of monetary policy? when you look back over history, since early 80s, you've seen nominal and real interest rates fall dramatically. when you look across advanced economies the real ex-post rate on ten-year government bonds average 8% in 1985 it's down to

less than 1% today. the nominal rate on ten-year bonds fell from 14% in 1985 to 2% on average today. so, what explains this? as an economist we learn that interest rates are the product of this supply and demand for funds. and ben bernanke, who is not with us, pointed out long before the recession that on one hand, we have the rise of surplus savings in many economies of the world, returning large sports surpluses and commodity exporters but the same time we've seen a dearth of demand. the u.s. and other advanced economies since the u.s. '80s you see a secular decline in the amount invested in physical capital. part of this reflects a shift to knowledge economy and intangible capital. part of it reflects the fact that capital goods cost less they've had tremendous economies of scale and cost has come down

but overall, we're living in a world in which companies don't see investment opportunities. you read about share buy-backs every other day in the press. many companies have now decided to start to return some of that cash to shareholders because they don't see where productive investment is coming from. so if we are living in a world of secular stag nation and one in which there are too few productive investment opportunities, then i think there's a real question about even with large-scale asset purchases have ended on net when we're going to see interest rates rise. and the phenomenon of very low interest rates may well persist, you know, for years to come and maybe outside of the control of the federal reserve or other central banks. and i'm out of time.

. >> that's what happens when you don't read directions. thank you, both for clear and succinct statements. we're going to -- i'm going to keep this session short with questions so we can get to the rest of the papers. but, josh, let me ask you so there are two things that just intuitively in my gut i wonder about. one is it seems to me there's a bit of attention on one hand, you say all of this monetary policy did a great job at restoring getting us closer to full employment so that's a plus. and then you say, but it didn't really have much impact on the stock market so it doesn't have as bad effects on distribution as some people think. well, aren't those two ideas that at odds with each other? if the quantitative easing packs such a big punch that we think

it helped get the economy stronger, how can we say it had such a small effect on stock prices? >> yeah. i wouldn't say it impacted -- packed thatting about a punch. i think the direction of the lsaps on economic activity and unemployment were in the right direction, it boosted economic activity lowered unemployment. i don't know it was an enormous punch like the slide i had up there, i think it said 2% of gdp peak effect which happened some time in the past year or so. i guess my -- so that, to me is consistent without enormous impact on the stock market. i guess my bigger point is that any positive impact in lower unemployment is strong progressive distributional consequences even if the fed's actions didn't have enormous impacts on unemployment, any impact on unemployment is strong and progressive that channel works well if you're concerned about equity. >> susan can you talk about the impact of very low interest

rates and government bond purchases on governments? you wrote about this in the mckenzie report. that goes to the hardance counterfactual, did governments benefit from this, and shouldn't have that made -- shouldn't that have made it easier to do more aggressive poll circumstance though they didn't. >> yes. thank you for remembering, one conclusion governments are clear, direct beneficiaries. when you look at net interest payments, it's been much cheaper for governments to borrow. that's particularly true for the united states where the start of the recession, average matureity on government bonds in the neighborhood of 4, 4 1/2 years and they've extended maturities but it's a direct savings to government whose can raise money more cheaply than they had. i would never suggest the federal reserve made monetary policy decisions with that as even a remote consideration. but it did enable the u.s. government to have a healthier fiscal picture than it would

have otherwise. we can debate whether that should have been used on more aggressive fiscal stimulus early in the recession, or even now. >> do you agree with josh's view that -- josh made the point there's no reason the federal reserve could have expected had it done less, that congress would have done more? do you agree with that? it's a political question. >> i will say i'm not a political scientist. but i am an economist. empirically we need look back at last seven years to see that there has been gridlock in washington. >> you don't need to be an economist to see that. >> i think we can take a couple of questions. brendan has a mike, if there are any. andy levin here. no, no, don't. we don't want to miss anything. >> i really like your paper josh. it's a really important question. i just checked on a couple of

