Tuesday, December 14, 2010

"Restoring" The Middle Class - Promise or Threat?

It's pretty well known that the middle class is shrinking in this country.  Usually, when politicians cite the shrinking middle class, they are implying that most people are getting poorer - possibly that the "elite" are somehow actively campaigning to drain the middle class to turn more Americans into wage slaves.  Both parties seem to use this rhetoric - Democrats use the shrinking middle class to justify additional spending on social programs.  Republicans use the shrinking middle class to justify stances on taxes and immigration.

The issue has become intensely politicized, but let's take a look at some fundamental data and we may see something really interesting going on!

I am going to cite to this excellent article with regard to "How The Middle Class Is Shrinking."  The author worked with an economist and identified approximately 1969 as the high point for the percentage of American population included in the middle class.  The middle class is defined as those earning between $35,000-$99,000/year in inflation-adjusted 2009 dollars.  Admittedly, that's a pretty wide income spread, but it seems like it captures the middle class decently.

Here is the percentage of the population that fell in that income range over the previous years:
2009: 43.7 percent
2000: 45.6 percent
1990: 47.9 percent
1980: 49.3 percent
1969: 53 percent

As you can see, the total percentage of the population that falls in the middle class range has shrunk from 53%-43.7%, a difference of 9.3% of the population.

Our immediate assumption (and the assumption that most politicians push) is that that 9.3% (or at least the majority of it) has now fallen into the lower income brackets and is now destitute and in need of immediate help.

However, take a look at the percentage of households with income over $100K, adjusted for inflation:
2009: 20.1 percent
2000: 20.6 percent
1990: 15 percent
1980: 10.4 percent
1969: 7.6 percent

WOW!  We note that the percentage of "high-income" households has risen from 7.6% of the population to 20.1% of the population - a difference of 12.5%!!  That means that there are about TRIPLE the percentage of high income households now as there were in 1969.

Further we note that the increase in high income households (12.5%) is considerably larger than the decline in middle income households (9.3%).  Now, I recognize that we are talking about percentages here and individual households can go up and down.  However, based on these statistics we can see that the decline in the middle class is more than offset by an increase in the high income households.  Thus, it would appear to be appropriate to say that the primary reason why the middle class is shrinking is that they are making more money and thus being reclassified as high income!

This would seem to put a whole new light on politician's promises to "restore the middle class" - Frankly, if they were to "restore" the relative percentages of the middle class, then a whole lot of people would have to be making a whole lot LESS money.  For example, at least 12.5% of American households that are currently making more than $100K would have to have their income reduced below the $100K mark - considering that there are about 100M US households, that means that about 12.5 M households would have to see a significant reduction in their income to "restore the middle class." 

Consequently, when you hear a politician talking about "restoring the middle class," he or she is really not talking about making people richer - they are talking about making 12.5% of US households poorer.  That sounds to me like more of a threat than a promise.


  1. This is not the right way to define "middle class." The US economy has grown faster than inflation since 1969, meaning that GDP per capita is much higher now than it was then, so of course you'd expect that incomes have gone up and that people have moved up out of the 35-99k bracket. You have to adjust not just for inflation, but also for economic growth.

  2. I don’t like Mr. Newman’s analysis.
    (1) Piggybacking off Andrew above, income inequality is better measured in ratios to the median, not in an absolute dollar band between $35k and $99,999. Looking at it this way, it’s pretty clear that those with higher incomes gained more over the years than those who didn’t (http://en.wikipedia.org/wiki/File:US_Income_Inequality_1967-2003_relative_to_median_%28log_scale%29.svg). As a result, the US’s Gini index (which has its flaws) has risen over 0.4 in the last few decades. Contrary to what Mr. Newman said in a linked article (http://money.usnews.com/money/blogs/flowchart/2010/10/06/the-myth-of-disappearing-prosperity), income inequality isn’t caused by lack of education but by the fact that productivity gains have gone to producers rather than laborers. Citing education sees the symptom as the cause and not vice versa.
    (2) Notwithstanding the above, income isn’t the sole way to measure the middle class. Wealth and debt inequality shake up the situation. Many high earners also have high debts, such as the U Chicago law professor from a few months back.
    (3) In fairness, class and poverty in these discussions are defined in *relative* terms. For example, it’s cheaper to be poor in the United States than in Afghanistan. Put differently, poorer Americans have an absolute higher standard of living than poorer Americans did a few decades ago. We don’t all need to own houses rather than duplexes, two cars instead of mass transit, etc.
    (4) However, these discussions are important to politics for several reasons: (a) wealth concentration generally weakens democracies, (b) if I’ve learned anything about economics over the last decade it’s this: economies thrive on spending. Poor people don’t spend money; rich people don’t spend money. If we want to escape our balance sheet recession, we must address these facts.

