Inequality. A Closer Look. February 2014

Transcription

1 A Closer Look Inequality February 2014 Joshua N. Feinman Chief Global Economist Deutsche Asset & Wealth Management Tel:

2 A Closer Look Inequality Inequality is in the news, to put it mildly. The Occupy movement, the Pope stressing capitalism s ills, and President Obama citing the growing income gap as the defining issue of our time, offer just a few examples. And it s not hard to see why inequality has taken center stage. Few economic issues inflame the passions as intensely even in the best of times, and the years since the Great Recession have been anything but. With many people still struggling, and the benefits of what growth there has been continuing to accrue disproportionately to the well-to-do, long-simmering antagonisms about inequality and opportunity have boiled over. In this paper, we take a closer look at the extent, causes, and consequences of rising inequality including any effects it may have had on the economic mobility so central to perceptions of the American Dream. A few stylized facts more detailed measures aimed at specific pieces of that distribution. All in all, it s enough to make a person s head spin. Fortunately, we don t have to get bogged down in most of this detail to distill a few widely accepted, stylized facts about inequality. For starters, most studies agree that inequality has been on the rise over the past few decades (roughly since the late 1970s/early 1980s). Not every year, not without occasional backsliding, and more so on some metrics than others. But in general the trend has been toward widening gaps between the haves and the have-nots. Consider, for example, the often-cited Gini coefficient a single summary statistic that ranges from zero to one, with increasing values indicating widening inequality. 2 The Gini coefficient for household income in the US has risen since the late 1970s, suggesting increasing inequality (chart 1). 3 There have been interruptions, There are many ways to measure inequality. Some studies look at the distribution of wealth, which tends to be the most unequal, others zero in on consumption, where gaps are smallest, but the majority focus on income, where inequality falls somewhere in between. Data sources vary too, from tax returns to surveys of income and expenditures. The unit of comparison is sometimes the individual but often the household, with some adjustment typically made for size. 1 Different metrics are employed as well, from single statistics designed to summarize the entire distribution in question, to Figure 1: Gini Coefficient for US Household Income 1 It is common when comparing living standards across households of different size to divide household income not by the number of household members but by the square root of that number. This implies that, to secure the same standard of living, household income needs to rise with household size but less than proportionately (because members of the same household can pool resources). 2 See appendix for more details on the Gini coefficient. 3 Congressional Budget Office (2011). Trends in the Distribution of Household Income Between 1979 and Congressional Budget Office (2012). The Distribution of Household Income and Federal Taxes, Increasing inequality Source: CBO Pre-tax and transfers After-tax and transfers