websites medium real household income has dropped substantially, still since 2008. so i mean, to me it's hard to believe that that's part of a structural trend. so what you said at the beginning, that there are problems with intentional policy design, i think is the right answer here. and as you know, unemployment isn't the full measure of what's been happening in the labor markets. i think we should also be cautious how not exaggerating how much improvement there's been. so two points first you mentioned, well, the stock market goes up, but then later when they kind of withdraw the lsaps it it goes down. i'm not economically satisfied with that argument partly because i think that as susan said, the stock market dropped a lot in the aftermath of the recession. so part of the recovery that we see is a recovery back to something closer to a normal

level. it's not necessarily the case that ending the lsaps or normalizing the balance sheets can cause a fall in the stock market but we can think of this is a policy that brought the stock market back up closer to its normal level quickly. and that was a policy decision we could have made policy decisions to bring employment back quickly to its normal level and that didn't really happen. okay. the other point is that, you're comparing to a fiscal alternative. for the reasons we've talked about, including gridlock, i don't think that's the right comparison. let me suggest two alternatives. one is forward guidance. that's the other tool that the fomc has used and other central banks have used. it's an important tool. very nice empirical work examining forward guidance and lsaps, there are important differences how they work. i think there's a strong argument that, maybe don can weigh in over lunch about the extent to which the fed didn't use forward guidance

aggressively at all until around 2013 right. from 2009, 2010 2011, you know, kind of this extended period language which basically said we're only a few months or nine months away from lift up. in the same way that critical of fiscal policy you can ask questions what would be counterfactual where the fed would have introduced aggressive forward guidance and look at distributional issues there. the other -- >> i want to ask, you think there's a different distributional different between forward guidance and asset purchasing? >> that's what mike's results suggests, premium versus expected future short rates. we could talk about that during a coffee break or something. but the other alternative not considered in the united states but has been considered elsewhere is to have more direct credit programs to try to provide credit to entrepreneurs

you know, and small businesses. and they're critical of distributional issues here of the extent to which the easing monetary policy helps large corporations reduce their debt obligations and has not helped this -- we know the business start-up rate has not improved since 2008. this is something to compare as a clean comparison give continue has to be monetary policy we're not going to count on the fiscal authorities jumping in. there's a couple of other comparisons and what are the lesson of the future. >> to be clear in no way my contention monetary policy satisfactorily filled the shortfall demands since the great recession. to me the most reliable way to have done that with aggressive fiscal policy and we didn't do that and i'm critical of fiscal policy makers. someone fumbled the ball in a critical way, it's them. how can we be sure they wouldn't

have acted better if the fed had not been aggressive, to my mind the fed chair's communicated really strongly every time they spoke to congress fiscal policy is a drag on recovery, and yet stil there's no response from congress. counterfactuals and other ways to do monetary policy that could have been better or amplified, i totally agree with that. the recent payment by simon wren lewis and mark blath figure out a way for the fed to provide $500 to everybody's checking account. they can't do it now for sure but that would be a great thing for future recessions to think about. housing policies a big part of this. we missed a key transmission wrecknism, so many whoulcould have refinanced couldn't. if we could have done some program to allow those people into the refinancing channel. >> we'll get do that in a

second. >> this is not to say the fed did everything right. what they did do pushed in the right direction from a distributional angle. >> i would make two points. i think that the academic literature on forward guidance and its effectivenesses a little bit more subject to debate than you would suggest. at the end of the day i think we all believed that should economic conditions change suddenly for the better or for the worst, the federal reserve, and in any other central bank in the world, would act quickly, no matter what they said previously. but the second point on funding for small businesses, europe has been the most aggressive with directed lending towards smes and trying to promote sme securitization to get credit into small businesses. i know this only because we're doing new resent on this. but the results are quite astonishing. if you take eu as a whole and the u.s., they're roughly the same size economies in terms of gdp.

