spending other than social security and health care, would shrink relative to the size of the economy. by 2019, outlays in the three letter categories would fall below the current -- latter categories would fall below the earliest year for which comparable data have been reported. revenues have been projected to rise significantly by 2016. buoyed by the economic expansion. projections based on current law, revenues equal 18.5% of gdp in 2016 and remain between 18% and 18.5% in the coming decade. revenues at that level will represent a greater share of the economy than the 50 year average of 17.5% of gdp. it would still be less than outlays by growing amounts in the coming decades.

revenue from individual income taxes are expected to rise relative to gdp, because incomes will move into higher tax bracket as income gains outpace inflation to which the brackets are indexed. but those increases are expected to be offset by reduction in the corporate income tax and other sources. turning from the budget of the economy, we anticipate that economic activity will expand at a solid pace in 2015 and the next few years, reducing the amount of underused resources or slack in the economy. in our estimation, increases in consumer spending, business investment, and residential investment will drive the next few years. the growth in those categories of spending will be run mainly

from increases in hourly compensation, from rising wealth , from the decline in crude oil prices, and from a step up in the rate of household formation. as measured in the change from the fourth quarter of the previous year, real gdp will grow by 3% in 2015 and 2016 and by 2.5% in 2017, we expect. the difference between actual gdp and our estimate, which is a measure of slack, was 2% of gdp by the end of 2014. during the next few years we expect actual gdp will rise more rapidly than its potential gradually lead eliminating -- gradually eliminating that slack. the labor market, we anticipate that will disagree -- disappear by 2016. this is close to the expected natural rate of unemployment the right arising from all sources except fluctuation in the demand for goods and services. the increased hiring will make people stay in the labor force, boosting the participation rate, which is the

percentage of people working or actively looking for work. our projections are not based on estimates of cyclical developments in the economy. the cbo does not attempt to predict fluctuations in the future. they are based on estimates of underlying factors that affect productive capacity. 42020 through 2025, we project real gdp will grow by 2.2% per year, which matches the estimate of the potential growth. potential output is expected to grow much more slowly than it did during the 1980's and 1990's, primarily because the labor force is expected to expand more slowly than it did then. the growth will be held down by the ongoing retirement of the baby boomers, by relatively stable labor force participation

rate among working age women after sharp increases from the 60's to the 90's, and by policies in current law. the elimination of slack will eventually remove the downward pressure on the rate of inflation and interest rates that existed in the past several years. by our estimates, the rate of inflation as measured by the price index will move up gradually to the federal reserve goal of 2%, hitting the mark at 2017 and beyond. interest rates on treasury securities, which have been low since the recession, will rise considerably in the next few years, but remain lower than the previous decades. between 2020 and 2025, the projected interest rate on treasury notes are three poi -- 3.4% and 3.6% respectively. we will be happy to take your questions. please do it by identifying yourself and the publication you

work for. >> william hall. do any of your projections [indiscernible] >> our baseline economic projections and budget projections are consistent with each other and that the budget projections are what we think will happen to the federal budget given the economic outcomes we projec -- predict. and the outcomes are what we would expect given the baseline projections. the baseline projections are always dynamic in the sense in which you mean that word, in

which the budget and economic effects are calculated and projected together. >> if i may, i guess i am using a term [indiscernible] >> the interest of people in congress is focused on the cost estimates and not the baseline. cost estimate there has been a long-standing -- for cost estimate, there has a long-standing tradition that they have not included the effects of the proposals on overall macroeconomic conditions and the feedback on the macroeconomic changes to the federal budget. however, we and jct have produced studies on the effects of those overtime and reported goes to congress. nearly every year, we do an analysis of the president's budget that starts with the revenue provisions without allowing macroeconomic conditions to change. then we go back into an economic

analysis including the feedback to the budget itself. we have done that essentially every year for a dozen years. we also do that sort of dynamic analysis when we do different paths for federal debt and the long-term outlook. we did it for comprehensive immigration legislation a few years ago. and our colleagues at jct have done it for tax legislation including the tax reform proposal last year. we and jct are accustomed to doing analysis of the economic effect of major pieces of legislation. what is different in the role of the house has adopted is that the budgetary feedback of those economic changes would be included in the official cost estimate rather than being presented as supplementary information which is the way it

has been done in the past. and we are prepared to deal with the dish -- do what the house rule asks us to do. >> two things. one, you have verbiage about interest costs. there is often concern that those will skyrocket. will a run down before you see what treasury will be paying? also, much is that -- how much of that slowed inflation is people, for lack of a better term, being cowards. >> the first question, about interest payments. as you know, interest payments on treasury debt have been low

the last several years. as a consequence, even though federal debt is historically high relative to the size of the economy, interest payments by the government have been quite low. so interest payments this year 2015, we project will be 2020 $7 billion, 1.3% of gdp -- $227 billion, 1.3% of gdp. but we expect that interest rates will move up and as the treasury needs to finance new deficits and rollover the existing debt, the interest rate it pays all the outstanding debt will rise. -- on the outstanding debt will rise. with that as high as it is interest payments will rise significantly. interest payments that this year will be tortured $27 billion will be a hundred $27 billion -- $227 billion will be $847 billion in 2020. as for wage growth, compensation growth since the recession has been a good deal lower than it was before the recession.

those of measures of compensation growth picked up last year -- although some measures of compensation growth picked up last year, they remain below. that is a consequence of the slack in the labor market. when employers are trying to attract, if a lot of people are looking for jobs, employers do not have to offer as high of a as they would -- of pay as they would in a tighter market. as it continues to tighten, that will put pressure on employers to raise compensation. using the compensation growth will pick up. -- we think the compensation growth will pick up. the share of income going to

labor has fallen and we think it will rebound but not to the level that it was 15 or 20 or 30 years ago. >> is a permanent structural change -- is that a permanent structural change? >> there have been permanent structural changes, but knowing how important those will be is of course very hard. certain technological changes, globalization, or the forces that people look to a lot. but there are others at work as well. how important those factors will be in the future is very hard for us to know. the labor share of income tends to fall during the initial recovery from recession's, so we

have a good basis for expecting it will come back some. how far back it will come is hard to know and what we have done is to project it will come back way towards the normal average of the previous decades. yes? >> paul, "roll call." two-part question. number one, do you have a view on the house rule on dynamic scoring being a good idea? number two, will there be a greater level of uncertainty to a dynamic score? compared to the conventional score. >> we do not have a position about whether the house rule is a good idea or not. we think it is natural for the members of congress to be interested in the macroeconomic consequences of major pieces of legislation and that is why cbo has built models to estimate goes macroeconomic consequences.

we have in the past few years devoted effort to improving those models, getting feedback from outside effort, revisiting -- experts, revisiting the assumptions, and writing a collection of papers that spell out fiscal policy. there is a set of eight papers that go through the key aspects of the model, one of which is a dozen years old but the others written in the past three years. have understood the important better now -- we have understood the importance of that analysis to congress bought it is up to congress to decide how to use the information. if it wants to use that information to come in a supplementary form and not in cost estimates, we are happy to do that. if the congress wants that extra set of estimates to be included in the "cost estimates as the house will not require is for certain legislation, we're happy

to do that. it is up to congress to decide the form. as for uncertainty, the macroeconomic effects of legislation are quite uncertain. it is also true that the non-macroeconomic effects can be quite uncertain. it is always important for readers of our work to understand that there is a great deal of uncertainty surrounding the numbers and we give point estimates because the budget process basically requires point estimates. things have to at all, politicians are given to

community -- add up, allocations are given to committees. but we understand there is uncertainty. in some cases, the macroeconomic effects of legislation will be more uncertain than the effects we estimate holding the macroeconomic conditions fixed but i think that will always be the case. it depends on the legislation we are analyzing. as a general matter, legislation that makes all changes and programs of the tax -- small changes in programs of the tax code or programs, analysts have a reasonable basis for making an accurate projection. but for major changes in policies or comprehensive reform of the tax code or on immigration policies or the health insurance system, there is a great deal of uncertainty that accompanies the estimates we provide for that as well. i think that is a general feature of that people need to understand and that we need to continue to convey.