3 usually during economic downturns, but the longer-term uptrend is clear. And the key driver has been increasing inequality in pre-tax income. Yes, the Gini coefficient has risen on incomes after accounting for federal taxes and transfers too, but by no more than on pre-tax incomes (and even a bit less during the Great Recession, when the safety net really helped shore up incomes at the lower end), suggesting that increasing inequality is not a direct result of shifts in tax or transfer polices. 4 Put differently, federal taxes and transfers continue to mitigate some of the inequality of pre-tax incomes the Gini coefficient has always been lower on an after-tax/transfer basis but not decisively more or less than they always have, and not nearly enough to have stemmed the tide of rising inequality. Taking a deeper dive into the income distribution reveals that increasing inequality reflects a more top heavy income pyramid. The share of total income accruing to all groups except the highest quintile has declined in each successive business cycle peak since the late 1970s (chart 2, which compares years near business cycle peaks to shine a better light on longterm trends, less obscured by cyclical influences). And within the highest quintile, it s the very top who have reaped the lion s share; indeed, the slice of the pie garnered by the 81st to 90th percentile has actually fallen slightly, that going to the next nine percentiles has edged up a little, while the infamous 1% have seen their share soar (all of these trends hold up on both a pre-tax/transfer and after-tax/transfer basis). In short, it s not that the middle-income groups have been pulling away from those at the bottom; it s that those at the top especially the very top have been pulling away from everyone else. These trends were temporarily interrupted during the Great Recession, when the pre-tax incomes of the affluent were disproportionately hammered while transfers helped buttress incomes of those on the bottom, but since then the longer-term trend toward widening inequality seems to have reasserted itself. In fact, preliminary data for 2011 and 2012 suggest that income shares have risen sharply again for those near the upper tail (though some of the increase in 2012 may be exaggerated and temporary, as higher-income people likely shifted income forward in anticipation of the increase in tax rates that was known to be taking effect in 2013). 5 Longer-term trends Several ambitious studies have attempted to peer even further into the past, looking at inequality back a century or more. 6 Relying primarily on tax records, they find a broad, u-shaped historical pattern with the share of income going to the top decile relatively high in the early decades of the 20th century, then falling sharply and remaining low from the 1940s through the 1970s, before beginning its much-publicized, relentless uptrend over the past few decades, taking it back to the old highs (chart 3). It s a similar, if even more pronounced story for the top 1%. Viewed from this longer perspective, it is the middle decades of the past century, when inequality was relatively low, that appear anomalous. One key difference, though, between today s well-to-do and their early-20th century predecessors, is that many of the latter Figure 2: Shares of Household Income (after taxes and transfers) Percent Lowest Source: CBO Second Middle Quintiles Fourth Highest 81st to 90th 91st to 99th Top Percentiles Top 1 Figure 3: Income Shares in the US Over the Long Run Percent Top 1% (rhs) Top 10% (lhs) Without capital gains Source: World Top Incomes Database Percent 4 Transfers include cash payments from Social Security, unemployment insurance, SSI, welfare, workers compensation, and veterans benefits, as well as the value of in-kind benefits like food stamps, school lunches and breakfasts, housing assistance, Medicare, Medicaid, and CHIP. 5 Saez, Emmanuel (2013). Striking it Richer: The Evolution of Top Incomes in the United States, UC Berkeley Working Paper. 6 Atkinson, Tony, Thomas Picketty and Emmanuel Saez (2011). Top Incomes in the Long Run of History, Journal of Economic Literature 49(1). Alverado, Facundo, Tony Atkinson, Thomas Picketty and Emmanuel Saez (2013). The Top 1% in International Historical Perspective, Journal of Economic Perspectives 29(3). World Top Incomes Data Base, parisschoolofeconomics.eu/ 3

4 lived off income on capital, whereas the modernday affluent even the much-vilified 1% mostly work for a living, relying much more on labor and business income (chart 4). Not just a US story Increasing inequality is not exclusive to the US Many countries have seen similar trends. In fact, the Gini coefficient for household income has risen in a number of advanced economies in recent decades (chart 5). Although the US leads the pack, suggesting it has the most unequal income distribution among its peers, that s always been true, little more so today than in the past. The US doesn t seem to be an outlier in terms of rising inequality. Even over the long haul, the US is not alone in experiencing a u-shaped inequality trend; that pattern has been evident in some other countries too, albeit not everywhere, and rarely to the same extent (chart 6). Indeed, where the US does seem to stand out is the extent to which its increase in inequality has been powered by outsized gains at the very top of the income distribution. Emerging economies have not been immune from the growing swell of inequality either. Many, though not all, have also seen their income distributions grow more unequal over the past couple of decades (chart 7). Interestingly, some of the biggest increases in inequality have occurred within emerging economies that have grown most rapidly (e.g., China and to a lesser extent India), and that have been able to narrow some of their stillglaring shortfalls in living standards with the more advanced economies. By contrast, in emerging economies where growth has been harder to come by, less progress has been made in closing the gap with the rich world, and inequality was often higher to begin with (especially Latin America), it has been less likely to increase further. Though it s difficult to draw firm conclusions here, especially about causality, it seems as if narrowing inequality between emerging and advanced economies has often been accompanied by widening inequality within those emerging economies. Absolute vs. relative Inequality is a relative concept. To say that it is on the rise is to say that one group is advancing relative to another, but it implies nothing about how either is doing in an absolute sense. In the US, for example, rising inequality is largely a story about those near the top increasingly pulling away from the rest, but that doesn t mean that the rest have been stagnating. On the contrary, average household income adjusted for inflation and changes in Figure 4: Income Composition of the Top 1% in the US Percent of total personal income Labor Income Business Income Source: World Top Incomes Database 1956 Capital Income Total Figure 5: Gini Coefficient for Select Advanced Economies (after taxes and transfers) US UK Japan Canada Italy Germany Sweden 1985 Source: OECD Figure 6: Income Shares of the Top 1% in Selected Advanced Economies Percentage of total income US UK France Japan Canada Germany Sweden Source: World Top Incomes Database Figure 7: Gini Coefficient for Select Emerging Economies (household income) Mid 1980s Brazil Source: World Bank China Late 2000s India Indonesia Mexico South Africa Turkey 4