u.s. smes outstanding, according to a study done out of london has $3 trillion of outstanding credit and european ones, $1 trillion. despite the attempt to get credit flowing to smes in europe that has not been a successful policy. whatever the fed has done right or not done right, funding for small businesses in the u.s. is dramatic. >> what andy's saying anytime you want to feel good about what happened in the united states, use europe as a benchmark right. >> good point. >> the question is what could we have done that would have made the outcome different? one more question. we'll get more to this. bob, and then we'll move to the second paper. we'll come back to these issues as the day goes on. >> bob. bob samuelsson "the washington post." last week, byron wien a well-known stock strategist, now

works for blackstone put out a commentary in which he recounted the following arithmetic. since the low point in 2009 stock market valuation has increased $13 trillion of 0 that he attributes $3 trillion, almost a quarter, to the federal reserve's qe policies. that's a major part of the increase increase. by contrast, your appraisals, both of your appraisals is that the qe had, at most, a very modest effect on the increase of stock prices. could you explain why you think that -- i take wien's arithmetic to be typical of wall street, an obsession with what the fed is doing and has done and will do -- could you try to explain why there's a large gap between

people who are in the markets and people who are watching people in the markets and trying to decide what they're doing and why. >> wall street or economists, right? >> definitely go with the economists. but the stock prices have risen but they fell 35% okay. there's one factor. second factor, look in our analysis of qe boosting stock prices by no more than 5% we did not take into this broader based economic growth effect. so that if the recession had been dramatically worse and corporate earnings would have therefore been dramatically worse, you know, how bad could it have been? that counterfactual is something we didn't measure. but for the reasons i explained, i think that corporate profitability is up low rates have helped companies very slightly through lower debt service payments like they have helped households. you know overall, we're living in a world where companies have done very very well

particularly in the u.s., and that explains the outperformance. but i would -- lie take a look at those calculations and see where we come out so dramatically different. >> yeah. i haven't looked the detailed calculations either, i haven't heard of this. i would say i occasionally see traders overestimating, in my view, the impact of sort of federal reserve decisions. my guess it's because they live in such a short-term world. you look at some of 0 the empirical research on lsaps event study they look at what happened when it's announced and if you stop there you might get a dramatic effect. but then the price goes up when it's announces and it decays over time. getting impact of the lsaps over the entire period of time rather than that day that i remember when they announced it and stock prices jumped through the roof, that's part of the difference between the trader view of the world, what happened day to day, versus what has happened over the entire time span.

>> but let's say byron is right, after all, people who bet on the stock market tend to make more money at it than people who don't bet on the stock market. so if he was right, and if the effect on stock prices was a lot larger than in your -- as you discussed in the literature, given all of the other things, including house prices, would that have made a big difference on the distribution question? >> yeah. the real horse race is between what it does to stock prices, which are concentrated at top versus housing prices which are broadly distributed. if i'm wrong and there's a larger impact on the stock prices, absolutely a lot of my conclusions would need to be modified. >> i'll introduce the second panel. the second paper, which follows nicely after this, is by martin, erik hurst joseph vavra and andreas fuster. what they look at is, okay it's

great to say this was good for housing prices but if you happened to have the mismore tune of living in a community where housing prices were under water, it it wasn't going to lift your -- you were going have this luxury of refinancing and it might not have helped. their paper looks at inequality but regionally rather than across classes. why don't we sit down and then come back? >> our take is different than papers. we're not talking as much about people as we are about space. we want to ask the question, where does monetary policy have more of a bite? the places that were doing relatively bad or in the places that were doing rel tichbly good? in terms of the recent economic performance within the u.s. or within europe, we know there's a large amount of heterogenaity,

las vegas, dallas. unemployment rate in las vegas got up to you know, 13% or so dallas peaked about 7.5%, 8%. sent narrow's going on in europe spain is hit harder than germany. we think of monetary policy we think of it as a blunt tool. interest rate, some policy tool moving and that kind of because of talking about monetary unions, by definition that interest rate is constant across the member union -- the member states of the union. so, what we want to think about, though, is the spacial component of monetary policy also has some sort -- is the result of variation in collateral val use across space the transmission of monetary policy to real

activity. you heard how monetary policy affects real activity it boosts stock prices, boosts home prices making firms borrow more expanding investment which might also cause them to hire more workers. to the extent some of the transmission from a given monetary policy action to real activity is dependent upon local collateral values. to the extent they differ across space, monetary policy could have differential impact across space. and i'm going to show you this through some data that the places where collateral values are most oppressed sometimes see the least amount of response in terms of real activity to a given type of monetary policy. so, what i'm going to do what we do in the paper, two things. i'm going to spend almost all of

