Oren Levin-Waldman

  • 02 Jun 2016 8:56 AM | Mike Lillich (Administrator)

    By Oren Levin-Waldman


    Editor's note: This op-ed first appeared online in the Yonkers Tribune.


    As rational actors we identify our goals and objectives, evaluate the different paths towards achieving them, and then we make a choice. In the marketplace, rational firms seek to maximize their profits while minimizing their costs. Consumers seek to maximize their utility, which means obtaining goods at the lowest cost. And workers seek to maximize their wages without having to sacrifice their leisure time.

    I

    n the political realm rational voters will vote for the candidate and/or political party whose platform will most maximize their interests. Therefore, if one group of voters who are poorly paid are promised more programs and/or higher wages, they will naturally vote for the candidate making those promises. Conversely, if another group of voters is promised more favorable business conditions so that they can profit more, then that is the candidate they will vote for.

    What happens when voters are promised things for which they voted accordingly, and then those promises cannot be delivered because the political system militates against them? Do we then say that they acted irrationally? Consider the following scenarios. In the first, we have Bernie Sanders’ voters choosing a candidate because he promises to deliver free college education, to make Wall Street pay for its abuses, and to raise the minimum wage. All of his supporters are “progressive” who believe that big business in cahoots with the Washington establishment have been riding roughshod over the hard working middle class.

    I

    n the second, we have Donald Trump who has promised to make America great again by bringing back manufacturing, making it more difficult for foreign goods to come in and threaten American jobs, and by stopping the inflow of immigrants, especially those who come illegally, because they take American jobs. Of course, he has promised to build a wall along the southern border, to pursue a foreign policy of America First, and to make our allies pay for their own defense. His supporters also believe that private interests in cahoots with the Washington establishment have been riding roughshod over the interests of the hard working middle class.


    Both camps ironically enough do share one thing in common, which is the belief that the political system, because it is in the pockets of special interests, has not been responsive to the will of the American people. Instead of public officials pursuing policies that the public wants, they simply ram policies down the throats of the voters. Then when polls show that there is actually great opposition to these policies, the standard refrain is that there was a failure in communication. In other words, there is no such thing as bad policy and those public officials who push these policies are never wrong.


    We have to assume that even if one of these two were to become president, it is highly unlikely that Congress would pass the policies that are being promised. Does that mean that the voters who voted for them have acted irrationally? After all, they must be aware that in a system of checks and balances it is very difficult for presidential candidates to make good on their promises. But then again, the candidates are preying on the ignorance of the typical voter.

    In truth, it would appear that there are multiple rationalities at work and perhaps in collision with one another. Voters vote the way they do because they are promised something. Politicians who need to raise millions of dollars for campaigns are also acting rationally when they put the interests of donors first at the expense of the masses. And groups within both parties who seek to deny both Sanders and Trump the nomination are also acting rationally because they seek to maintain their privileged position in the establishment hierarchy.


    Still, one wonders how it is rational for working class individuals to vote for a billionaire who has absolutely no connection with them. And how is it rational for affluent “progressives” to support a candidate whose platform is laden with proposals aimed at punishing them for their success? Unless they believe that either candidate will succeed in blowing up the establishment that has effectively stripped them of a true democratic voice for decades now.


    Both appear to have a populist message which resonates, even though they are coming from different directions. Both are appealing to angry voters who believe that the political system is non-responsive to the interests and needs of the middle class. After seeing middle class wages stagnate for four decades now, the decimation of labor market institutions, and the pursuit of trade policies that will only make monied interests wealthier at workers’ expense, voters understand there is little difference between the two political parties.


    And yet, ironically enough, it is rational for the establishment on both sides of the aisle to hunker down and deny that both Sanders and Trump are galvanizing large segments of the electorate because the political system is not working. For the establishment to admit as much is to acknowledge that the establishment and politics as usual is really the problem.


    Let’s see if we can get to the root of the problem. Wages have been stagnant and the manufacturing base that once made “America great” has been replaced with lower paying service sector jobs. Instead of labor market institutions that might have served to maintain wages in the face of greater globalization, the political class has maintained that workers need to be more flexible while the ratio between worker and CEO pay has gone from 40-1 to 500-1. Meanwhile, healthcare costs continue to rise and many are still without because nobody could deliver a single-payer system. Moreover, representatives in Congress are only responsive to those with money. Alas, the American public understands it has been had.


    Granted, most typical voters do not understand the basics of supply and demand. But they do understand that when more unskilled workers are allowed in, or even more skilled workers allowed in on H-B1 visas, this will suppress wages. They understand that the political class could not care less about them. They may even understand that Trump is not likely to care about them any more than anyone in the past has or that Sanders will not be able to deliver on any of his real promises. But they are frustrated which is reflected in their support for them. What they do understand is that the establishment is now running scared, and that, in and of itself, may be a reason to support them. Are they irrational? No, not al all. They are experiencing a different kind of rationality.

    http://www.routledge.com/books/details/9780415779715/#reviews

    Oren M. Levin-Waldman, Ph.D., Professor at the Graduate School for Public Affairs and Administration at Metropolitan College of New York, Research Scholar at the Binzagr Institute for Sustainable Prosperity, as well as faculty member in the Milano School for International Affairs, Management, and Urban Policy at the New School. Direct email to: olevin-waldman@mcny.edu




  • 08 May 2016 12:57 PM | Mike Lillich (Administrator)

    By Oren Levin-Waldman


    First published in the Yonkers Tribune.


    As more states, and even localities, enact minimum wages that are higher than the federal minimum, are there contributing demographic factors that can easily be identified? It would be easy to assume that these are simply blue states and therefore they would be naturally predisposed to passing them. Or it could be that they are simply places where the cost of living is higher and that policymakers understand the challenges of trying to support a family on less. And yet, that would appear to be assuming too much.


    If we follow the logic of the median voter theorem, then states, where the distance between the median voters’ income and average income of society is greater than this distance for the nation as a whole, should be pursuing policies intended to achieve greater redistribution. This redistribution is to be achieved through greater taxation on the wealthy in order to pay for programs that assist those at the bottom of the distribution. In fact, the median voter theorem postulates that tax rates are dictated by this distance between median and average incomes.


    What then, are the characteristics of states with minimum wages currently above the federal minimum wage? These states typically on average are blue states where the cost of living is higher. But data from the 2014 Current Population Survey (CPS), a year when many states began raising their minimum wages, add some interesting dimensions. In many of these states, the value of the dollar is lower than the rest of the nation, meaning that it buys less. Translation: it costs more to live there. But these states also have a higher tax rank, meaning their taxes are among the highest in the nation.