  3. Hi Andrew - Thanks for the comment! I certainly agree that the US economy has grown faster than inflation. However, it seems like the best measurement of the middle class over the years is based on the relative value of its spending power - and that would be adjusted based soley on CPI inflation as in the article, right? In other words, why do you think that adjusting downward based on the economic growth rate of the US provides a more accurate picture of the middle class than that?

  4. Hi LSTB - thanks for the comment! Your comment may be giving me more insight into Andrew's comment. You seem to be going more toward "income inequality" rather than "middle class". I'll agree with you in general, but posit that these are really two different issues. Personally, if we want to talk about the evolution of the middle class over time, I think that basing our definition on the adjusted spending power of their salaries is a pretty good way to remain consistent in general.

    However, I agree with your point that the middle class today may have more debt - and thus less real spending power than the previous years - and that consequently focusing on income-adjusted salary alone may not represent an accuarate reflection of the true spending power. Consequently, just relying on income alone may be overestimating the spending power of today's middle class. This would seem to not include mortgage debt - which should be rolled into the CPI housing cost increase. Further, I don't think that it would be fair to include credit card debt here either because it represents previous spending power that has been expended. However, student loan debt would certainly seem fair to include.

    It would certainly be interesting to see how much the spending power of the middle class categories above is depressed due to student loan payments. My feel is that it would be significant, but not significant enough to overcome the 12.5% transition of American families from middle class to high income. Maybe a third of that (4%-ish)?

    With regard to wealth concerntration and democracy - I am not sure that the data supports a cause-and-effect relationship like you posit instead of jsut a correlation relationship. For example, I'll agree that high wealth concentration is correlated with weak democracies that typically benefit the wealthy. However, I am not sure that wealth correltation actually weakens democracies as you suggest.

    Also, I'll agree with your comment that economies thrive on spending - but reasonable, sustainable spending that does not creat asset bubbles. However I'm not sure what you mean when you say that neither poor nor rich people spend money.

  5. Just a thought to chew on: perhaps its not just income that defines the middle class and the definition includes a sense that one can move ahead. That perception of upward mobility was lost in teh end of the 20th century along with the frontier and desire to take on finanical risk. The middle class has internalized the old saw, there are no second acts in American [finincial]life.

  6. 8:38 - Great comment. The phrase "middle class" has not be adequately and concretely defined in public discourse. In the article above, middle class is defined as those earning beterrn 35 and 100K in adjusted 2009 dollars, but when people read "middle class", they may import other definitations as well - beyond the purely economic ones. None were meant in my post above.

  7. MP, you’ve made me write a manifesto, but it’ll help me in a later blog post.

    First, as to income inequality vs. middle class, I’m starting from Newton’s income-based definition, not my own. Second, I appreciate that you included student debt in measures of real spending power, especially because I wrote on that last week (direct link in my sig).

    Perhaps to answer your other points, I should start by how I define “middle class,” and Anonymous 8:38 is on the right track. I don’t use the term because it only tells us that there’s a class between an upper and lower class without telling us how those two classes relate. Allow me to manifestify:

    Definition of capitalism:
    *An economic system in which producers compete to sell goods and services to consumers at the cheapest price possible*

    It assumes that consumers demand the goods/services, can afford them, and rationally choose to buy the cheapest ones.

    For capitalism to work, the increased productivity resulting from the competition must benefit the producers, and this is effected through property rights. Thus, as productivity increases, goods/services become cheaper for consumers, and producers benefit by innovating before other producers.