as clearly as we can. for a number of our macroeconomic analyses, we have provided close to a central effective convey the uncertainty -- the fact to convey the uncertainty. >> "new york times." the changing major health care programs -- can you talk about the changing major health care programs? >> yes.there are a few components of the major health care programs i think we are discussing. for the coverage provisions of the affordable care act, and to be clear what we mean about that, the affordable care act changed features in the tax code.

we don't about them in large pockets, one being the provisions that expanded health insurance coverage, with the expansion of medicaid and the offering of subsidies. four other provisions, -- for other provisions, changes to medicare and the tax code, the provisions we are focused on are the expansion of health insurance. for those provisions, we have made a significant downward revision to the estimated costs over the next decade, about $100 billion against 7% less than we expected last year. relative to our projections, in march of 2010, the cost of those insurance coverage provisions between 2015 in 2019, which was the last five years -- and 2019, which was the last five years amid estimates in 2010, we have revised down the cost my about 20%. -- by about 20%. that has been the collection over the past five years. there are a number of factors that underlie the set of revisions. perhaps the most important one is a slowing in growth of health insurance premiums. but there are others as well. including the supreme court decision of a few years ago and other external factors and

improvements in modeling. we also have revised down over the past several years our projections for meta-eight and medicare's funding -- medicaid and medicare spending a great deal. there are a number of factors that have fed into that. there are, apart from economic changes, for medicare that has not been significant changes in this projection relative to the last one of art from things related to the economic forecast. in terms of technical revisions, not related to economic conditions but focused on developments in the programs and our interpretations, i would not make a significant downward revision to the medicare production. but -- projections. all told, since 2010, because of this assessment, we have marked on projections of federal medicare -- download projections of federal medicare spending and premiums by 15% each for 2020. some of that has happened already and some of that is

projection of slower growth. we have gone farther in that direction this time in our projections of health insurance subsidies and the cost of the aca expansion and and medicaid. no movement in medicare. robert. >> "washington post." you have figures on the decline

in discretionary spending and the share of gdp. only figures adjusted for population -- were the figures adjusted for population change? >> we do have a continuous series. there are spots -- do not have a continuous series. there are spots that talk about that but i do not remember those numbers offhand. they are spot references and we do not report anything that adjusts for population growth. i am happy to 10 and all for a moment while youth -- hem and h aw while i sort through this. for example, in 2010 and 2014, on page 79, and quite down in 2014, pending for defense -- between 2010 and 2014, spending for defense declined from constant $2010. excluding the funding for -- constant 2010 dollars. also between 2010 and 2014, on the same page, funding for nondefense discretionary programs defended -- declined by 4.5% or 11% in constant 2010 dollars. >> "wall street journal." i was wondering if you could talk about what prompted the

downward revision on potential output? >> we revised down our estimate potential output by about 1%. from what we had in our august economic forecast. the primary source of that downward revision was a reassessment of productivity growth and especially productivity growth in the nonfarm business sector.

we had previously thought that the slow growth of an actual productivity, these loans are growth and productivity over the past several -- the slow observed growth in productivity, was due to a shortfall of growth relative to the potential productivity. we thus anticipated that actual productivity would rebound to its potential level over the coming years. and for some time that seemed like a reasonable expectation because productivity often fall short of its potential level

when the economy is weak, when demand for products is weaker. firms do not push workforce is as hard as they do when demand is -- workforces as hard as they do when demand is strong. we had growth in the non-foreign business sector. as the economy has strengthened over the past two years, when we would expect the actual growth to not be falling farther away from our projection of potential, but it was. that did not seem to us to be the best forecast in the economy that was strengthening. so we lowered our projection. it is still above actual growth, we think there will be a rebound but not as long a rebound as we were expecting.

this problem that we face, in trying to understand how much of what has happened to productivity over the past several years is due to a business cycle and how much is due to underlying factors, it is hard and becomes harder as the length of time during this week economic condition lasts longer. there are people who observe that productivity growth seemed to slow before the recession. that was a pretty short period. and then we had a short recession and rebound. it is hard to know how much of the past dozen years has been because the economy is weak and will go away as it recovers farther and how much is underlying structural condition that we face. on this particular variable, productivity, we think that the persistent slow growth in the face of an improving economy suggests that more of the slow growth is due to structural problems rather than cyclical. because of that we have marked

down expectations for productivity growth going forward. >> i have a technical question. on page 98, you talk about the erosion of tax, over 2.5%. you see that, it is in the right-hand column? the final factor, the subsidies, talks about pushing tax -- corporate tax revenue up and you expect the average tax revenue closer to the historic average. how does it offset? >> i think the way to think about or projection of corporate tax receipts -- our projection of corporate tax receipts is how much will be earned by corporations in this country.

does the profits earned in this country are taxed with more than profits of u.s. corporations earned overseas. we focused on the domestic profits. the question is, what share of that is paid in tax? one cannot tell by looking up the statutory corporate tax rate because the tax paid depends on the expense of the deductions and salons that companies will make -- and it's so on that companies will make. part of what is going on is that the effective tax rate on

domestic profits, the total amount of money the treasury has collected, expressed as a ratio of domestic profits, fell during the recession in a way that we cannot entirely explain. part of the reason we cannot explain it is that all we have for the past few years are the total amounts of money, the details before help unravel the puzzle that are only available to some. we have seen this tax rate defined as the tax collection divided by profits, it has been low. we think there will be recovery in the long-term average because we are making an educated guess that what happened in the last few years was partly due to particular economic circumstances so there will be rebound from the phenomenon. there's a different thing going on which is that companies are trying to find ways to reduce the corporate tax that they pay. as we explain here, some of what they do involves moving business income from the corporate sector to the noncorporate sector, out

of c corporations and into s corporations. there has been a downward trend in business income tax at the corporate level and we think that will continue. the second factor we describe is a set of effort by companies to increase income shifted out of the country. it is subject to less u.s. tax. there are various ways to do that. about -- we talked about inversions. corporate inversions are a part of roughly half of the erosion and that factor is at work in reducing corporate tax receipts over the decade. but that is offset in part over the next few years in an increase in the rebound in the tax rate so there are other factors that we do not have a firm factor own. -- factor on. we think that corporate profits will decline later in the decade because of rising interest rates and faster compensation growth will push down the corporate share. and that factor, which affects hundreds of denominators, we

think those factors will economic profits down relative to gdp and pushed on corporate tax revenue expressed as well to gdp -- wealth to gdp. i will check with my experts to make sure i got that right. other questions. >> the updated revenue and spending outlook, you will be looking at the debt limit this year? >> we do write about that in the report. the debt ceiling was suspended as you know, through the middle of march of this year. when that suspension ends, the debt ceiling, under the law as it stands, will be reset to equal the amount where the ceiling was before it was suspended, plus all of the data that have been issued since it was suspended. the debt will be equal to the

