 Good afternoon, everyone. My name is Marco Ramis and I'm an Associate Professor of the School of Public Health and Health Systems at University of Waterloo. I'm also an Associate Scientific Director of the Canadian Longitudinal Study on Aging. And I, along with health from Laurel Austin, Val D. Pietro and now Sue Johnston, and I'll put you together these monthly seminars for CLSA. So, I'll stop talking and just one note quickly. Because we often get feedback, we tend not to use the talk feature. It should be disabled. But if you have questions, please use the chat feature to type in your questions. And when Mike has finished his presentation, I will relay your questions to him. So, our talk today is going to be called The Income Distribution of Canada's Seniors. The top end and everyone else. And I'm happy to introduce our speaker. Professor Michael Viel is a full professor in the Department of Economics at McMaster University. And his main research interests include applied econometrics and economic statistics. Some of his focus recently has been on the distribution of income in Canada and tax transfer policy. He collaborates on projects involving productivity growth and the relationship of health and economic status. He was president of the Canadian Economics Association between 2011 and 2012. And he has also participated on the CLSA's social working group. So, Michael, give the floor over to you. And thank you very much for agreeing to participate and give this seminar today. Thank you very much, Mark. I will say first that I'm going to be talking largely about joint work with Brian Murphy, principally, and some with Hung Pham of Statistics Canada. And I thank them for their permission to do this. And I also have to disclaim that neither they nor Statistics Canada are responsible for any errors or any opinions in this presentation. I will note that the data I'm about to use or present is actually going to be updated probably within a month. And so already this presentation will be updated by the end of this month. Some of the data is to 2011 and some is to 2010. Because it makes so little difference, I will not be all that precise as to which, but if people have questions for me afterwards or want to contact me afterwards for some precision, on that point they can ask me. I want to point out that Statistics Canada now includes estimates of top shares in income distribution using tax file data, but cancer does not yet include information about the top end income distribution by age. So this is what I am trying to fill in a little bit of a gap here. For those of you who don't know about this data, it's anonymized tax file data. Statistics Canada is called the Longitudinal Administrative Data Bank. It starts in 1982 and nowadays has about 5 million observations per year. It's a one-kiss sample and it's got pretty good coverage from 1992 on. What makes the early 90s special is they started to become more and more tax-delivered benefits. And so therefore people at the low end of the income distribution had strong incentives to join the data in effect by filing tax returns because the filing of tax returns had monetary yields for them. There are lots of shortcomings however in tax data. One of them of course is misreporting. There could be misreporting in surveys, but perhaps a different nature of misreporting when it's for tax purposes. On one hand there are penalties for misreporting. On the other hand there can be advantages. We do know about the other kind of economy. We do know about tax havens. We've taken rather a lot of work on small business and small private corporation income. Private corporation income in general and people can ask me about that issue later if they want. As I just indicated there is a filing bias although that mostly is gone now in the data. There isn't that much filing bias left because it's still of course some people do not file. We're going to focus on before tax market income. That includes wages and salaries and self-employment income, investment income, pensions, things like that. I'm not going to include capital gains although I can. Capital gains are bumpy and of course the year you realize a capital gain is not quite the same thing with the year in which you truly earned it. So we're just going to leave that complication aside for today. Again I'll answer questions about it. I've done a lot of work with capital gains as well. And the transfer payments are included. We're not talking today or I'm not going to talk so much today about the welfare of the older population but I'm going to instead emphasis on their earnings process, how they participate in the labor market and other markets which earn income for them. And the other specialization is that I'm going to concentrate on the top 1%, which is what the Occupy Movement was about. I'm going to describe as we go why I think that's particularly interesting. So early questions you can ask me. First is the older population over or underrepresented in the top 1% of all income recipients. And the answer is you can read there is that they're underrepresented. There are fewer of them in the top 1% than you'd expect. Are older members of the top 1% different from younger members? The answer if you are in the top 1% and you're older you're likely to have a higher average income. Not surprisingly you're less likely to be earning labor income that is work less. But if you're employed you probably also earn less. So the big difference of course there is capital income. I'm not going to be talking so much about that. Instead I'm going to be thinking about the older group in itself. In other words rather than thinking of the role of the older population in the 1% of the whole economy I'm going to be talking about the shares within the older population. And also other age groups as we go along. So here's a first table and of