    There are perhaps a couple of ways this issue could be approached. The first is that because both taxes and the cost of living is so high it is only a foregone conclusion that minimum wages would be raised because low-wage workers need higher wages in order to meet their obligations. Of course, that assumes that policymakers are thoughtful people who truly care about the needs of the working poor. And yet, if this were true, they would not have waited until 2014 to really begin addressing the issue.


    The second is to view the high tax rank as a proxy for redistribution. In these states, which are also blue, there may be more redistribution precisely because the cost of living is so high and the low wages that those at the bottom earn are not enough to support themselves and their families above the poverty level. Therefore, people at the bottom need subsidies. And because these states are blue, legislators are more likely to vote for them because this is also a way to get the votes of those at the bottom of the distribution. In other words, low-wage workers in need of subsidy have a vested interest in voting for candidates who promise these subsidies to them, and these subsidies are achieved through redistribution.


    The data actually suggests that states where the distance between median and average income is greater than the nation, are blue and have high union density, and where minimum wages are higher are more likely to have higher tax ranks. In other words, they are more likely to be redistributive states. The variable that appeared to have the greatest effect was high union density, followed by the median voter theorem, followed by being a blue state. This, of course, confirms the hypothesis of the median voter theorem that the greater the distance between median and average earnings, the more likely there is to be redistribution. But this would only appear to be the case in blue states. The low dollar value variable wasn’t nearly as important as high union density and median voter theorem.


    But do these variables also affect whether a state is more likely to have a minimum wage higher than the current federal minimum wage? Here the median voter theorem variable has a negative effect, but the critical variables are low dollar value, high tax rank, and high union density. Interestingly enough, the median voter theorem did have an effect when it was the distance between median and average family income being greater than that distance in the rest of the nation. Moreover, the data suggested that being a blue state had no real impact. Therefore, states where the dollar value was lower than the rest of the nation, both union density and tax rank was high and distance between median and average family income was greater were  more likely to have high minimum wages than the federal minimum.


    One obvious conclusion might be that the public does not view the minimum wage in the same light as redistribution and that the two are affected by different factors. And yet, that the family median voter theorem has an effect for higher state minimum wages where the individual median voter theorem does not might suggest that higher state minimum wages are critical to supporting a family, especially in states where the cost of living is higher.


    That the minimum wage is higher in those states where there is higher union density is important because historically the minimum wage rose when there was a constituency behind them, and that constituency was always organized labor. Of course there is a logic that minimum wages would be higher in states that have higher taxes and where the cost of living is higher. But we have no way of knowing what motivates legislators to support higher wages. Still, the most important factor appeared to be the low dollar value in those states.


    We do know that where income inequality is higher that the distance between median and average incomes will be greater. States where that distance is greater are also more likely to adopt redistributive policies. But also states where the cost of living is higher and the tax rank is higher are also more likely to have higher minimum wages.


    I have suggested many times in this space that higher minimum wages are preferable to the standard redistributive policies supported by typical blue state politicians. In those states where minimum wages are higher, income inequality tends to be lower. To the extent that income inequality is strongly correlated with the distance between median and average incomes, it then follows that the distance between median and average incomes will also be less in those states where minimum wages are higher. In other words, there is less of a need to redistribute income through the standard model of over-taxation on the wealthy to pay for programs as subsidies to those at the bottom.

     

    levin-waldman_-wage-policy-income-distribution-and-democratic-theoryJust published: Wage Policy, Income Distribution, and Democratic Theory:
    http://www.routledge.com/books/details/9780415779715/#review

     

     

    Oren M. Levin-Waldman, Ph.D., is professor at the Graduate School for Public Affairs and Administration at Metropolitan College of New York, Research Scholar at the Binzagr Institute for Sustainable Prosperity, as well as part-time faculty member in the Milano School for International Affairs, Management, and Urban Policy at the New School. Direct email to: olevin-waldman@metropolitan.edu

    Tell Your Friends....Share on FacebookTweet about this on TwitterShare on Google+Email this to someoneShare on LinkedInShare on Reddit

  • 20 Apr 2016 2:04 PM | Mike Lillich (Administrator)

    By Oren Levin-Waldman


    First published online in the Yonkers Tribune.


    As more and more states pass legislation raising their minimum wages to $15.00 an hour, one might think that we will see greater pressure on the federal government to follow suit. And yet, what we are more likely to see is a rehash of the same debate that has unfortunately gotten us nowhere. Those on the right will no doubt claim that increasing the minimum wage will harm low skilled workers as employment opportunities for them will be diminished. And those on the left will harp on the necessity of raising the minimum wage in order to help the working poor.


    It is true that the competitive market model predicts adverse employment consequences due to an increase in the minimum wage. Actually, the model says that either there will be lower employment or productivity will increase. Productivity will either increase because employers invest in their workers’ human capital, in which case they will be more productive or employers will substitute more technology for workers. The problem with this model is that it is nothing more than a theoretical construct with inconclusive data to back it up.


    Consider that from 1981 when the Federal Minimum Wage Study Commission released its report until about 2000 that there was a consensus that a 10 percent increase in the minimum wage would result in a 1-3 percent reduction in employment among teenagers. But the same report also noted that the employment consequences would be considerably less among adults. Since 2000 numerous studies have shown that minimum wage increases affect different groups in the labor market differently.


    What the current data demonstrates is that the minimum wage debate is anything but settled. At best the data is ambiguous, which should pave the way for policy experimentation. In a book titled What Does the Minimum Wage Do? economists Dale Belman and Paul Wolfson did a meta-analysis of the studies on the minimum wage and concluded that there was no basis to conclude that the minimum wage had adverse employment consequences. On the contrary, on the whole, the benefits of the minimum wage were most likely greater than the costs.


    This does not mean that there isn’t a tipping point — a point beyond which we would see employment consequences — only that we don’t know what that tipping point is. A couple of years ago the median hourly wage for full time workers was about $14.90 an hour based on data from the Current Population Survey (CPS). It may be closer to $16.90 now. If we assume that the tipping point is between those two points, then a $15.00 an hour minimum is below that point and there should be no adverse employment effects.


    Still, the political right is strident in the claim that these increases will cost jobs. But what exactly does that mean? Will jobs be less because employers fire workers to compensate for higher labor costs? Or will fewer low-wage jobs be created in the future? The two are not the same. Of course, a higher minimum wage that attracts people into the labor market could result in lower employment because there would then be more workers chasing after the same number of jobs.