outstanding debt subject to plus all of the data that have been issued since it was suspended. the debt will be equal to the outstanding debt subject to limit. the treasury will immediately in march be unable to increase its net borrowing. at that point, we presumably will deploy is customary set of -- this customary set of extraordinary measures to obtain funds without increasing debt. we think those measures will be effective through september or october of this year. at which point we think they will have exhausted their extraordinary measures. shortly after that, the treasury will have run down its cash balance and be unable to meet all of its obligations in a timely way unless some other

action is taken. that projection of september or october is uncertain and depends on inflows and outflows of money over the next eight or nine months. so, you know from previous rounds of this, projecting the treasury's cash flows is a very uncertain business. so we will continue to keep the congress apprised but we are not likely to know more precisely when the treasury will run out of extraordinary measures or run out of cash until we get quite close to those events occurring. >> could you go over the major factors causing the deficit to fall in the next couple of years?

>> the debt? the deficits? >> the deficit. >> i will be happy to talk about the pattern. we think that the dollar amount of the federal deficit will be $468 billion this year and will begin to rise expressed as a share of gdp. the deficit will remain at 2.5% through 2018 and will rise beyond that, rising from 2.5% in 2018 to 3% in 2019 and 4% by 2025. that pattern results from a combination of factors. part of what happens is that tax revenues rise a good deal in 2016 relative to 2015 because of the expiration of some tax provisions in the improving economy. tax revenues we think that will

be 17.7% we project will be 18.4% next year. there is an increase in tax revenues, a market increase. -- marketed increase. after 2016, they remain the same relative for the rest of the decade. they move up quickly but then are about flat expressed as a share of gdp. the retirement of the baby boomers is putting an upward pressure on spending for social security and medicare and medicaid throughout the decade. there will be more than one third more people receiving medicare and social security benefits that are receiving it today. that pushes up spending. but at the same time, the other

thing is rising interest payments. we think interest rates will move up over the next few years the government interest costs respond slowly to that because the government does not roll over all of the debt in one year. the average interest rate paid on treasury debt responds gradually to mortgage interest rates. we actually reported in table 1.3 on some page of the report i do not have the page number, the average interest rate on debt held by the public rises. we have the aging baby boomers rising health care costs per person, the expansion of health care subsidies, rising interest payments. all of those are pushing spending up. on the other hand, there are declines in defense purposes and other purposes like highway infrastructure, r&d support,

education and training support health research and veterans health care. and in the non-social security non-health care benefit programs like snap and ssi. the particular pattern from year-to-year results in this combination of factors. again, the underlying point is that we have a handful of very large federal programs that provide benefits for older american or provide health care benefits or both. and with the rising number of older americans and the rising cost of health care, those programs get more expensive. by 2025, one quarter of total projected spending will be for social security benefit. some of the other biggest pieces

are medicare, defense spending interest payments. those four things alone, social security, medicare, interest payments, and defense spending those four items together represent two thirds of total federal spending in 2025. the growth in federal spending is not coming from a growth across the board in federal programs, is a handful of large programs. the other parts of the budget as i said earlier are getting smaller relative to the gdp and that damps some of the effects of the growing spending. so it is a combination of these factors but the trend is clearly for rising deficit for reasons that are very well understood by analysts for decades.

>> david, reuters. could you clarify your status? are you a contender for the job as chairman? >> my current status is to be the director of cbo. my term ended on january 3 but under the budget act of 1974 directors can continue to serve until after the terms of expired until new directors are appointed. and i like the job a lot so i am continuing to serve. i expect to do that, unless and until a new director is appointed. whether a new director will be appointed and when that would occur, i do not know. and i will not be the first person to go on the first person to tell you. that decision rests with the senate.

and they will speak to it when they want to speak to it. >> liz. can you talk about your estimate that by 2039, public debt will exceed 130% of gdp? and you say the last time that happened was after world war ii. >> this is a projection from the long-term outlook last summer, we look out 25 years. we thought that under current laws, that debt would exceed 100% of gdp within 25 years. we have not formally updated the projections but the projection in this report of debt relative to gdp at the end of the decade is quite close to the production -- projection we used last

summer in building longer-term projections. so we have not done anything since last summer that would change our view that unless some significant changes are made in spending or tax policy or both that debt will exceed 100% of gdp within 25 years. and be on an upward trajectory at that point in time. the only other time in our history that our publicly held debt was at 100% was at the end of the second world war. what was different then was that the debt ran up sharply during the second world war. but then started to come down relative to gdp. that was not particularly, on that picture, because large surpluses were run, but because as you mentioned, gdp grew rapidly in the budget was close to being in balance.

the run-up was because of the cost of fighting the war. the run-up that we have seen is heavily because of the financial crisis and recession and the policy responses. so those are, one hopes, one-time events. but behind that, is this ongoing and highly anticipated event of the retirement of the baby boom generation and the movement of a large number of americans to an age where the federal government provides significant benefits. and also underlying this is the rising cost of health care per person which has been less over the past years than projected but we think the

best projection is a continued growth. with many more people eligible for social security and medicare, and medicaid as time goes on, and with the cost rising, we face a structural long-term challenge that is quite different from the situation at the end of war. the fact that debt doesn't rise very much relative to gdp over the next decade, does not mean that is high-level doesn't have significant cost. with debt this high, there will be significant interest payments or interest rate rebounds. there will be crowding out of capital investments. as time goes on, there is reduced flexibility for policymakers to respond to future financial crises or

recessions or international events. there is heightened risk of fiscal crisis. even if that was -- that was lessened. moreover, we don't think the debt will be stable. another point we emphasize is that even though the reason that debt only raises a little bit relative to the size of the economy, despite the demographic changes and rising health care cost, is really because it is set into current lot a very -- law a very sharp reduction in spending everything apart from social security. compare the historical experience. nondefense discretionary spending. the congress appropriate money each year. it is about the same this year as it was 50 years ago. this spending has fluctuated relative to the size of the

economy over time but it has shown no evidence trend over 50 years. under current law, given the caps, it will fall to a lower percentage of gdp than we have seen. it turns out it was not available that way. defense spending is a little more complicated. it has trended down relative to the size of the economy over the last 50 years. but not over the last 20 years. it fell a fair bit in the 1980's and 1990's. but, it has stabilized this year, over the last 20 years. we think that need depends on what you think will happen in international events. defense spending is also on track to fall to a smaller number relative to the size of the economy. we have the data.

so, there's this very sharp shift underway under current law, and what the federal government spends its money on. because of the appropriation process, it occurs one year at a time, the caps have been set without decisions made about which programs will be provided. in the future it would have been provided under historical experience with those government programs. we highlight the issue in our report. as a risk going forward, that the caps have been set but the decisions about what will the cut and what will not be cut have not yet been made. other questions? yes? >> can you go back and talk a little more about inter-spending on projections, and although it

is lower than the past it is still substantially higher. i wonder whether not you consider the possibility that interest rates have declined structurally and are going to be lower in the future then anyone had predicted four or five years ago, and today what that would do? >> as i mentioned, we expect interest rates will rise considerably. but we do not think they will rise up to the levels they were at over the past few decades. we wrote about this at some length in our long-term budge outlook. we had a brief discussion. we recapitulate in this report. we got about what would be different in the future relative

to the 1990's to 2007 period. in that era, we think that inflation adjusted treasury interest rates will be about three quarters of a percentage point lower in the future than they have been in the past. that three quarters of a percentage point difference is pretty substantial. we project that a nominal tenure treasury note rates will be 4.6% by the end of the decade. with cpi inflation of 2.4%, that means inflation-adjusted tenure rate of 2.2%. that compares to about 3% for real treasury 10-year rate in the 1990's. we made a substantial downward adjustment. that adjustment is the net effect of a number of factors. we list 4 going in different directions.