course there's going to be lots of numbers. I'm going to try to make it not too daunting for you. But those numbers give you the idea of what it's like to be the top 1%. So the first thing you can see in the top left-hand corner is that the threshold in 2011 to be in the top 1% was $208,000. So if you were in $208,000 you'd be in the top 1% overall. But that number would be lower if you were 66. So you can see those two numbers are just circles. And so it's a little easier to be in the top 1% of the 66 plus population than in the overall population. Then the rest of those numbers you can see they follow a probably unexpected pattern. If you're thinking about the 0 to 35 age groups, some filers could be very young. And it doesn't take so much income to be in the top 1% over $127,000. The place where it takes the most income to be the top 1% is the 36 to 55 year old group. And then after that you can see it tails off more or less smoothly with age. If we go on to the average income for the top 1%, 5% and 10%. Now we're talking about not the amount to be in the group, but the averages of those that are in the group. And you can see again that the average income of the top 1% in Canada is $441,000 in 2011. And the average income of those 66 plus is $403,000. That is at the top 1% of all of those who are 66 plus is $403,000. And you can see the rest of the pattern more or less looks the same as we described before with a couple of bumps. But basically it is a little higher in the 50 to 65 group now, but then it starts smoothly declining with a couple of bumps. So let's look at these other ones. Basically you can see it gets to a point in all of these age distributions near the bottom where there's actually not that great a difference. So there's not a big difference between the 66 to 70 year olds in all those numbers than there is between the 80 plus, even to some extent in the lower 90 although that's perhaps the biggest thing. Notice of course that sharp contrast when you go into the lower 90 at the far right, how much the numbers fall off and that's something we're going to be talking about more as we go on. Now let's shift to consider top shares within different age classes, i.e. distribution of income by age group. So we're looking at the distribution of the wages. So you can see the top 1%, normally have pretty good labor market participation. When you go to the 66 plus of the top 1%, their labor market participation is only 56%. That is, I shouldn't say labor market. This is two T4 incomes. This means employment income as employees, 56%. But still it is the case that the top 1% and the top 5% have much higher labor participation rates than say the bottom 90%. So in fact, that's the numbers I'm going to concentrate on. If you look at that 66 plus age group, you can see that there's a huge difference in labor force participation, employment income between the top 1% and the bottom 90%. That is only 13% of the bottom 90%. That's 9 times the population, obviously. Only 13% of those earn any T4 income at all. Sorry, I think I just went the wrong way. So let's try to fix that. Okay, how many are self-employed? And again, you can see that same kind of pattern although they're lower numbers. Of the bottom 90%, only 6% are self-employed of those 66 plus. But of those 66 plus in the top 1%, 27% of them are self-employed. How many receive investment income? Well, again, there's a sharp contrast there. You can see of the top 1%, 93% earn some investment income and only 53% of the bottom 90%. How many receive private pension income? Well, there the contrast is in some sharp, isn't it? You can see, in fact, that the top 1% aren't the biggest holders of pension income. So this is largely not a pension story, the top 1%. That's not where their income comes from. So before I go on, I'm going to tell you a few sides here. I think I framed some of my major points. Top 1% and 66 plus receive 20% of all market income. So that's in fact more concentrated than for the all-age distribution. It's only 12% for everybody, but 20% of the top 66 plus. So market income is concentrated in fewer people. In fact, if you go to wage income, you can see it's really concentrated. 43% of all 66 plus wage income is earned just by the top 1%. Comparable for all the rest of the population is 11%. So you can see being a major wage earner in the top 66 plus age group, being the top 1%, there's a much more concentrated thing. It's a much rarer thing. And then you can see, if you look at investment income, it's 32% of all investment income. So in fact, wage income, labor income is much more concentrated for the 66 plus population within the top 1% than its investment income. And before I knew these numbers, that would have been my guess. And then, consistent with the slide I just showed you, you can see the pension income is not particularly concentrated in the top 1% of those 66 plus. So I've shown you a lot of numbers, so I'm going to take a little bit of a break and tell you my conclusions so far before I show you some more numbers. Top share inequality is partly the not solely a life cycle effect. You can see some things that go through the life cycle, but that's not everything that's going on. Senior top share inequality is labor income inequality. Even more than its investment income inequality. So that's that last slide I showed you with labor income being so concentrated, even relative to investment income. A bunch of senior labor income is concentrated at the top end. And that's interesting because that's at levels largely beyond transfer system disincentives. Okay, so if we think about things like the GIS system or even the can attention plan, I'm sorry, the old age security clawback, most of those disincentives have long gone away before we're talking about the top 1% where a big chunk, almost the majority of labor income is earned. And