    For the business community, the issue is the shock of raising the minimum wage to even $10.00 in the first year, let alone $15.00 by 2020. But perhaps the real issue is that economic models, far from being self-evident truths, are merely bases upon which different groups can cloak their self-interest and effectively make political arguments. Each group in the minimum wage debate will use the model, as well as rely on the data, that best supports its respective interests. But at least the political right appeals to a model.


    The political left doesn’t so much reject the model of competitive markets, as it says so what? To them, the model is irrelevant. All that matters is that businesses are profiting off the backs of their workers and the current minimum wage of $7.25 is insufficient to support themselves, let alone their families. It is a matter of fairness: people who work full time should earn a liveable wage. Of course, they aren’t completely wrong, but the argument has limited political appeal. Rather than arguing for the minimum wage on the basis of class warfare, why not argue that there are indeed positive welfare effects?


    In other words, if minimum wage supporters would couch their arguments in the language of economics and demonstrate that there will be benefits for the middle class, they would actually succeed in the art of triangulation, which President Bill Clinton perfected so well. And that was coopting the opposition. Here is the argument that supporters of the minimum wage should be making, but are not: The minimum wage should be raised because as the statutory wage increases, so too will the wages of those workers earning in wage ranges above. This will have spillover effects, which will work their way up the income distribution.


    As wages increase, the wage stagnation that the country has witnessed for the last four decades will be arrested. Moreover, as wages rise through the distribution, workers will have greater purchasing power and the demand for aggregate goods and services will increase, thereby leading to more job creation. Even if we focus on the old arguments for the minimum wage and accept that there may even be some job loss in the short-term, it is well accepted that in the long-term more spending in the economy due to greater purchasing power will lead to macroeconomic benefits. Lastly, an increase in the minimum wage will narrow the gap between the top and the bottom, thereby resulting in slightly lower wage inequality.


    The minimum wage is important because it speaks to the importance of labor market institutions that serve to bolster wages, and the main reason wages have been stagnant is because these institutions have been in decline. Those who oppose the minimum wage will continue to cloak their selfish interests in the language of the competitive market model claiming to speak the larger public interest of saving jobs for low-wage workers.


    The real public interest lies in rising wages, and jobs that allow for people to live independent lives and not need to rely on the government for subsidies and other handouts. The argument of the political left for a $15.00 an hour minimum would also be stronger if it called attention to the $152.8 billion the nation spends on subsidies for low wage workers, and that $15.00 an hour is the point where we would see less of a need for those subsidies. The political right always speaks the language of personal responsibility and self-sufficiency. Here might be an opportunity for the left to actually speak the same language.

     
    Just published: Wage Policy, Income Distribution, and Democratic Theory:

    http://www.routledge.com/books/details/9780415779715/#reviews

    http://www.routledge.com/books/details/9780415779715/#reviews

    http://www.routledge.com/books/details/9780415779715/#review
    Oren M. Levin-Waldman, Ph.D., is professor at the Graduate School for Public Affairs and Administration at Metropolitan College of New York, Research Scholar at the Binzagr Institute for Sustainable Prosperity, as well as part-time faculty member in the Milano School for International Affairs, Management, and Urban Policy at the New School. Direct email to: olevin-waldman@metropolitan.edu

    Tell Your Friends....Share on FacebookTweet about this on TwitterShare on Google+Email this to someoneShare on LinkedInShare on Reddit


  • 06 Apr 2016 1:59 PM | Mike Lillich (Administrator)

    By Oren Levin-Waldman


    First published in the Yonkers Tribune.


    According to the Median Voter Theorem, when there is greater inequality there will be a greater tendency to redistribute. Why? Because self-interested politicians seeking to gain or retain power will strive to enact public policies that appeal to the median voter. As income inequality increases, the median voter will seek policies that redistribute. What, then, are the implications of this in this election year?


    In its original conceptualization, Anthony Downs postulated that political parties converge upon the position of the median voter, i.e, the voter whose preferences are right in the middle of the distribution of voting preferences. That means that there are as many voters to the left of the median voter as there are to the right. At a minimum, this would imply that parties ought not to be ideological; rather they move to the center in order to win votes. The party that fails to do so does not get elected.


    Still, what does this have to do with redistribution? As incomes are skewed to the right, i.e. more at the top, the preferred amount of redistribution will be a function of the relative position of the median voter on the income scale. The greater the distance between the median voter’s income and society’s average income, the greater is the preferred amount of redistribution. A couple of columns ago, I noted that in 2013 median family income in the U.S. was $56,000 while average family income was $76, 966. And in New York State, median family income was $52,000 while average family income was $98,071.


    The implication would be that income ought to be redistributed so that the median and average match. It would also imply that redistribution ought to be greater in New York State than in the rest of the nation. Still, why redistribution per se? Because the median voter in New York State has as many voters with incomes below him or her as he or she has with incomes above. That isn’t the case with average incomes because a few very high incomes at the top can skew the income distribution to the right, when in reality there may be more people to the left.


    If we follow the logic of this theorem, then both parties ought to be advocating redistribution of some sort. Of course, that assumes that it is in fact agreed that there really is growing income inequality. Our Democratic candidates acknowledge growing inequality while at best the Republicans acknowledge stagnant wages. Moreover, there is no reason to assume that any of our candidates read Downs’s classic An Economic Theory of Democracy, from which this theory came.


    The median voter theorem also implies a role for taxation as the mechanism through which government will redistribute income in order to achieve a more equitable distribution. If inequality suggests the need for more redistribution, then taxes on the wealthy will be raised in order to finance programs and policies that benefit those at the bottom. In theory, this will achieve a more equitable distribution. Although it won’t result in an equal society, it may result in a narrowing of the gap between the top and the bottom.


    Even if higher taxes on the wealthy aren’t used to finance programs for those at the bottom, the gap between the top and bottom will be narrowed in terms of after-tax income. If inequality, however, is denied, then there should be no need to redistribute at all. And yet, some questions remain. Would a supply-side tax cut, often advocated by Republicans which effectively redistributes from the poor and middle class then may it be said to represent the median voter theorem in reverse? Or is inequality being addressed simply through a policy that stimulates economic growth on the assumption that those at the top will take their tax cuts and invest and create jobs?