number one, forward growth of the labor force, number two slow growth of productivity. both of those factors we think lower interest rates because it basically affect the product of capital in the production process. we also think greater income inequality is pushing up saving which tends to lower interest rates. and we think they'll be greater risk premiums on private security. treasury rates will be held down because of their greater demand. those four factors alone would've argued for a lower -- larger revision act. on the other side, 4 other factors that will push up interest rates. one of those is greater federal debt. another one is slightly smaller capital inflow from overseas. a third is fewer people in their prime saving years.

another way of saying baby boomers are moving into older ages. people in their prime saving years. and also the higher capital share argues for higher returning cap that compete with capital for investors. as best we can judge, the factors pushing down rates relative to this, will be more powerful than the pushing up rates. that is why we made this revision relative to history. nonetheless we think rates will move up from where they are now. that is consistent with expectations of participants in financial markets and consistent with our own modeling of demand for a fund as the economy strengthens. again, these are changes we mostly made last summer in the long-term budget outlook and incorporated into our august projections. we made it further small

downward nudge to interest rates because of a declining interest rates in other countries that makes u.s. assets relatively more attractive. and pulls more money into this country. this is no doubt one of the sources of great uncertainty in our projections. we offer an appendix to the report, and a rule of thumb for how the projections could be different if economic projections were different. we offer rules of thumb for faster or slower growth in gdp or higher or lower inflation and higher or lower interest rates. not surprisingly, budget outcomes are quite sensitive to interest rates. that is much more true than maybe a decade ago. debt is so much larger than it was a decade ago.

i think we have balance the risks in this production but the risk on both sides -- there is risk on both sides. [indiscernible] >> and trying to get a better handle on it about subsidies. people in the exchanges. reduce the little bit of your outlook for that. this is based on one year of experience. now, i guess, given the path how confident you feel about your original score? and things like subsidies. [indiscernible] >> you are absolutely right. to say we have only observed

exchange enrollment in subsidies and medicaid enrollment in people who are newly eligible for one year. we don't have detailed data in all cases. we will learn a good deal more in the years to come. on the other hand, the source of the revisions that we made, i think have more foundation then just looking at last year would suggest. part of what we have done is take have more information about private health insurance premiums over five years since we did these estimates. it is true that insurance premiums could jump up next year. they could fall next year or rise that small amount. there is a lot of uncertainty. we have seen now for a number of years significantly slower growth in private health insurance premiums than expected in 2010.

that is including premiums paid in the insurance exchanges last year. and the premiums that we see in 2015. i think we have a strong basis for making a significant downward revision for the protected costs, although we have the amount precisely right now, we don't know. similarly for medicaid, we have seen a number of years of slower growth in medicaid costs for -- per beneficiary than we expected in 2010. although there are particular aspect of the affordable care act we have only saw for one year, we had observed others for some time. we have a much stronger foundation with a cumulative set of foundations with the cost is the medicaid expansion than just from looking at last year's data. although once again there is

considerable uncertainty. [indiscernible] >> how important is that moving forward? >> as you know the excise tax on higher premium insurance plans have not yet taken effect. it is particularly hard to -- we don't even have the one year in that sense. there is anecdotal evidence of people reporting making adjustments to their plans. to keep premiums on a trajectory below what they think they will be subject to that tax. those are just anecdotes that this point as far as i'm aware. we expected that most of -- there will be a significant adjustment by employers to try to keep premiums below the

threshold low. of the total revenue that we show occurring because of the tax, only a quarter of it is to come in the form of excise tax receipts. three quarters of the revenue comes in the projections because employers hold down premiums and therefore pay more in cash compensation to their employees because they will be paying less than they would advertise in a nontax health insurance. the extra cash will yield higher tax revenues. three-quarter of the revenues will never show up. they will nonetheless occur because of the excise tax, if that were to induce a behavioral response. we don't know what that would yield. it is true that this all equals the slower growth in the health insurance premiums in the economy of the whole.

lots of other things are going on. we incorporate all factors in making projections. >> you popularized the term spending the curve. we have had five years now. you feel relatively confident at this point. how good a job do you think aca has done of spending the curve of national health expenditures? >> we don't know. the reason we shied away from using the term bending the curve is because people tend to do it -- view it as a permanent state of affairs. what we do know is the curve meaning both federal health cost and national health costs, has been flatter over the past several years than we anticipated. the curve has been flatter. whether it is flattened -- and

we think it'll stay flatter for a little while. but our projections don't show it being this much flatter indefinitely. we think it can be faster health and -- growth in health care costs again. but we don't know. we are trying to balance the risks of putting too much weight on a phenomenon that has gone on for a little while. or putting too little weight on. but the further question, is given that these have been flatter, how much is that is attributed to the affordable care act. we just don't know. so the slower growth of payments to a number of providers, even medicare, is a specific aspect of the affordable care act that we thought would lower medicare

spending relative to the pre-aca world. and it has. but, whether other provisions of the act have had other or more indirect effects on federal health care spending or national healthcare spending, we just don't know. anything else? ok it is great to have you here as always. thank you very much. [captions copyright national cable satellite corp. 2015] [captioning performed by the national captioning institute, which is responsible for its caption content and accuracy. visit ncicap.org]

>> coming up, a discussion on how government policies are affecting the internet. then senator lisa murkowski response to president obama's call for wilderness protection for more than 12 million additional acres within alaska refuge. then a later, a panel of print and tv journalist discussed how the media covers race issues. >> today, u.s. trade representative appears before the senate finance committee to answer questions to president obama's 2015 trade policies. see it live at 10:00 a.m. here on c-span. >> the house select committee on a benghazi meets today to continue its investigation into

the 20 12th attack that killed ambassador chris stevens and three other americans for state department officials and representatives from the cia are expected to testify. you can see it live at 10:30 a.m. on c-span 3. >> next, a discussion on government policies and they are affecting the internet. representatives from google, you all, facebook, ebay joint government officials at this event held by harper's institute of politics. this is 90 minutes. >> good morning. i feel like we should a vote on something. on behalf of the institute of

politics and the internet association, i am delighted to welcome everybody here and in the virtual world because we are live streaming right now. we are live streaming as far as i know. you already know it if you are on the live stream. if you are here, you don't need it. we are being recorded for posterity by our good friends and neighbors at c-span. they will be putting this up against the friday night battle between the house of representatives' subcommittee on veterans affairs. i think we are looking pretty good. be aware. we are broadcasting. we are in the penthouse on the fifth floor of the kennedy

school. the school is named after john f. kennedy. the institute of politics here has been from the start designed it to support the political aspirations and public service of young people, something president candidate was well-known for. who can forget his speech? how far seeing it was to see the role of coding and the app ecosystem for the future of the country in the world. we will unfold as two panels. i will be moderating the first one on will the government break the internet? the answer to any question in a headline is no, but we will have to investigate. then we will have a second panel moderated on why internet policy matters.