so those disincentives really don't have much to do with the majority of labor income that's earned by the older population. And then there's a tension between intergenerational equity and sensitivity of senior income to taxes. The trouble is that when we estimate these things and I've done some work doing this is we find estimates which suggest that senior income labor participation and supply is much more sensitive to taxes than everyone else, essentially because they have a retirement decision. And so when we think about perhaps some people's idea of a quote remedy, unquote, to this distribution of income being so concentrated at the top, if you go with the tax route, it might be that you in fact discourage a lot of that earning. So at some level you'll reduce the unequalness of the distribution, but the main way you'll have done that is by lowering the amount of labor income earned by the older population as a whole. And then finally, as I mentioned before, your private pension income is not spectacularly large to the top end. So I'm going to conclude by talking about some trends. And so at least we get to switch to pictures now. So here's the trend since 1982 when this particular data source started of average market income. And you can see the top 1% for all ages and the top 1% to 66 plus. And I think the main story there is particularly when during the period where we started to be a little more confident in the quality of the data in the early 90s, those two pictures are basically identical. The only thing is the average market income of the older population is a little lower and that is basically because they earn less employment income. But it's not a different story and you might even not notice those two lines of the bond for all ages and all fires, five to 66 plus. You can see the market income is pretty much flat and low during this period whether you're an all father or you're 66 plus. That's a very stable trend. During a period in which the top 1% income actually went up quite substantially and this is in real terms. So this is something that indicates that there has been a trend over the last 30 or so years that has led to a higher concentration income both for seniors and for the general population. So aside, why might this matter? And I'm not, this is not the talk in which I go in that direction but just to make a couple of general observations. When I think of this for the general population, I personally frame the issue is not so much that I'm worried about the affluent getting richer but that the less affluent are. In fact, I don't have a real problem with the affluent getting richer but it's just the unequalness, the consequence of the unequalness means that the lower income people are not getting big increases and I think that's a matter of concern. The obvious big policy question, which I'm not answering is the reason the affluent are gaining is that at the expense of the less affluent? Is that the reason for this or are those two things not associated? I won't answer the question today but I will say that I think to some extent I believe that the less affluent are not doing as well as they might because of institutional factors that favor the affluent but again, that's not my main theme if you wanted to come out in questions. As you can read, I think the biggest danger from top end inequality and not particularly senior inequality is probably through the political system. There's much writing on this topic but the basic idea is as the elites in society become more powerful, they start to set the agenda and they set the agenda in ways that tend to favor them. In the senior case, I don't think that this is really and us versus them in the conventional sense but I do think it is the case that sometimes high income seniors are more influential in the way you think about problems. So one of the problems we think about at the high about seniors of course is their participation in society in different ways, their ability to keep in the labor market, things like that is being very important issues and no doubt they are but we shouldn't let those issues dominate ones that are probably more important than lower income seniors who already have basically made the decision of not participating in the labor market where they are more interested in income support and whether they can have a higher standard of living. For people who by and large are not what we in Canada classify as poor but are really just a little bit above groups that we usually call poor. Now we go back to income source and now what I'm doing is I'm comparing 2011 and 1994 and there are the percentage differences and this is all seniors not just top 1% seniors and you can see the big trend during this period for all seniors has been an increase in self-employment income that's the middle bar there. Rages in fact has gone down so people are participating less in the labor market rather than more over this period. But there is that big self-employment number. Investments in fact have gone down a little. Now that I think is just an artifact of the fact that these numbers even though they are put across the dollars they have to reflect the fact that interest rates in the early 90s were much higher because inflation was higher and now interest rates are lower and returns on investment are lower and just adjusting for the CPI won't fix that up so I think that's the main reason for that. I skipped over pensions and you can just go to see RSV withdrawals which I'm mainly mentioning now because it shows up big in the next graph. You can sort of see that not too much has happened there fewer RSV withdrawals, a little bit more money we'll discuss that point a little bit more in the next slide. So the next slide just goes with top 1%. Now if we go back and forth between these two slides one of the things I want to point out is the