    In practice, though, we have a problem. Not everyone is a producer. Some people are workers, who don’t appear in the definition. So there are two classes, and unlike the upper/middle/lower one from above, we understand more about how they relate: As productivity increases, the producers’ need for workers diminishes (layoffs/downsizing/outsourcing). The Luddite fallacy prevents us from having a permanent mass-unemployed underclass. The workers find new jobs that are often safer.

    Here are three factors that distinguish producers and workers:
    (1) Workplace Autonomy
    (2) Job Security
    (3) Income

    The less of these you have, the more likely you are to be a worker than a producer. However, we end up in situations where we have uneven distributions of them. I’ll give two examples: the first is the branch manager of a bank. Branch managers have some autonomy and some income, and they’re unlikely to be replaced by a machine like a teller. They may be a hybrid, deriving income from salary and maybe from investments as well. They can’t be easily categorized as producer or worker.

    Second example, I have a childhood friend who owns two LLCs (a band and a recording studio). He obviously has autonomy within his companies because no one is going to fire him, but his income isn’t remarkably high (to my knowledge, and plausibly less than the branch managers'), and his job security is dependent on demand for recording services and his band’s performances--industries that’re easily affected by recessions.

    It’s in these circumstances that I think we come up with terms like “middle class.” It’s a recognition that the above three factors aren’t always black and white. In the American reality, though, I think most Americans are working class and not producing class, and definitely not some underspecified middle class. However, our politicians *certainly* do not like using these terms. I fear discussions of economic policy will be nonsensical until they do.

    To editorialize slightly, because the definition of capitalism barely implies workers’ existence, it’s clear it prioritizes producers’ privileges. Capitalism fails when producers aren’t rewarded for productivity gains (as we learned in 20th century communist states), but when they gain too much from productivity gains, the excess labor pool suppresses wages and grants producers monopsony buying power over the labor market. Thus, economic spending slows because the deadweight loss discourages producers from spending their reserves, yet workers have no money to spend to buy their products. This is what I mean when I say rich people don’t spend money; poor people don’t spend money. The system stagnates until workers’ spending power returns, and it’s certainly going to take a while if they’re paying down debt, as is happening now.

  8. LSTB - Thanks for the clarification - and for spending the time! As for your worker/producer dichotomy, it is interesting to consider, especially in a historical context. For example, jump back 125 years and you have about 90% of the American population working as producers - farmers, ranchers, fishermen, etc. However, they then give up their jobs as producers to take higher-paid jobs as workers - jobs less exposed to income risk. Less workplace autonomy, but more job security and income.

    Consequently, with your worker/producer factors, the only one that really works for me is autonomy - producers have more of it. However, that often comes at the expense of income. Further, the "job security" aspect does not really seem to work. For a producer like a farmer, if the income goes down to zero they are out of luck - are they "fired"? No, but is that meaningful when their income is crushed? Conversely, if there is a worker at a company and the company's profitability goes down one year, the worker may very well be insulated from the economic decline.

    Interestingly, your definition of capitalism seems to have a kind of imported Marxist element where it imports "workers" into the equation. I doesn't seem like they belong there. Instead, I would proffer that capitalism definitionally just has two parties, the producer and the consumer - and that capitalism is for the benefit of the consumer, not the producer. In capitalism, we want producers to compete to deliver their goods for the lowest price - so that the consumer wins. Whatever method the producers decide to use to produce their goods (within the constraints of the law) is up to them - they may use workers or they may use independent contractors (which seem more like "producers" rather than "workers", no?). For example, if a producer wants to sell baskets, they can either build a factory to sell baskets or they can just tell people that they are interested in buying baskets from them - the people then make the baskets at home as independent contractors and the producer purchases the baskets - or is the producer something else here? Is the producer now just a merchant buying and re-selling? Does this still comply with the worker/producer duality? Also note in the hypothetical above that the people making baskets at home may be better or worse off than they would if they worked in a factory. For example, in the factory they get steadier pay - and maybe higher - but less autonomy. At home, the producer has shifted most price fluctuation risk onto the basket makers and has spared itself the capital expenditure of building a factory. Under which set of facts are the basket-makers better off? It depends.