this is a rare and special opportunity to gather the amazing group of people we have around the table today. we hope we make the most of it. before we start the first panel, i am pleased to turn the mic over to michael beckerman from the internet association. >> thanks to the harvard institute of politics for partnering with us. we are going to ask the question will the government break the internet? these are important to the future of the internet and the internet users and companies around the world. as we get into an election cycle that voters care about, we're looking forward to a great conversation. thank you for moderating. >> i am going to take off the introducers have to put on the moderators hat.

i'm going to get into the middle of the well. this is where we need jeopardy music. let's see if this will work. all right. as you may have noticed, i have called this panel. this is a roundtable that has right angles. we have approximately 25 people on the panel. there are roughly 85 minutes to do this. introductions would not be done before the time is done. we have a special configuration. i want to use this to our advantage. this is how we suggest we do it. to bear in mind and adopt the ideal as we proceed, we are

speaking for posterity. don't forget that c-span is recording this. imagine the 20 years from now, when people have a jet packs and nick bostrom's ai has taken over. it doesn't look good for us. they will be looking back nostalgically at 2014. maybe the names of the companies will have changed. i urge us as we discussed to think about what we would be saying to somebody in the future about the issues we are facing now. i think that will keep us at the right level of importance for what we want to say and it gives license to explain a little bit the issue that among many people here dealing with governments all the time on this who may represent government, we have

shorthands for all sorts of things. there needs to be a translation available for others. i think that would be quite good. that also means that since we are not doing introductions around the table, when you first speak up, feel free to introduce yourself and do an extra beat. don't just say ebay, say as you may remember as... or maybe not. echo bay. we have already learned something today. [laughter] an extra beat on the introduction.

the one other guideline i would give is often and especially in government affairs, our points are usually three in number. i like to adjust that to one. give or take. how about just one thing with an elaboration and we can keep the conversation flowing rather than there being three different reasons. that will make it harder to follow. sound good? yes? as michael hinted, we have a mclaughlin group-esque set of issues. we will talk about liability and a surveillance and net neutrality, this is in the news recently. finally, that governance.

the thicket from which there is no return. we want to touch and integrate on all of them. i want to think a little bit about the category of intermediary liability. governments regulate people. at other times, they realize don't regulate people, regulate any institution in between the government and the people and put mandates there and then you can still effectuate regulation in a more powerful way. copyright infringement defamation, these are examples. as the internet has grown in the american context, we've seen some balance struck in what the government may require of intermediaries. one example is coming from europe with a newly recognized

right to be forgotten. adam, where are you? i don't know why it occurred to me that adam should be asked first to weigh in on the right to be forgotten. tell us what company you are from and how is it going? >> i am adam and i lead googles u.s. policy efforts. google is a search engine company -- was a search engine company that did a lot of great things. [laughter] >> humble. >> the right to be forgotten is an interesting issue that a lot of us in the u.s. look at the ruling of this european court of justice and think this is outside the bounds and not something that would gain traction here due to our first amendment and strong

intermediary liability principles. i think it's important to acknowledge some of the feelings that motivate the desire for a right to be forgotten. there is more information about all of us online than ever before. that is true. a lot of the search costs for finding things have been obliterated. it used to be you could find a criminal record in the basement of city hall. the internet has obliterated that. people are concerned about the impact. what concerns me about the right to be forgotten is it has not been balanced with some of the other competing values such as the values of free expression and the benefit of the consumers right to know about things like whether a vendor has had bad reviews against them or a

babysitter has had criminal offenses. these are things that people could request to be removed from search results and i think that's very anti-consumerist. >> is this a shift in thinking with google from a stance that used to say we index the internet. if you have a problem with that, please consult the internet. to a stance that says, there are a lot of tricky things that need to be balanced and may have quibbles with a privately activated process where a person comes to google and maybe a link goes down between a person and the information. this is a new reality. >> the reality is and why the court case in europe has been heavily dissected is it hinges on whether google is a data processor, which the court ruled it is, or a newspaper or

journalistic organization. we are showing some prioritization, which we are. that is the value of search results. search results do both. they believe that these results are more important to the user than others. >> on the mechanics of it, how is it going? give us a sense of the flow coming after the opinion came down. are there plans to open it up voluntarily to other citizens, because i have a small list in my pocket-- >> we have received 100,000 requests. we have a backlog of requests. we have to go through these one by one. >> are their jobs posted for google right to be forgotten processors? >> some of the cases we have seen have been former politicians asking to have news of their criminal convictions

removed. in our view, these are not in the public interest. there is some interest in the part of regulators and policymakers in parts of latin america and asia, looking at that, this is something we should consider. >> so far, you are limiting this right to where it is required of you to implement. >> only within our european domain. >> only in localized portals. if i am alert enough to say i can perform a search in england. are there other companies at the table that watching this very closely?

if you're not a search engine is right to be forgotten implicating what you do, either now or in the future? it's remarkably quiet. this is a search engine specific thing? wow. braddock? i thought i saw a hand go up? >> no you didn't. >> fair enough. >> i am a senior counsel at trip advisor. we were once the world's largest travel site. that point in time is now. it will be disingenuous to say we're not watching with interest how this unfolds. we have looked at the european court justice decision and we have seen how google has reacted

to it. there is a proposed directive that we don't know whether it will be enacted or not. we don't know what form it will take. and if it is, how they each country will implement that directive. the decision raises more questions than it answers. who falls into the exception? what is a relevant and who decides that? >> should google not allow the intervention? >> matters of public interest. where is that line drawn? a politician is one thing. a b&b owner providing poor service to customers, where does that fall in the line it? i think there are a lot of open questions. >> adam was saying its 2014.

it's not as simple as let the internet deal with it. there are lots of equities to be balanced. this is the new reality. is there a similar thing going on on sites like trip advisor that solicit reviews and other information from people that might make institutions -- it might make or break a hotel if it gets a certain ranking or review. this generally gets sorted out in your commitment to your users. are there ways in which this could go awry? >> i think each individual platform has the rules and ethos that users come to expect. i don't know when you what the government determining what's irrelevant. take them down. >> we got rid of the

cockroaches. the prior owners are in jail. how would i do that through a contact form? >> at trip advisor, there are situations where we will remove reviews. change of ownership is the most common one. major renovations, you just put in new wings and hired new management. there are judgments. >> are those judgments made according to your conscience and what makes for a good business? is there a shadow of government intervention should you do that wrong? i know that your lawyer is at

the table. >> he is sitting right here. [laughter] those are made for what is best for the community. the more experiences you are able to offer up to someone who is trying to figure out what to expect when they take time off and spend their hard-earned money and take their kids somewhere, the quantity is very helpful. there are certain things we determined that are irrelevant. maybe they should be forgotten. in certain situations, you put new linens on your beds, we are not going to change everything. >> anything else on the right to be forgotten before we move along? i can't tell it's because it's uncontroversial or because it's totally controversial.