5 household size has continued to increase within each and every income quintile, even as inequality has risen over the past few decades (chart 8). Of course, gains have been greatest in the highest quintile, and especially among the top 1%, but other quintiles have advanced too. In the bottom quintile, for example, after-tax/transfer income (adjusted for inflation and household size) rose almost 50% between the late 1970s/early 1980s and To be sure, increases were more rapid in the 1950s and 1960s, for all except the upper echelon. But it s simply not true that the lower tiers have been treading water (or worse) ever since; reports to the contrary tend to focus only on wages, ignoring the disproportionate growth in benefits, often look only at the individual, failing to account for the increased prevalence of two-earner households, and don t adjust for the declining size of households. In sum, today s poor and middle class are clearly worse off relative to today s affluent than they were a few decades ago, but they are better off than those who occupied the lower and middle rungs of the economic ladder a few decades ago. Mobility And those aren t necessarily the same people. People can move across the income distribution over time, and gauges of inequality like Gini coefficients, or income shares don t address this kind of mobility. Snapshots of the income distribution at different points in time can tell us how far apart the rungs of the economic ladder are, whether they re getting further apart over time, and whether the entire ladder is getting taller (making all rungs higher). But they can t tell us who occupies those rungs, how easily they can move between Figure 8: Real Household Income by Quintile in the US (after taxes and transfers) them (both up and down), and whether that mobility has changed over time. In short, inequality and mobility are not the same thing (though they may be related, as we ll see). Measuring mobility requires tracking individuals over time. Of particular interest is intergenerational mobility, or the relationship between parents incomes and the future incomes of their children. How likely is it, for example, that a child born to parents in one part of the income distribution will move to a different part of his/her cohort s distribution when becoming an adult and has this likelihood changed over time? 7 A number of studies have delved into this, and though results vary, the bulk of the evidence including the most up-to-date, comprehensive study suggests that intergenerational relative mobility has remained remarkably stable over the second half of the twentieth century in the United States. 8 That doesn t mean where you start in life has no effect on where you end up. Of course it does. Just no more than it always has. For example, recent estimates suggest that for children born in the earlyand mid-1980s, every 10 percentage point increase in their parents rank in the income distribution raised that child s rank in his/her cohort s income distribution of by 3 to percentage points (chart 9 ). 9 Crucially, the same study finds similar estimates for children born in the 1970s. For those born in the late 1980s/early 1990s, college enrollment rates which predict future income fairly well suggest that their mobility rates will be similar when they reach adulthood too. And all these estimates are in the vicinity of those from previous mobility studies that looked at cohorts born in the 1950s and 1960s. 10 Figure 9: Intergenerational Mobility Estimates for the US Indexes = 1 in 1979 Lowest Second Highest Top 1% Adjusted for household size. Source: CBO Middle Fourth For a good review of the literature on intergenerational mobility, see Black, Sandra E. and Paul J. Devereux (2011), Recent Developments in Intergenerational Mobility in O. Ashenfelter and D. Card, eds. Handbook of Labor Economics, Vol. 4, Elsevier, Chapter Chetty, Raj, Nathaniel Hendren, Patrick Kline, Emmanuel Saez, Nicholas Turner (2014). Is the United States Still a Land of Opportunity? Recent Trends in Intergenerational Mobility, NBER working paper 19844, January, pp Chetty et. al., ibid. 10 Lee, Chul-In and Gary Solon (2009). Trends in Intergenerational Income Mobility. Review of Economics and Statistics 91 (4). Hertz, Thomas (2007). Trends in Intergenerational Elasticity of Family Incomes in the United States. Industrial Relations: A Journal of Economy and Society 46 (1). Less mobility Child s birth cohort The figures plotted are estimates of the effect of a 10 percentage point increase in parent income rank on a child's future income rank (in percentage points). For children born after 1987, college attendance is used to predict future income. Source: Chetty, et al. (2014)