my time today on the first part which is data. i'm going to show you a bunch of data around the first quantitative easing not because i believe quantitative easing is different than other types of monetary policy but because we could date it exactly. i'm going to show you the response across different types of regions within the u.s. to some type of collateralized lending. in this case i'm going to focus exclusively on refinancing. our point is this applies to broader types of lending and much broader types of monetary policy but i'm going to use as a case study quantitative easing right around quantitative easing in the refinancing channel. and i'm going to show you a bunch of data how refinancing was much stronger in parts of the economy, parts of the thaus were doing relatively stronger and i'm going to show you that's going to translate into some aspects of spending local real activity. the second part of the paper, i'm not going to spend much time on today, we twriry to write down

a quantitative model to put in the collateralized lending channel, where there's two types of shock in the model. one a shock to call it local income, local productivity, something that makes las vegas doing worse than dallas. and then there's also another type of shock collateral values. they might be causatively core correlated. they might be ashrbitrarily uncorp lated. it's going to determine the distributional effects of monetary policy and also thing eing e agragregate effect. when we lower interest rates and the places in the country that are wanting to borrow the most, having the most demand for credit aren't actually able to get that credit because collateral values are low it can make monetary policy less effective in the aggregate than

it would have otherwise. correlation between collateral shocks and real productivity shocks or some other local driver of local income is going to be a key correlation that determines distributional and the aggregate effects of monetary policy. i'm not going to spend much time talking about that today but that's the key insight from the model we write down. from the empirical part we use a lot of data. i'm going to try measure as a case study kwan easequantitative easing the first one, and look at refinancing behavior and use data from the home mortgage disclosure act which measures mortgage origination and mortgage application activity in real-time. using the microdata that is able to be accessed from within inside the federal reserve that can date this by day. so you can get very fine type of temporal response in terms of when the monetary policy -- when

the mortgage activity takes place. we're also going to use data from equifax that has their credit risk insight servicing data set which merges credit records into applications with mortgage origination activity. so then we can actually see large amounts of information about the borrowers and their loan-to-value ratios at the time of refinancing. and a bunch of other data sets as well to merge in but those are our main two. the first picture you should see here is the solid line is refinancing activity in the aggregate measured by the mortgage brokers association. the dash line is just the 30-year mortgage rate rolling average over time changes. you can see right the at the moment, every picture i show you the rest of the day the red line is going to be announcement of

the first quantitative easing program in november of 2008. you can see, at that moment, mortgage rates fell, and refinancing activity increased in the u. at that moment. the second picture shows you the dark shaded line, how much of that mortgage activity was due to refinancing and how much was do to new originations. almost all of that activity that responded after the first quantitative easing was in mortgage refinancing. there wasn't a bunch new home ownership buying at that time. it's just a refinancing of existing mortgages. that's what we're going to focus on refinancing in the paper. in the back of your mind we should note what defines space as we've done -- been talking about. it's the value of collateral, these locations are going to have at the moment that quantitative easing takes place. so what i'm showing you here is

the loan-to-value distribution of borrowers in five different u.s. cities at the moment quantitative -- the first quantitative easing took place. you can see that bottom line is las vegas, and that red line in this picture is how many borrowers have at least 80% equity in their home at the time of quantitative easing. and at the moment that takes place in las vegas november 2008, only 20% of borrowers have at least 80% of equity within their house at that time. where the comparable number in philadelphia is like 70%. so you see the type amount of borrowers at risk to be able to refinance difficults are dramatically across different types of cities at the inception of this first monetary action or monetary action that we're focusing on in this paper. the second thing i want to show you, not surprisingly, this is

well-documented, the correlation between how many borrowers are under water on the vertical act axis or at least have loan to value ratios above .8 and the change in the unemployment rate up to 2008 november is highly correlated. the places that were hit hardest in terms of real activity measured by unemployment rate going up the places where most borrowers were under water. there's a strong correlation between real activity and house price declines making borrowers under water in the different cities. here's kind of pictures i'm going to show you a few like that in my remaining time. let me set them up for you. what i've done is clumped all of the msas for which we have data. think about 300 msas in the u.s. population weighted into core tiles based upon how many