    This assumption, however, assumes that wages will rise with those investments. If wages rise for those at the bottom and the middle, and especially if they do at a higher percentage than they do at the top, then this could narrow the gap between the top and the bottom, thereby reducing inequality. The problem with this assumption is that it assumes that wages will rise because productivity has risen. But the evidence in recent years suggests otherwise. Despite productivity gains following the end of the Great Recession in 2009, wages did not rise.


    Perhaps the way out of this conundrum is to focus on wages. The logic of the median voter theorem and redistribution in the face of inequality could be applied to the minimum wage. Instead of redistributing income through taxes, there should be redistribution from profits to workers in the form of higher wages. Those opposed to minimum wage increases argue that minimum wage increases are nothing more than redistribution because it is taking from employers and giving to workers who offer nothing more in terms of value.


    Early institutional economist John R. Commons acknowledged that institutions like unions and minimum wages represented a form of redistribution, but that since it was coming from the profits that workers’ labor contributed to it, it was indeed preferable. Arguably, since the firms would not be profitable without the labor of their workers, it is difficult to make the case that it is redistribution in the classic sense.

    And yet, with more localities passing $15 an hour minimum wages, especially in localities with higher levels of inequality, it is hard not to see a different application of the median voter theorem. Although it is not being articulated by policymakers and public officials, the new version might be stated as follows: communities where income inequality is high may seek to narrow the gap between the top and the bottom adopting wage policies that not only raise the wages of those at the bottom, but also effectively raise those in the middle through spillovers.


    Politically, this approach has to be preferable to simply overtaxing the wealthy as some advocate. Were Republicans really smart, they would advocate raising wages too, if for no other reason that to distinguish what we will call “good” redistribution from the traditional redistribution which they obviously consider to be “bad” redistribution. Why is this good redistribution? Because at the end of the day it can be justified on the grounds that workers worked for it. Their labor contributed to the profits from which their higher wages are coming.


    Were political figures to argue this point — to actually re-calibrate the median voter theorem — we might actually find ourselves having a serious discussion about what would benefit the middle class. And what benefits the middle class is what benefits the median voter.

    http://www.routledge.com/books/details/9780415779715/#reviews

    Just published: Wage Policy, Income Distribution, and Democratic Theory
    Oren M. Levin-Waldman, Ph.D., is professor at the Graduate School for Public Affairs and Administration at Metropolitan College of New York, Research Scholar at the Binzagr Institute for Sustainable Prosperity, as well as part-time faculty member in the Milano School for International Affairs, Management, and Urban Policy at the New School. Direct email to: olevin-waldman@metropolitan.edu

  • 24 Mar 2016 1:47 PM | Mike Lillich (Administrator)

    By Oren Levin-Waldman


    First published in the Yonkers Tribune online (http://www.yonkerstribune.com/2016/03/market-economies-require-regulation-by-oren-m-levin-waldman-ph-d) on March 21, 2016.


    Joseph Schumpeter famously described capitalist markets as processes of creative

    destruction. The old and obsolete are replaced by the new and technologically more advanced. This, of course, was supposed to be the measure of progress. In practical terms this meant that buggy whip manufacturers would be replaced by automobile manufacturers. It also meant that in the process of creative destruction there would be serious dislocation.


    Those who performed skilled labor for higher pay would be replaced by low-skilled workers on the assembly lines. And in today’s economy it has meant that higher paying manufacturing has been replaced by lower paying service sector jobs. In short, the process of creative destruction has resulted in the two-tier economy that we have become all to familiar with: an economy of highly skilled and highly paid workers at the top and poorly skilled and poorly paid workers at the bottom.Still, the process of creative destruction is supposed to be a measure of progress. Why?


    Because all workers who have been displaced will in theory be reabsorbed back into the economy. And yet, Schumpeter never really addressed the issue of how long periods of dislocation would last and how long it would take for workers to be reabsorbed. These periods could be as little as a few months and as long as fifty years.


    To the extent that creative destruction speaks to market dynamism, there are two ways to understand it, with clear implications for the political debate over economic policy. One understanding is that competitive markets operate according to natural processes and should be free to run their course unfettered by governmental interference of any kind. If we allow these so-called market forces to run their course, workers will be reabsorbed faster, even if their new jobs don’t pay as well as their old ones did. After all, they are always free to retrain and acquire the necessary skills to obtain better paying jobs. A new economy requires new levels of education.


    The other understanding, of course, is that government in a market economy has a responsibility to ease these transitions and to facilitate workers’ reabsorption. Not only does government have a responsibility to facilitate reabsorption, but a responsibility to ensure that the new economy will pay workers liveable wages. In other words, if more education and skill is required, then government needs to invest in the development of human capital. And if wages are low because higher paying manufacturing has been replaced by low-paying services, then institutions need to be in place to bolster wages.


    I have many times in this space talked about the importance of labor market institutions like unions and minimum wages for bolstering the wages of not only those at the bottom, but of the middle class. But the process of creative destruction actually speaks to a larger issue, which is that if market economies are to survive, then they need more regulation to ensure their survival.


    What we are seeing in the American economy today, which we have been seeing for several decades is what the Marxist economist Michel Aglietta called a crisis in capitalism created by Fordist production. Fordism referred to the early assembly lines that were created by Henry Ford. With these assembly lines more goods and services could be produced more efficiently. But it also meant that skilled craftsmen who were paid relatively well were being replaced by unskilled assembly line workers whose wages were considerably lower. Over time through unionization, these workers’ wages would rise, allowing them join the ranks of the middle class.


    And yet, capitalism, especially under the pressure of globalization is in crisis. What crisis? In order to be competitive, firms need to cut costs and the easiest costs to cut are labor costs. If workers won’t be flexible in their wage demands, then firms will relocate their capital elsewhere. Still, if wages keep falling, then there is no longer a middle class to demand goods and services. As aggregate demand drops, the economy will slow down.


    Of course, we have been hearing from market purists and those on the political right that this crisis only requires greater flexibility on the part of workers. Even politicians from the Democratic party will espouse this position because investment bankers who make large political donations subscribe to this orthodoxy. Aglietta’s point, however, was that capitalist markets actually required greater regulation.

    The nature and scope of changes that have occurred in the conditions of those who work for wages have necessitated what Aglietta called the “development of state influence within structural forms.” By structural forms, he meant a regulatory framework that would allow the market place to survive. These structural forms would surely include labor market institutions designed to maintain wages. And in today’s economy, these structural forms would include a wage policy.