6 Other metrics paint a similar picture. For example, the probability that a child reaches the top quintile of his/her income distribution having parents in the bottom quintile a true Horatio Alger remains low, at about 8% to 10%, but has been broadly stable over time (chart 10). 11 The chance of such a child entering the middle class (second to fourth quintile) is a much more promising 60% or so (Table 1), which has also changed little over time. Not surprisingly, those born in less austere circumstances have a better shot of reaching the top, but not more so than in the past. As for downward mobility, nearly 65% of those born in the top quintile don t stay there, and about 10% slip to the bottom. Mobility may not have changed much over time, but it does vary a lot across regions. Within the US, areas with less intergenerational mobility tend to have: greater degrees of residential separation (particularly between the middle class and the poor); higher income inequality (not so much at the very top of the distribution, but everywhere else); lower school quality; less social capital (proxied by things like participation in community organizations and voter turnout); and weaker family structure (especially, more children living in singleparent families). 12 Looking around the world, the US ranks fairly low among its rich-country peers on intergenerational relative mobility much as it ranks near the top on inequality. 13 This is part of a broader pattern, as countries with less equal income distributions tend to have less mobility, a relationship dubbed the Great Gatsby Curve (chart 11). 14 What s puzzling in light of this relationship, though, is why the rise in inequality over the past few decades within the US hasn t further reduced US mobility. One possible explanation is that much of the rise in US inequality has been driven by disproportionate gains at the upper tails of the distribution which recent studies have found do not matter much for mobility. 15 It s not how far the top tiers are from the rest that seems to affect mobility, but how far the middle and upper-middle classes are from the poor. And here, the US hasn t seen too much of a change, which may explain why its mobility hasn t deteriorated. Still, regions of the US that have always had greater separation between the poor and the middle classes spatially, socially, and economically have always tended to have lower mobility, and it may be this persistently greater isolation of an entrenched lower class in certain regions that makes overall mobility in the US less than in other rich countries (albeit no more so today than in the past). Although mobility hasn t deteriorated, it does matter more than it used to because of the rise in inequality. To use the ladder analogy, the likelihood of moving between rungs hasn t changed, but the Table 1: US Intergenerational Transition Matrix Child Quintile Parent Quintile Lowest Second Middle Fourth Highest Lowest Second Middle Fourth Highest Note: Each cell reports the percentage of children born in who wound up with income in in the quintile given by the row conditional on having parents who had income in in the quintile given by the column. Source: Chetty et.al. (2014). Figure 10: Probability of Child in US Reaching Top Quintile Given Parent Quintile Figure 11: The Great Gatsby Curve Parent Quintile: 0.50 UK Italy Lowest Second Middle Fourth Highest 0.45 Across countries, more US 35% inequality is associated 30% 0.40 with less mobility France 25% % 0.30 Sweden Germany Japan 15% 0.25 New Australia 10% Zealand 0.20 Norway Canada 5% Finland % Denmark Child s birth cohort The series without dots are based on the birth cohorts. The series with Gini coefficient (more income inequality) dots are for the cohorts. Source: Chetty, et al. (2014) Source: Corak (2013), OECD 11 Chetty et. al, ibid. 12 Chetty, Raj, Nathaniel Hendren, Patrick Kline, Emmanuel Saez, (2014). Where is the Land of Opportunity? The Geography of Intergenerational Mobility in the United States, NBER working paper 19843, January. 13 Corak, Miles (2013). Income Inequality, Equality of Opportunity, and Intergenerational Mobility. Journal of Economic Perspectives 27 (3). 14 Krueger, Alan (2012). The Rise and Consequences of Inequality in the United States. Speech at the Center for American Progress, Washington D.C. on January 12, Chetty, et. al., ibid Intergenerational income elasticity (less mobility)