borrowers are under water measured by loan to value ratio of .08 november 2008. bottom line is going to include msas like las vegas where there are a lot of borrowers under water. i'm going to refer to those as high loan-to-value ratio places. top line is lowest core tile of borrowers, these include philadelphia and seattles these types of metric. and at the moment that quantitative easing takes place the first one, you see a spike in refinancing activity. trending very similarly before. you see a spike in refinancing activity that is much larger in the parts of the country doing relatively well. these are going to be these low loan to value-type places. this an application data moment using high resolution data. you can see applications spike immediately after the

announcement. differentially in parts of the country doing relatively well. this is actually on origination data using our equifax merging of credit records and mortgage data. you can see, again, you see that increase in refinancing activity, even in this data set, between the high and low loan-to-value ratios but it's delayed a month or two. why? it takes time when we apply for the mortgage and when the mortgage actually closes. this is actually origination of the mortgage and that's kind of even though the policy took place in november you can see big effects in january and february because it takes time when an application turns into an origination. so what about people removing equity? this is one channel for which we're going to be linking to real activity in the economy. you're basically lost your job, economy's suffering you have some equity in your house you

basically want to refinance when interest rates go down tap into home equity, you might remove some of the equity you might be able to use for current consumption. you can see the amount of equity being removed from the refinancing during the refinancing process in response to the quantitative easing was much bigger in the places that had equity. this is not surprising. you can see that they tapped into their home removed some equity during the refinancing process more so than in the cash out refinancing. i -- before i show you the next picture we kept talking about the effect of monetary policy on house values something you think about. the key thing to keep in mind is that house prices were falling a lot in all areas in this period. monetary policy might have stopped the house pricing from falling a little bit more in one region than another but it didn't cause people to get equity in their house, people in las vegas still had low -- no equity in their home in november

of 2008 and monetary policy might have stopped it from getting less but the key difference is, it just didn't help put equity back into people's houses. house prices continued to fall for the next 2 1/2 years from the start of the recession. okay. this in is just a measure of spending. one measure of local spending we look at in the paper and that's new auto purchases. you can see between the high and the low core tiles of msas in terms of their spending on cars or at least purchases of cars they were tracking near will identical with each other until qe. there's a delay by a couple of months. why? by the time the origination started, applications started, in turn it origination it too three months before it showed up in spending which is what we would expect if people were liquidity constrained and needed to tap into equity before they

could increase spending you see a pretty big difference in spending measured by new car purchases between places that had equity in their home and the places that didn't have equity in their home at time of the monetary policy extension. i have two more slides before i conclude. is this a common feature of all recessions? the answer is we don't find evidence of. what i'm comparing in the paper is the 2008 recession, each one of those dots are different msas, about the correlation, change in house price with the change in the unemployment rate. in this recession, a strong unemployment growth and much housing changed. the gray was not there in 2001 recession. there was weak correlation between house price growth and unemployment changes at the local level during that recession. then when we look at differential refinancing

behavior across msas based upon, in this case not how hutch equity but unemployment changes, we don't see any differential response between the two regions in this recession. if anything, you might see a stronger response in the parts of the country with the higher unemployment rate in terms of propensity. this correlation between collateral shocks and something like unemployment that is driving the differential hetrogenaity during the recession. back of the number on some of this -- i'm out of time. as we go through and look at how much due to refinancing cash-out, how much is due to mortgage reset. in the paper i didn't talk about it here, we have a bunch of variable rate mortgages we look at. we look at home equity line of credit and vectors ways households can get cash into their hand through the interest

rate change from qe. you can find of the total effect that we find in terms of cash out into turn consumption home equity lines of credit, refinancing cash out, reduction or changes in mortgage payments about 15% of it went to the bottom core tile. about 30% went to the top core tile. you can see that the distributional effects were helping more in places that were doing relatively better than the places that were doing relatively worse. this is my summary. it's important to understand interaction between monetary policy in the u.s. but europe as well. we might want to think that the collateralized lending channel might have an effect that we need to think about. at least during the great recession that monetary policy at least by our estimates, was ex-as sper rating regional dis