    Let’s remember, in the absence of a serious wage floor, employers will not raise wages because they cannot be sure that their competition will do so. A wage floor in the form of a minimum wage ensures that everybody does because it applies to all. But the development of state influence within structural forms means more. It means regulating the financial sector so that there will be stability. It means investing in education and human capital so that workers can more easily be reabsorbed into the new economy. And it means that actors from the public and private sector need to work in partnership in order to create an effective market regime.


    In urban politics, regime theory has long referred to the relationships between public and private actors for the realization of the public interest. Here it would mean working together for the attainment of an economy that benefits everybody, especially the middle class. It would not be political actors beholden to special interests who seek to enrich themselves at the expense of others. It would not mean a return to laissez faire economics and absolute free trade as espoused by the political right. It also would not mean the class warfare espoused by the political left.


    Until these points are understood, it is not clear that the political debate over how best to restore the American economy will ever progress to one where we can truly believe that the political class, regardless of political party, really cares about the middle class. At the end of the day, it isn’t about economic growth, but economic development.


    http://www.routledge.com/books/details/9780415779715/#reviews

    Just published: Wage Policy, Income Distribution, and Democratic Theory:

    http://www.routledge.com/books/details/9780415779715/#review

    Oren M. Levin-Waldman, Ph.D., is professor at the Graduate School for Public Affairs and Administration at Metropolitan College of New York, Research Scholar at the Binzagr Institute for Sustainable Prosperity, as well as part-time faculty member in the Milano School for International Affairs, Management, and Urban Policy at the New School. Direct email to: olevin-waldman@metropolitan.edu

    Simple Share Buttons Adder (6.1.4) simplesharebuttons.com


  • 11 Mar 2016 3:21 PM | Mike Lillich (Administrator)

    By Oren M. Levin-Waldman


    (First published in the Yonkers Tribune, March 7, 2016)


    We hear a political class profess to care about the middle class, but the obvious question is just what does that mean? How is the middle class defined and can the definition for the nation fit all the states? This is a critical question because the term is often bandied around, but it isn’t clear what it means. Were we to look at workers in families, and define the middle class as these workers falling within certain parameters of family income, we would have a very different definition than if we were simply looking at these workers on the basis of their income alone. And yet, when we talk about growing income inequality, it is often taken as code for the disappearance of the middle class.


    The growth in wage inequality between say 1982 and 2013 is often considered to be problematic because it suggests that there may be fewer people in the middle of the wage distribution between the bottom and top quintiles. This, however, is not easily borne out by the data. In 1982, the average family income of full-time workers in the bottom quintile was $8908, and the average family income of full-time workers in the top quintile was $56,572, yielding a ratio of 6.35 percent. In 2013, the average family income of full-time workers in the bottom quintile was $24,376, and the average family income of full time workers was $217,277, yielding a ratio of 8.9. on this basis we would say that income inequality rose by 40.2 percent.


    Still, how do we define the middle class? In 1982, mean family income was $28,857 and in 2013 it was $94,433. That is the family income of those individuals according to the Current Population Survey (CPS) who were specifically working for wages. The mean family income of all families, however, was only $24,117 and $76,966 in 1982 and 2013 respectively.


    And yet, the percentage of those falling into this definition of the middle class is approximately 40 percent, and it does not really change between 1982 and 2013. The percentage of those individuals earning wages defined by family income between the 40th and 80th percentiles in 1982 was 49.2 percent and 48.7 percent in 2013. Can we really say that this decrease represents a shrinkage of the middle class?


    On the other hand, if our definition of the middle class are those with incomes between the 40th and 80th percentile, and it is only 40 percent of all families rather than those earning wages in families which is close to 49 percent, then we may indeed have something to worry about. We may indeed have a bell shape distribution in terms of wages, at least according to the CPS, but the middle is really not that big.


    Moreover, these definitions would surely vary from state to state. Consider New York State where the cost of living is higher than much of the rest of the nation. The mean family income of individuals earning wages was $29,507 in 1982 and $98,071 in 2013. But the median family income of those working for wages was $26,381 in 1982 and $73,000 in 2013. And the median family income for families was $20,374 in 1982 and $52,000 in 2013.


    This means that by the 40th to 80th percentile definition individuals earning wages were in families with incomes between $22,308 and $43,100 in 1982 and between $57,984 and $135,075 in 2013. And family incomes were otherwise between $16,292 and $37,088 in 1982 and between $39,193 and $113,469 in 2013. Again, about 40 percent fall within those parameters, and there was little change from 1982 to 2013. This would suggest that the claim that the middle class is disappearing is simply not true.


    What, then, is the point of all this? It might strike some as an overly technical discussion of what constitutes the middle class. By some accounts the 40-80 percentile definition is perhaps too broad. It certainly encompasses more people than those earning wages that were typical of the manufacturing sector more than 30 years ago. For those who believe that the answer to income inequality is to raise marginal taxes on incomes over $250,000, the middle class would indeed be defined as those in families earning up to $250,000. But would that be an accurate definition in New York City? And yet if we do define the middle class on the basis of the 40-80 percentile definition, then it isn’t hard to see some politicians calling for raising marginal taxes on lower incomes.


    Peripherally, the reader may want to conclude that there really is no objective definition of the middle class; rather it is all subjective. The data presented in this column is data I ran from the CPS on the computer. Which is to say, we all manipulate data, and politicians with a particular political agenda are no different. What is clear, although not everybody agrees that it is a real problem, is that inequality, particularly wage inequality rose between 1982 and 2013. This was largely because the incomes of those at the top rose at a higher rate relative to the incomes of those at the bottom. Meanwhile, the wages of those in the middle simply stagnated. And this happened because institutions like unions and the minimum wage that used to bolster wages and maintain the middle class were in decline.


  • 10 Feb 2016 11:52 AM | Mike Lillich (Administrator)

    By Oren Levin-Waldman


    (First published online in the Yonkers Tribune)


    As the Democratic Party chooses a nominee to run for president in the General Election, an interesting debate is occurring over the meaning of progressivism. Senator Bernie Sanders accuses former Secretary of State Hillary Clinton of not being progressive enough. She claims that she is progressive because she supports all the right “progressive” policies: universal healthcare, a higher minimum wage, college education, etc. And yet, the progressivism that she and Sanders appear to speak in the name of is really not the same as the progressivism of the Progressives of the early Twentieth Century.