    Furthermore, I have to disagree that capitalism prioritizes producer's privileges. Instead, it prioritizes the benefit to the consumer. You note that capitalism does not mention the worker, but then it should not - because it is not about the specific modality that is employed to produce the product, instead it focuses on the exchange/sale of the product. As a counter example, mentioning that capitalism does not address workers seems as mis-directed as condemning Marxist thought because it barely mentions "consumers" - when it doesn't mention consumers because it is not primarily directed to the consumer/producer exchange and instead is directed to the means of production.

    Finally, lest I inadvertantly give offence, I want to thank you for greatly enriching the discussion with your comments. Nothing in my words above should be taken as discouraging your voice - it's boring when everyone agrees! I hope that we are able to continue to (as Shakespeare said so eloquently) "Strive mightily, but eat and drink as friends." I look forward to considering more of your comments with an open mind!

  9. MP,

    Thankfully, my last post didn’t take long, though I need to be more concise. Hopefully this one won’t either. I also don’t mind a civil disagreement that I can learn from, and your previous comment was very helpful. So, I have a few points, in no particular order:

    (1) If you want to condense class distinctions to workplace autonomy, I see no reason to disagree. So long as we’re approaching a distinction that isn’t solely based on income, I’m satisfied.
    (2) I realized the problem here is that I equivocated producer and producer class (if you’re coloring me Marxist, I’ll just call them “capital owners”), which makes a mess of my previous post. All workers are producers; all capital owners are consumers. Reflecting on your basket-seller example, I risk over-simplifying things to buyers vs. sellers, but I’m willing to live with it.
    (3) You are right that capitalism discusses only producers and consumers, which is why I hedged myself by saying, “In practice,” when we’re discussing class. There appears to be a commonality between producers of unskilled labor and one of capital owners that isn’t inherent to discussions of economics.

    To simplify:
    We can discuss capitalism without class.
    Can we discuss class without talking about workers? I don’t think so. Newman does.

    Why did I bring up capitalism, then? It has to do with productivity and Newman’s definition of the middle class. Capitalism promotes productivity, but its indifference to class distinctions means it won’t tell us what happens if some classes benefit more from productivity gains than others. Newman thinks this doesn’t matter, that productivity ultimately benefits everyone.

    At the risk of straw-manning him, he seems to think that the middle class is growing, and it’s doing so via education because in the New Economy educated people have higher incomes. His income-defined lower class has had its jobs outsourced or bubble-busted out of existence, so it needs to adapt because the Luddite fallacy tells us that for every low-income job lost, logically a higher-income job must come into existence elsewhere. It’s just a matter of getting them there.

    By contrast, I think discussions of class require investigating whether productivity benefits some groups over others. When that happens (and this is something I’m still thinking through), the system gets gummed up. Sure more people may be earning more, but spending isn’t growing. People are either paying down their bubble-level debt or don’t have enough money to buy anything, reducing overall spending and creating unemployment. More education doesn’t help (at least not directly).

    PS I appreciate the Shakespeare.

  10. This post is idiotic and based on a complete misrepresentation - it's a statistical manipulation meaning nothing, and the author should be embarrassed.

    Check your definitions of middle class, compare purchasing power for the stated time periods, and otherwise come up with a basic statistical model that MEANS something. This is so stupid, and it's presented like, "Look what I did with my smarts!" Infuriating.

  11. Hi Liz, thanks for your comment. The original statistics were developed by a journalist and an economist and they are indeed indexed for inflation. You mention a misrepresentation, but don't actually point one out. You mention a statistical manipulation, but don't point it out. You assert that I should be embarrassed, but don't supply a reason. I have checked the definitions employed by the original authors of the data and they appear sound.

    In a broader sense, you appear very angry and I am not sure why. What is "infuriating" about actually taking a look at the statistics? If anything, I think that it would be more infuriating to realize that you have been misled and manipulated by politicians that paint a picture of your economic condition that is not in accordance with the reality portrayed by the statistics.