[laughter] >> i just want to add one thing. this won't go away as an issue. the debate will be the next arena where this takes place. we have precedent under the fair credit reporting act where we say that information should be used like bankruptcy or foreclosure, it can't be used to get somebody past 10 years. we've made a judgment through legislation in this country that information has an expiration date. we acknowledge that there is value in the meantime. we don't want it to be held against somebody forever. we will have to weigh a lot of these things legislatively about how we feel a person's history ought to follow them prefer that certain types -- for certain types of information.

we require sex offenders to register forever. if you look at france, they believe any crime should be expunged from your record once you have done your time. they want people to have a clean slate. we will have to make these judgments legislatively. >> it's easier when the information is in the hands of the government. once it's in the public domain it's much harder to effectuate putting the genie back in the bottle. five to 10 years from now, will there be an american right to be forgotten recognize legislatively or as a matter of customer service? >> not on the scale, it will be miniature laws and other steps taken to it dress certain types -- to address certain types of information.

>> are some of the accordance is with american law, before the modern lab and the companies here were up and running, are those holding up well? defamation in another state law claims that 70 might bring often involved an intermediary, those tend to be off the table through the communications decency act. we see section 512 of the copyright act which encourages intermediaries in many circumstances against copyright infringement claims, even if the site is the way in which the infringement is happening if they take things down when asked.

is that holding up well? is there a problem anybody would have to tweak that in america one way or the other? this is the most wonderfully -- >> i am with ebay. we do payments through paypal and we have a small company called a stub hub for ticket sales. this goes to your questions about the communications decency act. if you look, we would have said today that the internet had been broken. the irony of section 230 is a law meant to limit speech on the web ended up being one of the

greatest proponents. part of this is the government acting in ways that doesn't come out the way it came out. i do the two did get anywhere close to passing section 230 in the united states. as the far end of the pendulum it allows everybody to start from a position of we are going to work to keep openness as the basic principle and work act from there with the exceptions. >> it sounds like you are saying, congress, keep up the good work. thanks so much for your enlightened legislation. >> a senator from nebraska ended up being a visionary that he never expected to be. >> you can layer on acceptable

use policies, things that in the company's judgment like auctions that are prohibited. >> part of the protection that these acts provide is the ability for companies to exercise judgment and to get rid of things they believe are in the consumer's interest. >> when there's been a conflict in a commerce, the rolex company has a litter of kittens over fake wetsuits -- watches being sold on ebay, is that settled now? >> i would say that the great copyright and trademark wars over the internet are not settled. there is the ability for people to push the envelope.

for the vast majority of commerce, it's a much better situation than it had been. i think there is no reason to reopen the digital millennium copyright act and most companies have adopted similar statutes and regulations. >> can i ask you to weigh on this? >> i am also from trip advisor. we used to do an online review. we confront this regularly. i would agree with todd. as the state of it is today in the united states, 230 is a terrific help to our business model and others here today. it is under threat from time to

time. there was a group of state ag's a year or two ago trying to claw back some of those restrictions. europe is more unsettled in terms of the regularity and the rules of the road there. it is a less friendly environment for the intermediary. >> nothing in particular you would ask of the government? nothing you are fearing it will do? >> on the u.s. side? no. internationally, i think clarity. the big ask would be what it looks like. it's unlikely that would be passed today. it may be a difficult road to get past internationally. >> maybe i can ask you, tell us where you're from.

>> i am at the republican senatorial committee. the republican party over the last 20 years has had an amazing resurgence, it is a great, golden age, as we all know. [laughter] >> and how much would an issue like the one we were just talking about be on the radar, whether for a campaign, or for policymaking, or is this kind of in the weeds? >> presently today is an issue that is largely to the weeds which is a benefit to the tech industry as a whole. tech policy is an area of remarkable consensus more or less, and as we have seen in recent policy debates on the hill, there are issues that are

forming unusual coalitions, so issues where you have darrell issa working, and that is to the benefit of my view of the tech industry that it does not become overly aligned with either political party. as you zoom out the lens on the issue of the right to be forgotten and these other issues, within the construct of how you set up this roundtable and you say what is the perspective in the future, i think it is an inevitable shift as society changes their perspectives, so-called digital natives rather than the 30th 40's, and 50's, and frankly we have a different view on things like social posts, search information, and the rest. the baby boomer generation immigrants if you well, perhaps will be more concerned about some of those issues than digital natives will be ultimately in the end. >> we are hearing about the lifestream people, they are not

hearing you. >> they are not hearing me? apologies to the lifestream for the distance of the mics. got it. let me ask a question knowing we have a good friend from facebook at the table. here is just a quick question -- suppose we have some form of unrest, a la ferguson of the last few weeks, and there are posts on facebook about it, and facebook has an opportunity in the interest of public safety to decide whether to subordinate posts, that's a something like let's meet at the corner of x and y and really show our rage maybe facebook and say well, we will escalate a little bit video of a cat.

[laughter] >> they might do that anyways. [laughter] >> let me turn to joel kaplan of all people so he may express his outrage that i would even ask the question. >> yep, i'm just recovering from the shock that my friend, matt would suggest that. joel kaplan. i'm the vice president of u.s. public policy for facebook which was a small startup social network with a goal of connecting the world, and hopefully by 2024 we will achieve that. it is a great question, and i think it does, it extends out the conversation about intermediary liability in a way that i think touches on some of the broader issues that we see not necessarily the united states. i think i agree with the earlier comments that u.s. law is actually pretty good as a result of section 230 and other

efforts, but we do see in other places around the world efforts to hold internet companies responsible for the content of the people who are posting it, user-generated content. one of the places where it manifests itself is when the citizens are using the internet, not just facebook, but other sites, to communicate their dissatisfaction and potentially plans for meeting, and i can turn into unrest. at a summit you see governments around the world i think very conscience of and increasingly conscious of after the arab spring in 2011. >> what is your baseline way of dealing with that? is it analogous to adam's original, as google, google's original view of it is the internets, if you have a problem with it, go to somebody else

and facebook is saying look, we have a secret sauce that organizes what posts rise in the feed. we do not tweak it so much, it does what it does, don't blame us. or is their licensors or to do exactly that in the interest of the customers or the public good? >> i think you are conflating two issues, which understandably, healthy algorithms for the newsfeed operates, and whether facebook will be responsive to a government that wants to crackdown on dissent. >> right. let's talk about the first one first, which is wholly on the absence of government wanting to do anything. with facebook on its own, the way trip advisor might decide there has been a change in ownership, we are going to make a change to our algorithm -- there is an acceptable use policy.

>> the algorithm that facebook uses to determine what to show in an individual's newsfeed is a somehow or effort to show that individual the information we think is most relevant and interesting to that person, and that is a constantly evolving determination, taking into consideration lots of other factors. so, i mean, i don't really view that as facebook -- i do not think facebook would view that as its response ability to determine what the people who are using its service ought to see. >> now this is getting into my oath as a law professor to ask the hype that all follow up -- there is some video from the islamic state that if other users of facebook subscribe into the ideals of the islamic states want to see, you and reserve the right to intercede? >> again, i am trying to disaggregate the question of how often the algorithm will show what is on the site, period. we do have community standards that exclude some contents, and mostly that stuff that we view as causing direct harm, so if somebody is directly inciting violence on our site, that is precluded under our community standards.