    This raises the obvious question of just what are the implications of different definitions of progressivism for economic policy. Traditionally progressivism was a doctrine of efficiency. The early Progressives sought reform in local government on the grounds that it was corrupt. A system based on the old Jacksonian spoils system — to the victor goes the spoils — could not deliver services in an equitable fashion. In a city like New York residents could count on not receiving prompt services if they backed the wrong candidate.


    The early Progressives, however, were also elitist who sought to displace the ethnic minorities who were dominating municipal government. Born of the patrician classes they felt that it was their natural right to govern, and unless the rules of the game were changed they would never again have their chance. Therefore, they introduced reforms designed to restore themselves to their rightful place.


    First, they supported a civil service system in which workers would be hired on the basis of merit. Their merit would be based on their mastery of scientific principles of management. Second, they introduced at large city council elections in an effort to weaken the hold of ward leaders. And third, they introduced the concept of primary elections so that the voters themselves could choose candidates running for office. This was intended to weaken the control of party bosses making decisions in the old smoke filled rooms.


    They also pushed for ratification of the Seventeenth Amendment to the Constitution in 1913 that resulted in the popular selection of the U.S. Senate. Prior to this, the legislatures in each state simply chose two senators and sent them to Washington. While this along with primary elections were seen as steps in the greater direction of democracy, their real goal was the weakening of party bosses who either were not members of the right social class or had not attained the appropriate level of education.

    Of all these reforms, perhaps the most important one was the idea that principles of scientific management should be applied to public administration. The father of scientific principles of management, Frederick Winslow Taylor, sought to achieve greater efficiency in the new industrial production economy. Industrial production, often referred to as the Fordist Production Process named after Henry Ford, revolved around the assembly line.

    Taylor conducted time and motion studies at the Midvale Steel works in Midvale, Pennsylvania, and later at Bethlehem Steel, in an effort to determine what would make workers more productive and in the end more efficient. It was management’s responsibility to take “soldiering” workers, by which he meant lazy and second class, and turn them into first class workers. If management could not do this, it was a failure. Taylor also assumed that if workers became more productive, employers would reward them with higher wages. Although Taylor never specifically spoke of efficiency wages, he appeared to be introducing the concept because he certainly believed that one way to get product out of workers was to pay them more.


    Sidney Webb, also writing during the progressive period argued that a minimum wage would achieve greater efficiency because better paid workers would better maintain themselves, and in turn would be more productive, thereby adding to the efficiency of their employers. Moreover, higher wages would encourage their employers to invest in their human capital by offering them training so that they would be worth the higher wages.


    At the heart of progressivism was efficiency, whether it be the production of goods and services in the private economy or the delivery of services in the public sector. Today progressivism is more about social justice and achieving the “just society.” But social justice often turns on questions of redistribution, especially from the rich to the poor, as any number of “progressive” candidates today support over taxation of the wealthy to pay for more programs for the poor and maybe even the diminishing middle class. It is not clear, however, that the old Progressives would have recognized themselves in the New Progressivism.

    A

    ll of this raises a fascinating question: Could a Republican running for office supporting a higher minimum wage on the grounds that it is an efficiency wage claim to be a progressive, too? Would this represent a threat to the so-called social justice progressives? Neoclassical economics does stress that if a minimum wage does not result in the layoff of those workers whose value is less than the minimum, it will result in an increase in productivity among low-efficiency workers.


    The neoclassical economist often presents the efficiency side in terms of the anti-shirking wage. Workers paid a higher wage will avoid shirking on the job for fear that if they do they will lose their higher paying jobs and be forced to accept lower paying jobs. Moreover, because they won’t shirk, the employer can save money on monitoring costs. Granted, this is a more negative formulation of Webb’s argument, which we often refer to as the “Webb Effect,” but it does recognize the greater efficiency to be derived from higher wages, even higher wages that are legislated.


    Can our competing definitions of progressivism be reconciled? I have argued many times in this space that the path forward with regards to economic policy favorable to the middle class does not lie with redistributive policies. Rather the path lies with strong wage policies that through wage contour effects can lift, not only the wages of those at the bottom, but those in the middle as well. Although such an outcome will no doubt further the objectives of social justice, it would be in lines with the old progressive vision of the efficiency wage. Perhaps it is time to resurrect the efficiency wage argument as the best way to appeal to those middle class voters who remain.



  • 20 Oct 2015 8:25 AM | Deleted user

    Following the first Democratic presidential debate, one could not help but wonder what exactly is the party’s economic platform. At times it appeared that each candidate was attempting to outdo the other in terms of how much the wealthy should be taxed in order to pay for more programs for the poor and the middle class. Although each candidate professed his or her love for the middle class, it became apparent that nobody was really addressing their needs.

    Sure, they all talked about the need for higher minimum wages, but none of them seem to grasp the importance of the minimum wage for the overall health of the economy. The minimum wage is critical because it serves as a reference point for wages in specific industries — what we would term the low-wage industry — with potential spillover effects for others. In other words, if you were setting up a businesses in services, how would you know how much to pay your workers? You would most likely survey other businesses in the same industry to get a sense of what the prevailing wage is. But this is not what neoclassical theory tells us about how wages are set.

    In the theory of competitive markets, wage setting is fairly simple: wages are set at the point where the demand for workers intersects the supply of workers. The more workers there are the lower their wages will be because more workers are competing for fewer jobs. Therefore, they will lower their wages until their labor services are consumed, i.e. purchased by employers. The problem, however, is that one would think there is one labor market, when in reality there are multiple labor markets. The labor market where the minimum wage has its greatest influence is in the low-wage labor market, which also happens to be the lowest skilled market.

    In this labor market there is also an oversupply of low-skilled workers, which only drives their wages down further. Now for those industries that would like to pay as little as they can, which is also consistent with the basic tenets of micro-economics, these workers precisely because they have no real skill to offer are in no position to bargain for higher wages. A wage floor becomes necessary to ensure them a measure of monopoly power they otherwise would not have.

    This is precisely why collective bargaining was so important to the building of the middle class during the Twentieth Century. It gave workers a measure of monopoly power to bargain for higher wages. In other words, it gave workers voice. Moreover, factor worker at the end of the Nineteenth and beginning of the Twentieth centuries were considered to be no more skilled than low-wage service workers are today. The fundamental difference is that they were able unionize, and unions were able to afford them dignity in their work, while low-wage workers are not able to organize.

