Causal Misatribution, Markets Did This, Not Government

David Brooks has a very wrong-headed column this morning.  My primary problem with it is his usage of the culture of dependency argument, if not the term.  As I've said before, there is very little proof in favor of this argument and a great deal against it.  Let me explain.

According to Brooks:

The Republicans, and Rick Perry in particular, have a reasonably strong story to tell about decline. America became great, they explain, because its citizens possessed certain vigorous virtues: self-reliance, personal responsibility, industriousness and a passion for freedom.

But, over the years, government has grown and undermined these virtues. Wall Street financiers no longer have to behave prudently because they know government will bail them out. Middle-class families no longer have to practice thrift because they know they can use government to force future generations to pay for their retirements. Dads no longer have to marry the women they impregnate because government will step in and provide support.

Let's examine this piece by piece

First of all, financiers.  I shouldn't need to explain in detail that financiers have never behaved prudently without heavy government regulation.  The prudent simply aren't the ones that get rich in this industry.  This goes back centuries to the Amsterdam and London Exchanges.  Tulip mania didn't happen because Dutch investors thought the Republic would bail them out.  It happened because there was money to be made, until there wasn't.  It seems absurd to credit government with something that was happening regularly for centuries before government got in the business of bailing people out.  This is all on the market.

There is something to the middle class thrift argument but there are a few details that have to be noted.  First of all, even during the early years of Social Security there was a very high incidence of poverty among the aged, about 40 - 50%.  (Patterson 78)  So it doesn't seem that even the middle class were saving adequately for retirement before Social Security, there's little reason they would do so today if it were cut.

But market forces and the ensuing cultural changes had more to do with the institutionalization of Social Security than poverty fighting  did.  By the 1920s market rewards had begun to erode the old way of life.  Before then, it was quite normal for at least one son to stay on and take over the farm or the family business and be in a position to take care of elderly parents.  As poverty statistics show, this didn't always work out.  As a result of the growing prosperity linked to markets however, fewer and fewer Americans wanted this.  They wanted to move to new job opportunities and pursue a career different from their parents.  Markets meant that there were opportunities that previously never existed.  These opportunities required individuals to move into a market economy where they generated wealth through their own individual efforts rather than relied on past accumulation by their parents of land and other assets.  People were eager to generate wealth on their own rather than depend on their parents and later have their parents depend on them in turn.  To quote from Patterson (73):

With economic progress in the 1920s, the middle classes had begun to get accustomed to better life-styles.  They had fewer children, and they frequently moved away from their parents.  They did not want to lose what they had by being forced to take in the old folks.  For these people, the allure of old-age insurance was not the benefits they would ultimately receive themselves, which few Americans calculated carefully.  Rather, it was the assurance that they would not have to take care of their parents in the here and now.

In case that wasn't clear enough the changes in American society that brought rise to the pension movement were not some kind of top down change by government, it was a bottom up movement created by changing market opportunities.  We have Social Security because it became increasingly necessary as people transitioned from a local, subsistence economy into an integrated market economy and were exposed to new opportunities.  The government response was a result of not cause of these changes.

Of course, aside from the historical factors, it can also be noted that savings rates are higher in Europe where old age pensions are much more generous than in the United States (the most recent evidence I have for this is the charts in Alesina and Glaeser).  While it may still be arguable that government policy is to blame the policies involved are not those that make current workers pay for retirees.



The third factor mentioned by Brooks I find highly chauvinistic, we can't just assume that it has anything to do with dads no longer being forced to marry the women they impregnate, maybe women are deciding that they don't want to marry the father now that they have other options.  While this may be a development many dislike, given the very high rate of increase in women's labor force participation it seem to be a more immediately plausible change than the idea that government policy has led to this.  The dates work better as well, welfare was very stingy in its early years and these changes began to become large trends in the 50s and 60s, not long after many women, or at least their mothers, had the experience of working for a wage in WWII.  Since the majority of single mothers aren't poor and do not receive benefits from the government it seems likely that their increased ability to participate in the market relative to the past has more to do with this development than any change in government policies.

Let's briefly review some key evidence.  It has long been established that illegitimacy goes along with poverty, while statistics are poor before the state started giving benefits, early studies indicated that one tenth to one third of births to poor black families were illegitimate in 1938 (Patterson 39).  While assistance programs had already developed by the time we start to get good statistics, early welfare programs often excluded illegitimate children from the rolls which makes it doubtful that government policies were linked to the increase in single parent families.  In the 1940s and 1950s these laws were known as suitable home regulations, these laws cut off aid to women in nonlegal unions and conducted midnight raids to check for a man in the house in order to enforce these regulations (Patterson 86).  Despite these restrictions, the numbers of unwed mothers continued to grow during these decades among all ethnic and economic groups.  Among whites, between 1940 and 1960 households headed by women under 55 doubled from 3% to 6%, among blacks from 7% to 20%, despite these strict laws and very scanty welfare benefits (and by no means were all of these women receiving assistance) (Patterson 99).  These rates would increase further   Research on this subject indicated that most of this difference could be explained by poverty rates, while other factors may have been involved there was very little assistance available to these women.  Due to the belief that welfare caused illegitimacy, many studies tried to look for it but found no correlation between benefit levels and illegitimacy rates or number of children per family (Patterson 169).