dispersion dispersion, helping parts of the country that were doing rel tvly better. going forward, who holds assets and whether we can think about optimal policy. that's it. >> now we'll hear from mark zandi from moody's analytics, as long as i've been a reporter in washington, someone who we turn to help us think about the way that regions of the country differ. mark? >> thank you dave. i can remember my first conversation with you was about -- was about the mountain west and i can remember you telling me that doesn't make sense. i literally remember, must have been 25 years ago, i literally -- actually, you were right, i wasn't making a lot of sense. i really wasn't making a lot of sense. let me say a few things. first, i enjoyed the paper very

much. i think it advanced the ball our understanding of the efficacy of monetary policy. and i'm very sympathetic with your two premises, one that if you've got an economy with a lot of different regional economies it impairs efficacy of monetary policy in term of sensitivity to the monetary change but also in term of the timing. i think that's intuitive. feels right to me. also that monetary policy changes can have different regional economic consequences. that also, i think looking at a lot of business cycles over 25 years, i think that's -- that makes sense to me. it's very, very intuitive. i do have observations though. first, i think you said it, you said it in your paper this is a case study. it's very partial equilibrium kind of analysis.

you focused very nicely on one aspect of qe monetary policy and the impact on regional economies and that's the refinance channel and cash-out withdrawal. very important channel no doubt. but only one of many. and i think to really truly understand the effect of qe, other monetary policy on regional economies has to be considered more broadly. so one example of that would be the impact of qe lower interest rates on a.r.m. readjustment. you mentioned this in the paper i think you're right, but you're using the wrong counterfactual. you almost said it up here, you were looking at the change in the payments resulting from the decline of interest rates but the real counterfactual is what would have happened had the

federal reserve not lowered interest rates? the impact on the economies would have been much more severe. go back to before qe six-month libor, key index rate for the two 28 subprime a.r.m.s was 4%. right after qe it was zero. that made a huge difference in terms of the two 28 a.r.m.s. there was a lot of concern of what impact they had when they actually readjusted and the fact they did not readjust was a huge boost to those regional economies, like vegas, california, miami arizona. so very very important. there's other channels that matter. you know lower interest rates obviously help out the banking system in lots of different ways. lowering cost of funds, that was very, very key to the hard-pressed banking system in the regional economies. one might argue the lower

interest rates were helpful in bringing out mom and pop investors that ultimately cape in and caused house prices to bottom out and to rise. without those lower rates investors could have come in when they did, certainly not as soon as they did. second order effects. one of the interest things you spent out why spent a lot of time on it the impact of the cash-outs on consumption particularly auto sales. well, the benefits of those auto sales actually go primarily to one of the hardest pressed regions in the country that's the industrial midwest, which was getting hammered by the collapse in the auto industry. there was a case where the lower rates, cash-out refis helped a region of the economy that was struggling to a significant degree. the other thing to consider is regional mobility, very important. even if lifting the stronger economies more than the weaker economies, because of the mobility and the economy that

allows for the economy to adjust much move quickly in unemployment falls much more rapidly in places like nevada, arizona, california, because of texas economy's doing better and you saw the migration flows pick up substantially even if you're only lifting the stronger -- if the official effect is lift the stronger regional economies also has a second order effect which i think is very important. the bottom line is that, when you consider all of these things, i think it doesn't obviate the result. i think it mitigated it to a significant degree. you can see it in the data today. look at employment in the states that got creamed in the downturn, california nevada, arizona, florida. since the bottom of the recession, use q2 09 the bottom of the recession, employment in these economies up 12%.