>> where i'm hearing you saying as it is on us. it is not that it goes down in the feed -- it is just gone. >> if it violates our community standards of inciting violence or otherwise causing direct harm or other things laid out in our standards, it will be precluded from being on our site. the way we enforce that is through a reporting mechanism, so in most instances we are not going out to police that. we're waiting for somebody, one of our users, to report that. >> you would not use the feed algorithm as a remedy, you would only use block it or not block

it, take it down or leave it up. >> yes. >> got it. you said there was a second point about government and what it might demand, and facebook of course is a worldwide phenomenon. >> we are obviously a global service, and we have to abide by the law of the countries in which we operate. so in addition to our community standards, basically, we will respond to demands from government if they conform to the laws of that country. now, we will not typically take something off of our site if we think it does not violate our community standards, but we may i.t. block it in the jurisdiction of the area. >> roughly but not exactly analogous to google implementing a right to be forgotten within the european portals but not elsewhere, and then the user will see a message -- "this is not available." >> right. one of the things we have done as have a number of companies

around the table is publishing government request reports, the purpose of which is to share with our users the number of circumstances in which their government has asked us to take something down. that gives the people who use our service a way of understanding the extent to which their government or some other government is insisting that content be precluded from what we have seen. >> can i jump in actually? >> yes, macon. >> i am macon phillips. i used to work for the organization that governed the united states before the great facebook -- [laughter] i work at the state department now. i joined it about a year ago and prior to that, i worked at the white house after the president was elected in 2008. it occurs to me that i'm in a room with people who can answer this question that i have had for a few years now, and hearing matt talk about how tech policy

is in the weeds right now and it is a good thing. i recall the great sopa and pipa debate of 2011 or whenever that was, and being in the white house and seeing this issue develop and how rapidly it developed in the public space once the internet companies started talking about it to the users. and as someone who has spent a few years in internet advocacy -- and advocacy generally -- it was something i had never seen before. to go from five miles an hour to 100 miles an hour. google put a link on its page, once expedia had a link at the top, and it occurred to me that these large sort of membership-based organizations that can rapidly communicate to millions of people and frame an issue where people who had no idea what it was before they saw it on this page had this great new power that could be, you know, used -- advocate for

policy positions when it is presented to people by companies -- they are advocated by corporate incentive, even though they say they are always looking at the customer first. and i think it is good. but being on the receiving side of it, it was remarkable to me to see the power that large internet companies have when they decide to present an issue. >> is there a question you want to ask? >> the people who are probably making those decisions -- are we ok talking about the users? when we talk about transparency reports, do we want to tell people how they could, you know, express their views on whether they like this or not? how would you advocate your responsibility of encouraging accuracy? >> i think we had someone here from reddit, ok, we have a campaign manager from reddit.

>> we are a public policy trade association represent public policies. if you are watching in 2025, we are also doing drone policy. [laughter] >> a bit too late. >> i think that is an interesting question. looking through the political lens, all politics are local and the approach we take with issues that we bring, we want to make sure that elected officials see our companies and our users as particular interest. if you're a senator from a big agricultural state, you care about the farmers and the crops and you may not have a google or yelp based in your state, but all of your constituents and voters care about the internet and they care about internet issues. they are using facebook, google, trip advisor, yelp, other services, and they care about what happens. that is what you happen with sopa and pipa. they are going to speak out.

>> macon is saying with an association like yours, you have a usual lever to pull. there is a special lever, "break glass in case of emergency," which has changed the homepage of google with the shroud around the wikipedia entry, and the users here, and i hear macon asking how frequently, under what circumstances will that lead? >> the great thing about the internet is the user is king, so it is not always top-down where the user is saying you need to care about this, it is what matters to users, and competition on the internet, unlike probably any other part of our economy, to click away. it is easy to click from one site to another, whatever is interesting to you, so our company and our industry is very responsive to what the interests of the user are.

>> trying to think what might be the next issue where you way to break glass and pulled lever and rally directly -- would it be over government surveillance? >> i think that is an important issue. joel made a good point on his transparency report, something that users care about and our companies and our industry have been leaders on shining the light on government practices and standing up and defending their users. the platforms are all global, so it is not just by users in the united states but about users around the world, and it is something that out think our companies have shown real leadership on. >> yes, lorelei. >> i am going to give an institutionally -- >> tell us who you are. >> my name is lorelei kelly. i am with x lab.

we are looking at building the next generation of platforms for public problem-solving. i work on congress and how to -- [laughter] here we go. who watches "house of cards"? congress is not organized enough to be that awful. it is an old jalopy with the hood up right now trying to drive on a modern highway. >> it is just a country legislature. [laughter] >> one of the things i am delighted about this meeting right now is to bring technology, policy, community, especially the commercial interests, into a conversation about long game policies, and not just showing up -- >> what would crystallize the ask you would like to make to this group? >> i would like to ask that you invest a new kind of, not necessarily think tank, but a nude kind of knowledge brokering, in a support system

with a neighbor like technology for decision-making in the policy arena, which is not crowdsourcing, it is much more curation, much more showing up at the right place at the right time. all information is not created equally. right now, if we do not figure out a way to flip the big data revolution into a competitive political constituency for evidence, we are not going to have a legislature that makes policy based on the best knowledge available, for example. >> let's just unpack that. the big data revolution, to be able to be used to produce evidence, you mean by that thanks to big data, there is lots more we can know about the world and about people, not just what kind of cat food they are likely to buy -- >> exactly, but they are not necessarily on rents into the policymaking process right now that are useful or that show up

at the right place at the right time. i mean, let's face it, the language between california and washington, d.c. is crazy, and hackers are artists. they are looked at as criminals in d.c., disruption of a business plan and san francisco, a national security threat in washington. there is a huge effort to simply apprehend -- >> is there a classic case study where you can imagine this being applied, just to be concrete? >> i was on the hill for 10 years, and we did not even have lcd screens or really basic technical compositions inside committee hearings. what if we created a way to create data, predictive modeling or probability modeling or context modeling, inside mark ups while they are voting on amendments? so you have a way to hold members accountable for decisions they are making in the moment about the outcomes for

society at large. there are all kinds of ways you can enable -- >> if i put the word "not" into this sentence, this will tell me what will change -- this is how many people -- >> it is very important to say right now it is not the quality -- congress has a huge data quantity problem. let's face it. but we do not want to get into the fight about the credibility of big data as much as the quality of it, and i am fine with co-locating the data if a committee chair will listen to it, which is from their district. it is that kind of mapping that gets to the -- it just gets into the inbox. it is really not that complicated. it is that this community, i think, is thinking that information is the answer. it is not. >> let me stop you there and turn to bruce schneier.

bruce, i want to ask you crisply first, after introducing yourself, how worried should the average internet user be about government surveillance in particular in 2014? use whatever unity think appropriate. second -- do you have a view on whether there is a way to deploy new technologies, new ways of sorting and searching new information to help this broken jalopy of congress? >> i am bruce schneier, i work in security and technology and policy at a lot of these issues. this is a comic it a question, and to be crisp, i think you have to be very worried about surveillance in general, whether it is government and corporate. because it is a new way of organizing society where everything is recorded and used and reused and saved, and the policy implications of that really have not been brought through.