    The neoclassical economist, however, might respond in two ways: First, collective bargaining is no better than a wage floor, rather the effect is to artificially inflate wages where natural market forces would otherwise leave them. And second, a wage floor results in lower employment because workers are unable to accept jobs at lower wage rates. If wages are artificially inflated too much, employers might seek to substitute technology for workers. In other words, we are led to believe that wage setting according to the laws of supply and demand are really natural forces. 

    Perhaps the real issue is that there is no such thing as natural forces when it comes to wage determination. The so-called laws of supply and demand work when we have skilled workers, but when our economy is only leaving us with low- skilled workers, then employers need a little push from the state. The purpose of the minimum wage then is to afford workers a measure of monopoly power.

    In other words, heterodox economists don't disagree with the neoclassicals; they simply say so what? What heterodox economists realized was that workers and employers are not equal in their bargaining power. The neoclassical model always assumed that they were. Early heterodox economists, like John R. Commons, emphasized that the labor market imperfectly gave employers superior bargaining power relative to individual employees. There is an asymmetrical power imbalance between employers and their workers. Employers are wants traders while workers are needs traders.


    The employer has sufficient resources that s/he does not have to hire workers immediately. Rather, s/he can wait it out until the price of labor drops to a more favorable level. The worker, however, is a needs trader who, because s/he needs to eat, does not have the luxury of waiting it out until employers raise their wages. Therefore, they will take whatever job is available. This, of course, gives the employer considerable power over workers.  Because of this inequity in bargaining power, there really was nothing to prevent the economic coercion of workers.

    A wage floor, then, increases the bargaining power of those at the bottom rungs of the labor market. But for those employers who might like to obtain more effort from their workers in the general industries where minimum wage workers are found, they can now use the statutory minimum wage as a reference point for the wages they will pay. An increase in the minimum wage will no doubt affect the wage rates of those earning between $8-10 an hour or even more. Moreover, there is no reason to believe that there aren’t spillover effects into those industries paying more. After all, if the wages of low-skilled workers rise, is it unreasonable to expect that more skilled workers will similarly demand higher wages?

    Now we can cut to the chase. The minimum wage does not really refute the laws of supply and demand, but the idea that it is a reference point does expose as hollow the idea that wage setting is really a natural process. It then becomes even more curious why those politicians who talk about the power of corporations don’t seek to explain the operations of the market place in these terms. Those who really want to help the middle class should seize on this idea of the minimum wage as a reference point, whose increase can help many more. As wages increase, so too does purchasing power, which in turn leads to more demand for goods and services in the aggregate. This is ultimately what drives the economy; not more programs paid for by more taxes on the wealthy.

  • 07 Sep 2015 9:20 AM | Deleted user

    Another Labor day has come and despite all the tribute to the hard work of those who toil for wages, wage earners have been losing ground. Wage inequality continues to grow and we see more calls for higher state-level and local minimum wages. We may never resolve whether the minimum wage is a net positive or a net negative for the economy because the issue is steeped in politics, ideology, and competing economic models that are manipulated by different groups for their respective interests. And yet, these are all symptoms of a larger problem.

    Of course the larger issue is economic transformation and how we deal with it. Industrialization no doubt resulted in the growth of an unskilled class of workers earning barely subsistence wages and plutocrats holding substantial wealth. As labor unions addressed the asymmetrical power relations between management, coupled with public policy, we began to see the emergence of a middle class.

    Then with deindustrialization and capital mobility, middle class and unionized factory jobs were lost and replaced with low-skilled and low-paying service sector jobs. It isn’t clear that retail jobs require that much less skill than the lost manufacturing jobs. What is clear is that the problems of growing wage inequality and the decline of the middle class are really symptoms of the larger issue of the “commodification” of labor.

    Labor has become nothing more than a commodity. As commodities workers are interchangeable, disposable, and expendable. If commodities, what difference does it make if paid $20 an hour or $7.25? In neoclassical economics, firms simply seek the cheapest inputs. As a commodity, labor is simply an input. If commodities or inputs, workers, then, cease to be people. In the context of corporate decision making they have no feelings and certainly have no needs.

    Arguably the transformation of workers into commodities began with adoption of the Fordist mode of production — the placement of workers on an assembly line. This was work that could be done by anybody. The Fordist mentality, however, didn’t become obsolete with the disappearance of factory jobs. Rather we see it in low-wage and low skilled service jobs. Those who flip burgers in fast-food restaurants are the new assembly line workers.

    So long as we spout the myth that these jobs are low-skilled we as a society can feel good about treating those who work them as commodities not worth more than they currently receive. Surely they are not worth any further investment into their human capital. But then why would we invest in their human capital if they are not really human?

    In her tale of living as a low-wage worker in Nickel and Dimed , Barbara Ehrenreich tells of how these workers really engage in tedious, difficult, and humiliating work. These workers actually have more skills than we would like to believe. To call them unskilled workers, after all, is to further demean their labor. Because they are unskilled, they aren’t worth the wages they are receiving, let alone higher wages. Because we see them as unskilled workers, we don’t really see them as human beings that work because they need to.

    We add further insult to injury by talking about what makes for an efficient economy as though treating workers as people would be anything other than inimical to efficiency. Again, we as a society can feel good if we can couch opposition to human capital investment, higher minimum wages, and just shoring up the middle class in the larger public interest of efficiency.

    One wonders, however, how efficient it is when workers too poor to sustain themselves are simply unable to purchase goods and services. Has anybody wondered how it is that the term economic efficiency has come to be a mask of simple greed? Surely employers aren’t going to say in public discourse that they are opposed to higher minimum wages because it means they can keep less in profits for themselves.

    It does sound so much more civilized to talk about how inefficient it would be if low-skilled workers — especially if they are not primary earners — were to lose their jobs. After all, that would be contrary to the public interest. And yet, it is easier to make this argument because workers have been reduced to commodities.

    Now you might ask just what it is that we celebrate on Labor day? Of course, that would be a good question. Politicians will be quick to offer the standard bromides and platitudes. If we really want to have a meaningful Labor day, then we need a serious program for shoring up workers and the middle class. At a minimum, we need to be concerned about wages. That means strengthening traditional labor market institutions like unions and the minimum wage that served to bolster wages. It is because of their decline that wages have been stagnant for more than thirty years now.

    The real work is for us to change how we view workers and their labor. We need to stop viewing them as mere commodities and begin viewing them as people who indeed have human needs. Until we do that it will be difficult to have a serious policy discussion that extends beyond the platitudes characteristic of most political campaigns. And until we do that, we are simply making a mockery of Labor day.