employment everywhere else up 8%. those economies have done marvelously growth. job growth no those economies have been stronger since 2012. current job growth in the four economies together is 3er. year-over-year compared to 2% from a nation. lots of other things going on, not just qe and you're focused on qe. but it highlights the point we need to think about this in a broader, more general equilibrium sense, i think that's very important. a couple of other observations. one, other observation is that this -- this is a case study. it feels very idiosyncratic to me. for example, constructing a sent scenario for a recession today, the federal reserve does this in stress testing every year, last fed stress test and the scenario that generated it, it was motivated by a -- motivated by

global downturn european crackup, chinese recession. okay. so if that is the motivation if that's the cause of the next economic recession that we struggle with and the fed -- federal reserve's response is lower interest rates then that would also the key effect would to be affect the value of the dollar and benefit the same regions that are getting hit by the slowdown in global activity. so, i think it depends. the case study's interesting but i'm not sure it translates into other types of economic shocks that really does depend on the shock in terms of trying to understand what the impact is in regions. the final observation i make is that this all -- the relationships here may be changing rapidly because the financial system is changing very rapidly. i think if we went through this kind of downturn 25 years ago before 20 years ago before interstate banking the regional impacts would have been much

move significant than they ultimately ended up being. this go-round we had a more nationalized banking system. when indy-mac went belly up in san diego, a large california lender wells fargo could come in and fill the voigt and do a lot of fha lending. the regional impacts are less significant. i think they'll be much less significant in the next go round because not own the banking system is more national but the shadow banking system is starting to fill in holes and that's national and international and that will, i think, in the future, mitigate some of the regional hetero genatity. what those means for policy. point number one i think the paper's really good. and i think i agree with it. it makes a lot of sense to me. it's very intuitive. take that as given. this gets to the final point, this has policy implications.

number one policy implicationing in economies that have regional, we have more aggressive monetary policy. this is patently obvious in the context of europe and european central bank. their regional dispersions are greater than the regional dispersions that exist in the united states and that would have argued for a more aggressive ecb than even the fed. and of course we got just the opposite, of course i think that's one of the reasons why we've got such very different results here. the second implication is that, in an economy with regional hetero genaity you need more fiscal stimulus and less fiscal austerity and that's a point that's clear in the context not only our experience but the european experience and the fact that they went to austerity almost immediately counterproductive and more so than it would have been in part

because of the great disparities in regional economies across europe. finally, on policy this might be more controversial, i think it also argues that fiscal policy should be used to help fa facilitate monetary policy. we got a good example of that in your case study, opening up the credit channel for more refinancing and that was the h.a.r.p. program, home affordable refinance program, first implemented in 2009, revamp in the end of 2011, this cleaned out that channel and actually as of today 3.3 million homeowners in those very distressed states got a refinancing because no longer was cltv a criteria for refinancing for a fannie/freddie loan. a huge difference to those

economies. and i think we need obviously a lot of political economy issues with regard to the statement the fiscal policy should be used to help support monetary policy and used to help monetary policy and vice versa but i think smart policy going forward does need to think about how to work together to clean out the channels to allow monetary policy to work more effectively to work out recessions. thank you. [ applause ] >> in case, the unfamiliar face here from the new york fed who worked on this paper and i'll say on his behalf, anything he

said has nothing to do with what they think at the fed. my favorite is someone at the ifm said when i use the word we and he put up the picture of janet yellen he said i'm not talking about her and me. and i want to pick up on where mark ended the so what question, and monetary policy works on the economy as a hope and it is hard to target it at particular regions. so is there any lesson here for how monetary policy could have been different and i'll pick up on the fiscal in a minute. >> do more. >> so in the end, the places where the people have the hardest effort to consumer aren't able to borrow and you want to stimulate aggregate and your target is aggregate unemployment rate, you might have to do more because you are not targeting the people with the biggest margin to kmum for a

biggest change in interest rate. >> i agree with that. >> and you raised a question about fiscal policy but i would have put it differently. >> of course you would have. >> if you know that your monetary policy is going to work better some in regions of the country than other then should you structure your fiscal policy, the housing policy to complement what the fed does. should they have done more on the fiscal side for the las vegas' a lot of acronyms before we got to harp is that the lesson if you have a housing bust and you know monetary policy in a world where you have fixed rate mortgages where peoples don't adjust automakers, should have offered to with fiscal policy in mind? >> well i think that is the lesson. i would say in all fairness to