7 rungs have gotten further apart (especially near the top), so a child s future income (rung height) depends more on his/her parent s rung than it did in the past. Put differently, the stakes of the birth lottery have increased as a result of rising inequality; people born on the lower rungs may have as good a chance as they always had of moving up, but the benefits of winding up on the upper rungs have increased, and those born poor are still less likely than others to get there (though again, not more so than in the past). Why has inequality increased? Several likely causes of the rise in inequality have been identified, though definitive answers remain elusive. Most studies conclude that technological change has played a central role, tilting the demand for labor away from the low-skilled and toward the high-skilled, apparently faster than the supply of skills has been able to increase. 16 To use a popular phrase, in the race between education and technology, technology has been winning, boosting the wage premium of those with key skills relative to those who lack them. 17 This wasn t always true; indeed, in the middle decades of the 20th century, the supply of skills seemed to be outstripping demand, helping level the income distribution. Exactly why education has been losing the race more recently is hard to say, but it is adding mightily to widening inequality. Globalization may have had a hand here as well. In principle, increased trade between emerging economies (that have lots of low-skilled labor) and advanced economies (that have lots of high-skilled labor) should boost the relative wages of the highskilled in the latter, while having the opposite effect in the former in essence, narrowing inequality in both sets of countries. Empirically, though, it has been hard to find much support for this proposition; inequality has actually widened in many emerging economies too, and in the US, which is still relatively closed, it isn t surprising that globalization s impact has been estimated to be modest (though perhaps increasing). 18 Still, a more open world may have amplified the effects of skill-biased technological change in raising inequality, in part by giving the high-skilled larger platforms to market their wares. And FDI from advanced economies to developing ones, especially to industries that require workers who are high skilled by the standards of the receiving countries but that outsource labor that is relatively lower skilled in the advanced economies, may have widened inequality in both places. Again, most studies don t find big effects here skill-biased technological change seems to dominate as a driver of inequality but globalization may have reinforced the trend. As for government policies, as we ve seen, taxes and transfers have not directly abetted rising inequality in the US, but neither have they mitigated it. And while average tax rates haven t changed much, marginal rates have declined over the past few decades, perhaps indirectly boosting inequality by increasing incentives to work and invest, disproportionately benefitting those with higher skills and more capital. Also, some contend that government policies in the US did a better job of spreading the benefits of income-leveling education in the mid-20th century than now. 19 Others point to the decline of labor unions, less-effective minimum wages, and regulatory reforms that have made labor and product markets more flexible as possible culprits. But while these factors may have increased earnings differentials, they ve also likely expanded employment opportunities, and the empirical evidence suggests their net impact on inequality has been small. 20 Societal factors have likely played a bigger role. The increased prevalence of singleheaded households, especially among the poor, together with more assortative mating people with higher earnings tending to have spouses in similar earnings brackets has significantly amplified the growing gap in individual earnings on household incomes. In sum, the rise in inequality seems to have been driven primarily by skilled-biased technological change, an insufficient educational response (that itself may have several causes), reinforced by societal shifts in household structure and perhaps by globalization. But part of the rise, especially at the very top of the income distribution, remains incompletely understood. And others also see much longer-term forces at work. The World Wars and the Great Depression destroyed a lot of wealth, sharply reducing inequality. Then, in the decades immediately following WWII, growth soared as rebuilding took hold and the benefits of previous technological innovations were reaped, generating rapid increases in incomes unrelated to past wealth, helping keep inequality low. But eventually, as 16 See, for example, OECD (2011). Divided We Stand: Why Inequality Keeps Rising and IMF (2007), Globalization and Inequality, World Economic Outlook (October). 17 Goldin, Claudia and Lawrence F. Katz (2008). The Race Between Education and Technology. Harvard University Press. 18 OECD ibid. IMF ibid. 19 Goldin and Katz, ibid. 20 OECD ibid. 7