  12. Most important, by definition, you can't increase the middle class by decreasing the upper classes, because they are RELATIVE terms. You have to be able to see what's wrong with your conclusion, don't you? You went to college, right?

  13. Hi Liz, please take time to read the post carefully before commenting. The post presents the statistical fact that the percentage of the US households classified as middle class (making between 33K and 100K adjusted for inflation) has fallen and the percentage of US households classified as upper class (making over 100K adjusted for inflation) has grown.

    I did indeed attend college - and two things that I learned in college were 1) to avoid ad hominem attacks, and 2) to back up my opinions with actual facts - or else re-examine my opinions. For the purpose of the original post, middle class and upper class are not relative terms - they are defined by the earnings of the households. When a household earns more than the 100K threshold, it moves from middle class to upper class - and vice versa.

    To directly address your incomprehension, the statistics show that over the last few decades we increased the upper class by decreasing the middle class - and it could be forced to go the other way under certain economic conditions. Please read the post - the classes are set by dollar thresholds, not relative percentages. If 5% of the US populationwas making $105K, but is now making $95K, then the upper class has decreased by 5% and the middle class has increased by 5%.

    If you want to define the classes some other way, that would be OK to talk about - but it is not how the classes are defined in the comment.

  14. I did read the post. The post presents a particular income bracket as being "middle class" without, apparently, realizing that the very nature of the term is relative.

    You continue to make this mistake through your long comment on my "incomprehension," which is still based on the original mistake.

    The ONLY possible reason to try to measure the middle class as an economic band, rather than a relative purchasing power and standard of living, is to play with numbers in the way this post has, in order to justify otherwise unsupported conclusions.

    So I understand how the classes were defined in the comment. That's the problem.

  15. " Please read the post - the classes are set by dollar thresholds, not relative percentages. "

    Argh. I know this. That's what I was pointing out as being ridiculous. It's not really the "middle" if it's set in dollar thresholds. That's like trying to calculate the median without relation to the other values in a set.

  16. Thanks Liz! I think you have done a great job pointing out the strengths and weaknesses of your personal conceptualization of the definition of middle class. Nothing further for me to add here.

  17. You're welcome! And I didn't even get around to mentioning that in creating this "dollar threshold" conception of "middle," the "journalist" who picked the income bands somehow neglected to point out that $99,000 is the threshold of top 5%, while $30,000 and below is actually a little more than 50% of all individual income earners. I wonder why he cherry picked this particular data set? It couldn't possibly be because that's the group that stands in starkest contrast to the top 5%, whose incomes have soared, and so would provide support for his pre-determined (yet still not well-supported) conclusion?

    I am going to assume that political prejudice is causing some otherwise reasonably intelligent people to turn a blind eye to the obvious flaws in this post.

  18. Again Liz, you may want to consider reading the article more thoroughly before posting. The 35K and 99K are household income numbers - not individual - and it appears that 40.12% of households make less than 35K/year.

    The article actually explains the author's choice in that the statistics had been broken into 9 brackets and he chose the middle three brackets to represent the middle class. Try reading the article.

    Also, you seem to be confusing the "percentage of US population earning more than 100K" with "the average earnings of those earning more than 100K". For example, even if the incomes of the top 5% have "soared", it would not change the percentage of US households earning more than 100K. That is, if everyone who earns more than 100K earns on average 500K instead of 300K, it is irrelevant because it does not change the percentage of the population earning more than 100K.

    Finally, you close with another attack - one that is actually pretty funny and ironic because of your obvious, glaring flaws and apparent political prejudice.

    So, to sum up: 1) you don't appear to be reading and understanding the comments and article, 2) you are posting unsubstantiated things as fact - that are actually wrong, 3) you don't appear to be able to logically support your arguements, such as they are, and 4) you repeatedly use ad hominem attacks.
    The bottom line is that it really doesn't appear that you are providing a lot of value to the discussion. Consequently, I am going to have to insist that your future comments be more thought out, logical, and refrain from ad hominem attacks or I may decide that they are not helpful and remove them from the site.