fiscal policy makers it the not like they didn't understand this and didn't try to grapple with it. i mentioned harp but there was hamp and there was a zillion and one things designed to address the region problems. but stating these things and putting them on paper and executing on them in the context of the different parties involved -- and also given the complete mess in lots of different levels, in the servicing and in the under writing other facings, a lot of different moving parts here, it was hard to make it work well. and the one program that actually worked in my view well took a little time to get it going, but at the end of the day was slam dunk success was harp. 3.3 million people, mostly in the hard-pressed areas. if you go into nevada in

florida, arizona california, these are mostly -- because they were under water. you could see eric's data, they were under water. and by the way, i live in philadelphia, i don't feel like i have a whole lot of equity. i don't know. should have bought earlier. i wish i had a home in las vegas right now. >> on the market. >> much less than there used to be. but i think the harp program worked exceptionally well. but that is the third thing i said about used fiscal policy to help monetary policy. bup of course you have -- but of course you have to have a prolonged economic problem like a great recession to have time to do that and a typical economic downturn doesn't work all that well. >> and it does have regional components built into it automatically. so we can talk about harp having some effect because it was treating people under water and

did have more effect in some of the sand states and that others but the key is the food stamps and all of the things set to act in when places and people are out of a job or don't meet some asset limits and those are automatically set up to fiscal policy is built in to do this. and we could cut a check to people in las vegas. some is extending benefits and getting them out of water. >> i don't -- cutting a check i don't consider this bailing them out. this is was a fannie to freddy and it reduces the credit. >> but it comes at a subsidy to other people that are being backed up taxpayer dollars so there is some in the background some of this is going on. >> i thought the loser was the

person who held the paper. >> exactly. >> well to the extent that we are subsidizing. >> from they are expectations they expect prepays but they didn't get prepays because they were under water. >> for harp it is a slam dunk, it happened rate, right it happened 2012 is when it really took off. nonagency borrowers were not out of the program so sub borrowers that were prevalent in vegas, they could not refinance under the program. so if you compare it in a world to which we had the trust mortgages and you don't have to under write the loan, even with harp the transmission was quite a bit weaker. >> what about the europe thing, do you draw parallels. mark made many observations about europe. >> this is it, when you starp thinking now, there are two things about europe. the dispersion is bigger and the spanish and the germans are bigger in dallas versus las

vegas and they don't have an inter grated fiscal system so some of the automatic stabilizers we have built in our systems doesn't exist and when we think of monetary policy as being the tool of the european union, if it has distributional effects and big value in collateral value you don't have anything in the background to come in and offset that and now you have to cut checks and the germans have to vote every time they want to cut a check to people in greece as well as into the u.s., it is built into the political process that las vegas got more checks than dallas. >> question in the back there. >> good morning. tom pally from the afl/cio. i would like to roll the

question back a little to something that josh asked and make the connection. we know that contractionary macro policy is legal legal terroran and [ technical difficulties ] and interest rates is sort of a payment between groups, but the question then is could we have done better and josh kicked it off by saying monetary policy versus fiscal policy. and i think since talking q.e. and we want to draw implications for future policy let's try to pull it back versus monetary policy versus another form of monetary policy and josh started to say let's try to get to the policy implications and how we might improve future performance and make monetary policy more

legality arian and can we make it even better and here josh started with the idea of giving the fed somehow or other the power to send people checks directly. i think that is probably -- would be a good way of doing things, address directly these things if you think monetary policy is transfer of funds going to groups and this goes to everybody and the fed can draw it back again and the fed can pull it back and use the existing power to raise interest rates and -- >> but do you see any line between fiscal policy and monetary policy. >> yes. fiscal policy is about buying resources, that is how we've seen it, government spending. but i want to make a list of thinks you might do that and some things. mark talked about harp.

anyone who -- this q.e. in terms of the housing market could have been so much for effective had people just been able to jump and if we could set up some mechanisms to jump through the process so you can get direct access if you want to. and i had to refinance a couple of types and hi to pay $1,800 to pay for mortgage insurance. and i can't believe there are defaults on property rights and so on. that is the justification. it is a fantastic fixed cost to refinancing. that is a policy reconsideration. >> and one more on your list and otherwise we don't have anything else. >> and something i've advocated for for a long time is asset based. and mortgage fixes that. have different reserve requirements based on the state