scales change. So the numbers on this slide are all bigger than the numbers on the previous slide and you can see they're mostly big green numbers and they're big positive numbers and big changes. The top 1% earnings have gone way up and that's one of the reasons that the income distribution has become more concentrated and that's partly wages and that's partly self-employment income. The self-employment effect is not that much bigger than it was in the previous case but the wages effect is way bigger and positive. The investment figures are positive too. I think I have some of these circles. Yes, they're all circles because they're all big numbers and the pensions number even though pensions is not all that important it used to be even less important and pension numbers are a lot bigger for the older population. Now the RSV withdrawals number that's a little bit artificial and that's got to do with remember we're going back to tax data. It turns out that the nature of the tax data is such that this isn't getting all your RSV income. Some of that is in fact embedded in pensions and there's different ways to take out and what's happened over time is more RSV money comes out this way than it used to be so some of this is not a pure trend. It's not just the fact that high income seniors have got really big RSVs now that's part of it but that's not the whole story there. That's partly an artifact which is unfortunately a shortcoming in the tax data. Two more quick trends which I will not go through in much detail if you look at labor market participation you can see that the blue line at the top that's the top 1% all ages and basically you can see that pretty flat number for all ages if you go down to the red number that's the top 1% in 66 plus maybe it is increasing but the change in participation rate isn't really the big thing that's going on it's the wages that are received by those participating in the top 1% that's changed a lot that's the big number where meanwhile you can see the green number that there's been a slide in labor force participation for all filers over this period and the data hasn't in the case of 66 plus population so it's gone up a little bit and then finally there's a slight trend toward having some private pension income but it's not there for the top 1% even though the pensions that the top 1% do receive are a lot larger it's not largely a participation effect so this is my final round of conclusions senior income is labor income and investment income and both have become more concentrated in the top end and that's increasing concentration over time high senior labor participation higher growing for both the top 1% and for the senior population as a whole a well-known trend that lots of other people have documented pension income is not so important in the increasing concentration it's not the reason it's not big pensions rather it's big continued at least what is reported on tax for tax data is big continued labor market participation and some things going on in the investment side and again some of these patients are tax transfer policies so again we economists spend a lot of time thinking about the implicit taxes on income that come from guaranteed income supplement and come from say the old age security clawback and programs like that and my point here is that for 42% of senior income and probably a lot more those clawbacks are completely irrelevant because these people in the top 1% are way above the range in which those clawbacks make much difference and so the great bulk of the income those clawbacks don't have much to do with that so that affects the way I think you think about those clawbacks and how they should be implemented but that is my last point I believe I think all it remains to me to do is to thank you for listening to this and I appreciate all of these numbers I noticed a lot of familiar names on the sheet and if anybody has any questions I'm most happy to answer Thank you very much Mike for your interesting presentation for those of you who may have missed the directions at the start of the presentation I had asked if you have any questions to please type them into the chat feature and I will relay them to Mike that's just because if everyone has feedback we can't hear anyone so any questions please type them into chat I'll begin with one question verbally for Mike and that has to do with some of the income assistance programs that we have in Canada such as old age security et cetera et cetera do those programs have an impact on how can I say this do those programs make a difference for seniors when it comes to their income and has there been any work looking at how those income assistance programs might have an impact on seniors health because you have one camp if those programs are in some way cut or the amount of money that's distributed to seniors those programs is reduced one camp is going to claim that it's going to be a huge disaster I'm just wondering if there's been any real research looking at how those programs really impact seniors especially with respect to their health okay I see Harry has a question I think it's going to be related to the first part of my answer and so I will fold that in a little bit and then I'll answer Harry's question as it comes up so first I want to talk a little bit about what we know about transferring system and income to seniors and then I'm going to I think answer your question more directly about what we know about the health effects which is not my own research so I've done other research about the other end of the distribution and basically the way international comparisons are done rightly or wrongly in this context is they look at what's called well various sorts of low income measures which are usually 50% of the median income for the population and we see whether say seniors whatever group we're interested in