    I am available for comment: (914) 629-6351

  • 26 Aug 2015 2:56 PM | Deleted user
    As the Dow Jones plummets amid concerns of a new recession on the horizon, there will naturally be those who argue that now is the wrong time to raise interest rates. Moreover, it will certainly be the wrong time to raise wages. But it is high wages that drive the economy, not lower interest rates. Higher wages after all, afford workers greater purchasing power, which in turn enables them to demand more goods and services in the aggregate.


    Over the last couple of years, many states have either raised their minimum wages or adopted them. One of the most interesting cases is that of Oregon. Between 2002 and 2014, the State of Oregon has had a steadily increasing minimum wage. In 2002 the minimum wage was $6.50 compared to the federal minimum wage of $5.15 an hour. In 2014, the state’s minimum wage was $9.10 an hour while the federal minimum wage remained at $7.25 an hour. Currently Oregon’s minimum wage is $9.25 an hour.


    The standard model of competitive markets holds that a rising minimum wage will lead to lower levels of employment. But recall the 2014 report of the Congressional Budget Office (CBO) that concluded that a rise in the minimum wage to $10.10, as was proposed by President Obama in his State of the Union address, would, despite resulting in as many as 500,000 fewer jobs, would on the whole greatly benefit the economy. As many as 16 million Americans were going to see their pay rise, and they would spend it in the economy.


    Oregon’s experience appears to bear this out. To understand what is happening and why the states may be a model for the nation we need to abandon the standard model of competitive markets and look at the minimum wage through the prism of wage contours. The concept of wage contours was developed by Harvard economist John Dunlop in the late 1950s, who would also serve as President Ford’s Secretary of Labor during the 1970s. Dunlop developed the concept to explain how a firm’s internal wage structure might be as much affected by external forces as internal ones.


    A wage contour was defined as a group of workers with similar characteristics working in similar industries and earning similar wages. For each group there would be a group of rates surrounding a key rate, and these group rates would be affected by changes in the key rate. Within an industry, the key rate was essentially any rate serving as the reference point for the industry. As key rates were specific to industries, they could also vary from industry to industry. Similarly the logic would apply to sectors.


    Therefore, if we understand the minimum wage through the prism of wage contours, then the statutory minimum wage is nothing more than a key rate for the low-wage sector. Low-wage workers working in an industry that pays slightly more than that key rate but are part of a group earning around that key rate who will also be affected by changes in that key rate.


    Data from the Current Population Survey (CPS) from 2002-2014 shows that the statutory minimum wage does have an impact on wages around it. Using this data, I constructed 10 contours. Beginning with the statutory minimum wage in each year the first contour included wages between the statutory minimum wage and 25 percent above. Each successive contour ran an additional 25 percent until a total of ten were created. At the national level median wages rose in each year the statutory minimum rose, and did not rise in years when the statutory minimum did not.


    In more concrete terms, between 2002 and 2006 when the minimum wage was $5.15 an hour, the median wages were $5.77 an hour in the first contour, $7.21 in the second, $9.13 in the third, and $11.54 in the fourth. The median wages in these contours did not increase until the minimum wage increased from 2007 to 2009. By 2009 when the federal minimum wage reached $7.25 an hour, the median wage in each contour was $$8.17, $10.00, $12.50 and $15.55 respectively. They have remained relatively unchanged since.


    In Oregon where the minimum wage was higher to begin with and was rising consistently each year, with only one or two exceptions, median wages also rose for the most part, although there were some exceptions. During the same initial four year period median wages rose from $7.22, $9.13, $11.54 and $14.42 respectively in 2002 to $8.41, $10.58, $12.98, and $16.83 respectively in 2006. By 2009, these median wages were $9.62, $11.54, $14.42, and $18.40 respectively. And in 2014, they were $10.00, $12.50, $15.38, $20.19 respectively.


    Meanwhile in Pennsylvania, where there has been no state set minimum wages, median wages have hovered around the national median wages. At the same time, unemployment, for the exception of when it peaked to 14.8 percent (according to the sample) in 2009 at the height of the Great Recession, it steadily decreased in Oregon. Between 2002 and 2008, unemployment fell from 8.8 percent to 5.6 percent. It then fell again from 14.8 percent in 2009 and 7.2 percent in 2014.


    If we consider the first four contours to be constitutive of the larger low-wage labor market — what we would define as the “effective” minimum wage population — it isn’t too difficult to see the similarity between Oregon and the conclusions of the CBO report. Moreover, that the minimum wage appears to result in wages rising is actually consistent with the preponderance of studies showing that there is a strong correlation between minimum wage increases and increases in average wages, especially at the lower tail of the distribution.


    Now I only presented what was happening in the first four contours; median wages following increases in the statutory minimum wage, were rising in the other contours too. On the whole, this would suggest that if we want to help those at the bottom of the distribution, the minimum wage may be an effective tool. That the median wages in other contours are also rising suggests that the larger middle class will also benefit.


    Contrary to the typical big government response that often comes from Washington advising that during a recession workers need more programs, they instead really need higher wages. It is rising wages that fuel the economy because increased purchasing power leads to increased aggregate demand for goods and services. If policymakers are looking for demonstration projects to support this position, they need look no further than the states. Oregon appears to be a good example.

    ------------------------------

    See article originally appearing at: http://www.yonkerstribune.com/2015/08/public-policy-what-policymakers-can-learn-from-oregons-minimum-wage-by-oren-m-levin-waldman-ph-d

    ---------------------------------------------------------------

    Oren Levin-Waldman is professor of public policy in the School for Public Affairs at Metropolitan College of New York (olevin-waldman@metropolitan.edu ) and author of several books on wage policy. They include the just published: Wage Policy, Income Distribution and Democratic Theory ( http://www.routledge.com/books/details/9780415779715/#reviews ); The Political Economy of the Living Wage: A Study of Four Cities (M.E. Sharpe 2005); and The Case of the Minimum Wage: Competing Policy Models (SUNY Press 2001). He is a researcher for the Employment Policy Research Network (EPRN), and some of his work can be found at http://www.employmentpolicy.org/people/oren-levin-waldman.

Employment Policy Research Network (A member-driven project of the Labor and Employment Relations Association)

121 Labor and Employment Relations Bldg.

 

121 LER Building

504 East Armory Ave.

Champaign, IL 61820

 

The EPRN began with generous grants from the Rockefeller, Russell Sage, and Ewing Marion Kauffman Foundations

 

Powered by Wild Apricot Membership Software