8 wealth was rebuilt, and the post-war growth boom started to slow in the 1970s, inequality began to creep up again. On this reckoning, the 1940s through 1970s were an egalitarian interregnum caused by a number of powerful yet ultimately temporary forces. 21 Conclusion Inequality is immensely controversial, not least because it touches deeply normative, value-laden nerves about things like fairness. But is there anything we can say that transcends normative judgments perhaps about the relationship between inequality and overall macroeconomic performance? Here, we are on somewhat firmer, but still shaky ground. Few would dispute that a certain degree of inequality is a natural consequence of free markets, differential abilities, and individual choices. Restricting market outcomes too severely in order to achieve some normative sense of equality could impede overall economic performance by reducing incentives to invest, work, save, and strive. Even aiming for perfect mobility could have adverse macroeconomic consequences if it reduced the incentives of higher-income parents to invest more heavily in their children. But inequality could go too far, reaching levels that impede overall economic growth by reducing the mobility of talented poorer people, and potentially rending the social and political fabric, provoking a backlash that weakened support for free trade, competition, and market flexibility. So what s the optimal level of inequality from a macro perspective? Not surprisingly, the evidence is far from conclusive there are too many moving parts to disentangle, and causalities likely run in too many directions. It s encouraging that mobility hasn t deteriorated in the US, and that support for free markets hasn t frayed too badly. But a tipping point could yet come, not only if inequality keeps rising, but also if overall growth remains lackluster and labor markets incompletely healed, keeping lower and middle-income groups from advancing in absolute terms (which they ve struggled to do since the Great Recession). Until more progress is made in these areas and likely even after inequality will remain a front-burner issue. Appendix The Gini coefficient To compute a Gini coefficient for, say, household income across a nation, sort households by lowest to highest income. Then, plot the cumulative fraction of income (measured along the y-axis) accruing to households at or below each point in the income distribution (measured along the x-axis). If, for example, the bottom 10% of households receives 5% of total income, this would be represented by the point (0.10, 0.05). The plot of all these points is called the Lorenz curve (chart A1). The Gini coefficient is a measure of the extent to which the Lorenz curve falls below the 45-degree line which is what would prevail if income were equally distributed. Specifically, the Gini coefficient is the area between the Lorenz curve and the 45-degree line, divided by the total area under the 45-degree line. Since that total area is ½, the Gini is simply twice the area between the Lorenz curve and the 45-degree line. The more unequal the income distribution, the further the Lorenz curve will lie below the 45-degree line, and the larger will be the Gini coefficient. In the limit, if all income were held by one household, the Gini would equal one. By contrast, if each household had the same income, the Lorenz curve would lie exactly on the 45-degree line and the Gini coefficient would be zero. There s another way of interpreting the Gini coefficient: it is half of the average difference in income between every pair of households in the population, expressed as a percentage of average income. In other words, a Gini of, say, 0.4 would imply that the average difference in income between all pairs of households is about 80% (twice 0.4) of average household income, which, in 2007 in the US, for example, would translate into roughly $45,000 (adjusted for household size and expressed in 2007 dollars). Figure A1: Derivation of Gini Coefficient Cumulative share of income Cumulative share of population (from lowest to highest income) Source: DeAWM Actual distribution (Lorenz curve) Gini = 2 x Area A Perfectly equal distribution (45-degree line) Area A Picketty, Thomas (2014). Capital in the Twenty-First Century. Harvard University Press. 8

9 Notes Deutsche Asset & Wealth Management represents the asset management and wealth management activities conducted by Deutsche Bank AG or any of its subsidiaries. Clients will be provided Deutsche Asset & Wealth Management products or services by one or more legal entities that will be identified to clients pursuant to the contracts, agreements, offering materials or other documentation relevant to such products or services. This material is intended for informational purposes only and it is not intended that it be relied on to make any investment decision. It does not constitute investment advice or a recommendation or an offer or solicitation and is not the basis for any contract to purchase or sell any security or other instrument, or for Deutsche Bank AG and its affiliates to enter into or arrange any type of transaction as a consequence of any information contained herein. Neither Deutsche Bank AG nor any of its affiliates, gives any warranty as to the accuracy, reliability or completeness of information which is contained in this document. Except insofar as liability under any statute cannot be excluded, no member of the Deutsche Bank Group, the Issuer or any officer, employee or associate of them accepts any liability (whether arising in contract, in tort or negligence or otherwise) for any error or omission in this document or for any resulting loss or damage whether direct, indirect, consequential or otherwise suffered by the recipient of this document or any other person. The views expressed in this document constitute Deutsche Bank AG or its affiliates judgment at the time of issue and are subject to change. This document is only for professional investors. This document was prepared without regard to the specific objectives, financial situation or needs of any particular person who may receive it. The value of shares/units and their derived income may fall as well as rise. Past performance or any prediction or forecast is not indicative of future results. No further distribution is allowed without prior written consent of the Issuer. The forecasts provided are based upon our opinion of the market as at this date and are subject to change, dependent on future changes in the market. Any prediction, projection or forecast on the economy, stock market, bond market or the economic trends of the markets is not necessarily indicative of the future or likely performance. 9

10 Deutsche Asset & Wealth Management represents the asset management and wealth management activities conducted by Deutsche Bank AG or any of its subsidiaries. Clients will be provided Deutsche Asset & Wealth Management products or services by one or more legal entities that will be identified to clients pursuant to the contracts, agreements, offering materials or other documentation relevant to such products or services Deutsche Asset & Wealth Management. All rights reserved. I (2/14) RETAIL-PUBLIC