what their income is as well as that measure which is set for the population as a whole in this case say 50% of the median income in this low income measure well Canada's transfer program the GIS has been set in such a way that it gets seniors very close to the low income measure and for a while it was a little bit above and now it's a little bit below and then you get some wobbling around of the poverty rates the measured poverty rates which are the rates above and below this measure because of that but still even fairly recent things seem to suggest that Canada has low poverty rates for the senior population and that's largely because of the guaranteed income supplement now the reason I mentioned this in the context of your question is just to say that there's a good chunk of the lower part of the income distribution which would be there without that program that's just pushed up just to that threshold but not much more that's a fair a fair number of seniors maybe a quarter of maybe 30% of seniors are more or less in that state now there are some below that range and that's largely because for whatever reason it's possible not to qualify it for the guaranteed income supplement it's part of just probably measurement error but various measures have suggested that there's not that many who do not even know quantitatively it's a large number but it's a percentage it's probably something like 5% so that means we've got a lot of seniors who are up to that threshold of income and of course there's publicly provided healthcare compared to some other countries although most countries have similar systems to Canada in that regard and so basically at that level we think that using that standard of poverty there are not that many seniors who are below it and that they have access to healthcare so therefore where does that come across in terms of debate seem to impact their health status and I think the evidence by and large is that there's weak evidence suggesting that it does positively sustain that health status the one place where I personally done work is I've done some work on drug plans and there the differences don't seem to be so profound some provinces have quite good drug plans for seniors and some have not so good and there doesn't seem to be a lot of healthcare status variation due from that but I think to some extent it's probably trumped by the more general population healthcare programs and by the minimum income systems provided to a guaranteed income supplement so that doesn't answer your question fully but I do think that there is weak evidence in Canada that suggests that the income programs have translated into better health outcomes for seniors other people on this forum there was the question from Harry recognizing the data sets are different what are the data for other OECD countries and so unfortunately I haven't done the seniors alone for other OECD countries and perhaps I should have or know more about it so I'm sorry Harry I can't answer that question very well I will say for the general population we know what the data looks like and the concentration of income in Canada is basically higher than it is anywhere else in the United States but there's a lot of places that are more tied with us and they would include Australia and Ireland and New Zealand and the United Kingdom and these are commonly described by and their work as the English speaking countries perhaps not entirely the best use of phrase I would say perhaps however it might have something to do with the British system of law which has governed all those countries but it's clearly true that measured income concentration is much higher in those countries that you did say in the continental European countries much lower concentration in France and Spain and Switzerland Germany there's a question about the quality of the data that I don't think has yet been resolved so Italy and Switzerland I'm not sure I mentioned as well as you'd expect the Scandinavian countries less concentration income at the top and the other countries that we have a little bit of information on which we're not sure how much to trust include countries like China if you include those in the pack that they have more concentrated income than Canada or the United States a few of those countries do but I think of the countries where both the US and Canada have broad economic system similar to ours Canada has relatively highly concentrated income I do not know but I expect that to be the case also for seniors income great thanks we have another question have you had a chance to look at individuals from a longitudinal perspective for example who leads or reenters the top 1% or the top 5 or 10% over the years in other words there's a long variation well again I'm sorry I haven't done this for the older population I've done it for the general population the general population I've done it pretty thoroughly it's an actual thing to have done it for the older population I haven't so I can just quickly report the general population result the general population result was I think not as much as you'd think or at least not enough to offset the trends so for example suppose instead of thinking about year incomes we think about incomes for 3 year periods or incomes for 5 year periods if there's a lot of going in and out you might think that those ones would have not this big surge of top end income that we observed in our data for the general population and the answer is you observe almost the same surge and it's also the case that the probability of what's being in the top 1% now compared to 10 years ago or 20 years ago or even a little bit more the probability of being in the top 1% and staying in the top 1% is in fact higher now than it was at previous times so it does appear there's a fair amount of persistence it does not appear for example but some economists might