we have kind of backed through carrying a cell phone or using a search engine without really thinking about what the ramifications are. policy fixes i think are going to come not directly but from the side. we do not really see appetite in congress, really in many countries, to attack these countries head-on. europe more than the united states, but even so not much. i look toward some of the regulatory agencies, some of the ways you can get policy, which is informed by technology, into action that does not require legislators. i tend to be near-term pessimistic and long-term optimistic, and i will leave you tantalizing at that point. >> a lot of the companies present are on the table are themselves on behalf of their users newly sensitized to maintaining a certain distance

from government, securing user data. i am curious -- is there anything, any question you will make around a table, or is this pretty much the issue and they are working on it? >> they want to do it themselves. google is not saying, "we do not like surveillance." google is saying, "it is our job, go away." we need to get to the point where users want business models that don't have surveillance. when duck, duck, go surpasses google in traffic, we know -- >> policy or something like it is adopted. >> or you go in through policy. we are at the point where technology can become policy

and policy can become technology. >> let me go to althea. tell us where you are from, and given the posture of the couple you are working with and the place in the industry, how much do these issues of surveillance, whether in relation to the government -- i am trying to figure out what sort of request the government would make of etsy -- we need to know where that scarf came from! [laughter] on maybe the corporate surveillance side that bruce was rather incendiary, striking the tinder on -- how much are you concerned about user rights within the commerce that you facilitate? >> i am with etsy, an online marketplace, and for us the government surveillance issue is not as you implied much of an issue, we just do not have government knocking on our door

all that often, which is great. it lets us not worry about that as much as up on the user privacy peace, for us it is really a matter of building and maintaining trust with our user base, so when we are making decisions about how to use data internally, we just got check ourselves. if i knew that was happening for me, would i find that creepy? that is important for us because if we use the data we collect about our users in a way that they -- that would lose trust without consumer base, they are going to go somewhere else, so that is sort of how -- we do police ourselves, i guess. >> i once went to hormel foods in minnesota and met their bacon taster, a guy whose job was to taste bacon all day long. >> great job. >> at some point he would go you know what, it is best before today, and they would stamp the package and be like, "now we know." etsy is like, "you know, that

policy, it would be best if it were not that way." >> right. >> maybe i should put it out voluntary questions before i call on someone, maybe from yelp or something -- [laughter] how broken, if at all, is the model of data gathering and usage that, in many ways, drives the free internet? maybe less so the transactional internet, but you get the free information -- somebody has got to pay for all of this stuff. advertising has a storied history back to free television. is this model -- you just need to adjust it to deal with the occasional creepiness, or is there something fundamentally worrisome about it? why don't i put it to you, laurent? >> sure. laurent crenshaw with yelp.

at this point, in 2014, i do not think it is broken. i will just put that out there. yelp is a platform that can next people to local businesses that are around them, so whether it be a restaurant or a dry cleaner or wherever, you want to find out what is best in your location for stuff you can going to yelp and look at user id's and content -- >> how does yelp make its money? >> through advertising. basically we get businesses to put ad packages on yelp, and if you are searching for a dry cleaner, a restaurant -- >> you have two lists, here are the good dry cleaners, and here are the ones that paid me, and they are marked. >> it is marked as an advertisement, so the user knows, and we try to do everything that we can to illuminate any sort of confusion there. at the same time, that is how we

make our revenue, through that manner, and i do not think there is anything wrong with it. in our sense, it is not about electing a ton of information about the user. it is easy to set up a yelp account, your first name, last name, initial, date of birth and an e-mail address. anything you want to put in after that is up to the user themselves. >> is the data you are collecting, i guess the targeting happens quite naturally. you are looking for the dry cleaner in this area, boom. are you looking on the basis of inference about a user rather than just on a search? >> we want to build to tell the business and providing most useful information about who is going there, so it is a matter of how many people who might be going to that page. >> what demographics would you give in to say old people really like your dry cleaning, maybe young people not so much? >> exactly.

various metrics saying the vast majority of yelp users are college graduates, most of them are in the 25 to 50 range and make over x amount a year, and also generally speaking have been able to sort of show that within a period of time, like within a week, you want to go to yelp to spend money, to buy something, so those are examples you can take to business owners, why you should advertise on our platform. >> leigh, from the point of view of your company, we are about to hear about, basically plus one or a difference from what laurent just described? >> thank you. my name is leigh freund. i am the head of public policy for aol.

we make our money primarily from advertising as well, and i would add to what laurent said. we talk a lot about privacy at aol, and i think joel talked about transparency in terms of government requests we get. transparency for users is the way we get our trusts, so in our privacy policy, we try to make things easy to read, we try to get people options if they do not want to be targeted on advertising. >> the privacy policy? >> it is much shorter as of september 15. >> a provocative answer. all of the fat has been removed from the privacy policy. >> legally, right. facebook, i read through privacy policies on facebook and yahoo!, and we all try to do the same thing because we give people information about the type of data we collect any

type of data we use, and one of the things i think that is an important consideration is the way we the sting was having personally identifiable information, like "i am michael beckerman, i live in washington, d.c., here is my social security and credit card information," and "i am a user who lives in washington, d.c. who searches golf equipment," hypothetically. we use the ladder in an aggregate form. if i were to use personal information to target michael beckerman for an ad based on cliques and the millions of impressions that we need for advertisers, it would not be successful strategy. so aggregating all of the information laurent was talking about in all of the information about user behavior to serve them targeting advertising is i think something that a lot of the companies around the table do, but one of the things that we try to do in a reasonable way is give transparency to our

users and choices about whether or not they want to be targeted. you have got an opportunity to opt out of that kind of data collection and targeting. and i think that there is always a balance in our industry -- >> i hear you -- >> between privacy and free internet.

it goes back to what she said in terms of if our users don't trust us they're not going to continue to use our service. so we are always trying to ensure that we are doing something that the users are going to be comfortable with. one of the things the companies are doing is starting to -- i won't necessarily say exeet on privacy but people are starting to see privacy as a way to differentiate their product. and to talk to users about privatesy more directly. so one of the things that you are seeing a lot of is contextual privacy notices and trying to make sure that when you're on an ap you're seeing what's happening on that ap not what's happening on the yahoo website. when you're interacting with an advertisement you're going to see privacy notices about the kind of data that is informing that advertisement. when you're seeing news or sports teams you're seeing on our sports website you're seeing your fivert football

team. you're going to -- there's going to be a way for you to understand why you're getting that information that's in context. i think that's one of the areas where you're seeing sort of development and trying to make that communication more clear. >> let me ask you in the point of competition, which is in a way, getting back to duck duck go is there and if people want it and enough use it others would model that kind of private policy. you're suggesting that here as well. to what extent if i'm the average internet user might something i'm doing on yahoo implicate something i see on a.o.l. or on yell p? are there ways in which the information about me that facilitates advertising moves among the websites? do you glean from your users

and those are not mappable on another site? >> there are circumstances where what is happening on one site is being used on other sites as well. that sort of mapping is for yahoo! always disclosed, partners we work with and have relationships with other websites. that is clearly disclosed in our privacy policy. when it is related to advertising, you can click and see who the partners are. we try to make that clear to users because we understand that is something they would want to understand. i cannot speak to every other internet company and how they handle relationships with their parties. >> on the regulatory front, do you see anything coming from government? is this stable? the jalopy is not chasing you? >> i don't think the jalopy is chasing us.

i think it is something because so many people use technology and are seeing the ways it is being integrated into our daily life, i think it is something everyone is paying a lot of attention to. there is a lot of attention on it and a lot of oversight, which is appropriate. whether that will lead to legislation or regulation in the immediate future, i am not sure. i think the dialogue and oversight is important. you're seeing a lot of companies work together on regulatory efforts. i think there are best practices and things that will move forward maybe more quickly than regulation or legislation. it is certainly an issue we are all focused on. >> in the future, i think we will see this change from a person-interface site world we are living in and moving to the