tell the story but all that's happened is we've got more developments in the general population and some people earn bonuses one year and some people earn bonuses in another year and really it all works out there's no big change it's just when you look at an annual basis it looks like there's been a change but that does not appear to be the story the story appears to be that there's been a permanent shift I will however say that some people also argue that the 2008 recession and the time after that was the time at which the top end of the population were doing well at the expense of the rest of us and all the data suggests that that's not true but in fact their incomes proportionally fell more after the great recession in 2008 than those of the general population still with significant top end concentrations great we have another question how do you think the baby boomer will change income distribution at the top? okay so this I think goes over to Thomas Piketty because we're now thinking about the forecast and his famous forecast is that the income distribution is going to become more unequal largely because what we've been observing so far has been a big increase in wage income concentration so much of the reason for the patterns I've shown you can be traced to higher labor income either received by the general population currently or by the retirees during the period in which they were in the labor market and Thomas argues that there will be this shift over to capital income partly because the baby boom will continue to have this unequal labor income distribution that will more and more translate into the amount of resources they have when retired and also because he argues that what matters in this context is the rate of interest which is the return to capital relative to the economic growth that's gone really high return to capital and that will just improve a lot of capitalists relative to the rest of the population more and more I think so Thomas is in effect predicting that the rate of return on capital is going to go up that right now we have this unusual period of relatively low rates of return on capital and it's back to go up and he's a pessimist about growth he thinks that the rate of growth is going to go down I'm not so sure about the rate of return on capital but I agree with him that unfortunately I do believe the best forecast is that the rate of growth for the economy is going to fall partly simplistically because part of that rate of growth is the number of children we have I do not see that trend changing greatly and the other reason is that there does appear to be at the moment some sign of diminishing returns in the innovation process that's leading to lower growth we're getting the fast growth economies like China and India are getting closer to us and there may be some reason to suspect that their rates of growth will start to fall and that not only will hurt them but that will probably hurt us as well so for a mixed set of reasons I think the baby boomer will face income distribution more concentrated at the top both for the general population and for the older population still more concentrated than we have now but it is a forecast and it is just that I do not have strong statistical reasons to believe that, that's largely based on the book and the parts that I thought were great thanks for the question for me have you looked or are you aware of any work that has looked at income distribution differences between Canadian born and immigrant seniors I'm just reading some data here that immigrants account for about 28% of all seniors in Canada and I'm wondering if there might be differences in immigrants and non-immigrants in terms of their income yeah they actually do a little better in terms of market income and the reason for that appears to be that they do worse in terms of transfer income because some immigrants don't qualify for the full range of transfer income so when I did my study now getting a bit dated 7 or 8 years ago and found that using that low income measure 6% of all seniors who are below the low income measure about a third of that 6% were immigrants and the reason for it was that they did not qualify for the guaranteed income supplement I see great are there any other questions from any members of the audience now is your last chance to type a question if you would like to have a question or even if you want to make a comment about some of the issues that we've discussed over the last little while please go right ahead people can feel free to send me an email if they have any questions later or things that they wish to talk about great that would be fantastic get in touch with Mike if you have any further questions and that will then conclude our seminar for today Mike thank you very much for presenting this interesting discussion this oh no sorry I thought we got a question it's complimenting you thank you this was super interesting indeed it was thank you very much for presenting this marks the first time we've been doing these seminars for at least a couple of years now this marks the first time that we've looked at seniors and aging from an economic perspective so it was certainly quite enlightening because of that novelty aspect we've tended to focus more on health outcomes in these presentations so once again thank you very much Mike and just before we log off I'd like to announce our next seminar it's going to be on December 4th between 2 and 3 in the afternoon Eastern Time and Heather Keller a professor at the University of Waterloo will be presenting older Canadians food intake and nutritional status how the CLSA will advance knowledge so we look forward to that one on December 4th and once again Mike thank you so much alright have a good afternoon everyone thank you for joining us today we appreciate your support for these seminars have a good day