In addition, according to census data the percentage of female household heads as varied a great deal, from 9.8% of family households in 1940 to a low of 7.9% in 1960 up to 12.2% in 2000 (there is some variation between census and current population survey (CPS) statistics, also, Patterson's statistics are for women under 55 and these statistics are not adjusted for age reducing their value since some households will be female headed due to the earlier death of the husband, finding better numbers I could link to is beyond the time I'm willing to devote to a blog post).  What is notable here is that according to the CPS survey, the percentage of female household shoots up in the 1970s and 1980s, rather bad years for the value of welfare benefits, before stabilizing after that (and slightly declining more recently according to more up to date data).  While I'm not prepared to run any real statistical tests myself, or to look for papers that have already done so, just eyeballing these numbers show that they are fluctuating independently of big changes in government policy.  These numbers are certainly varying, and there are convincing reasons to be concerned about these statistics, but it is very hard to see how these numbers are varying in relation to government policy as opposed to other variables.

While I'm not prepared to take the time to test for causality, it is worth noting that throughout the same period that these rates were rising, women were beginning to not only participate in the labor force in much greater numbers but were also able to achieve greater incomes.  What has changed here isn't that women have grown more dependent, they have clearly grown more independent.  According to Bureau of Labor Statistics (BLS) data labor force participation among women has increased from 35% in 1948 to 55.1% in 1975 and peaked at 76.8% in 1999 (if anyone is wondering, labor force participation was also notably higher among black women than white women until the 1990s, when labor force participation became and remained roughly equal, with 1972 the earliest year for which I have data, 49.9% for white women, 60.0% for black women).  I'm less sure of the causality here than I am with middle class savings, but the trends in family structure seem to precede major government involvement in income support and family life.  Changes in welfare policy appear to be following changes in family structure, not preceding it.  I'm not sure how much of this can be explained by changes in market incentives but the changes here are much greater than they are for government policy.  The increasing options available to women have reduced their dependency on men.  This makes it likely that it isn't so much that men aren't marrying women and being good fathers, it's that women aren't marrying men that they don't think will be good fathers because they have other options.

This has been a long discussion, but I don't think I can emphasize strongly enough how important I think getting the causality of these issues right is.  Many Americans are upset with government not just because they don't like government but because they believe it is causing social issues such as those mentioned by Brooks.

This belief is extremely problematic, I believe it is a cultural factor driving much of the rightward surge we have been witnessing in recent decades as well as the resistance to compromise and polarization that is becoming ever more evident.

Let me briefly explain the mechanism I have in mind.  Over the past few decades there have been a lot of compromises where the left gets some of what it wants and the right gets some small government concessions.  Yet, there seems to be little movement on any of the three issues mentioned by Brooks as being problems with government (though marriage rates are rising among the better off and teenage pregnancy has been falling, savings and cautious finance are nowhere evident however).  Now, this could possibly lead to a refutation of these ideas and a rejection of the belief that government is in fact behind these three things in favor of more plausible explanations.

However, pundits like Brooks reinforce the public's existing biases.  Even though there is not improvement on these factors when small government advocates get their way Brooks, and others, choose to continue to propagate the belief that government is behind these factors.  This confirms the public's prejudices and leads them to conclude that since they are correct in the causal attribution of these problems to government that it must be the compromises that are in fact behind the continued deterioration.  This will make them ever more resistant to the idea of compromise.

This tendency becomes a self-reinforcing spiral.  The market incentives reinforce this, columnists are paid to produce pieces that the public likes to read, reinforcing prejudices is a good way to get eyeballs, challenging them is not (anecdotally, I have been trying to sell several pieces on this to magazines recently and have not yet managed to do so, this failure is overdetermined, I'm an unknown writer and may not have a good feel for my own writing skills, but since pieces that reinforce these ideas are so common and pieces attacking them are so rare, despite overwhelming evidence that these ideas are false, I also can't reject the idea that there simply isn't a market in refuting commonly held but wrong ideas).  But, if it is market forces driving these social changes and not government then the more that small government advocates get their way the worse the problems driving small government beliefs will get.  As long as these ideas are reinforced by cultural outlets the greater the rage and frustration of those holding them and the more extreme their anti-government views will become as the social ills they want to combat get worse even as government becomes weaker. Until these ideas are successfully challenged and exposed as wrong, polarization will continue and the issues driving it will get worse.  Short term compromises, such as those Brooks propose at the end of his column, will only worsen the situation since these ideas will continue to be widely held and the reinforcing spiral will continue as market forces drive the changes that are making people desire small government policies to ever greater extremes.

Misatribution of the causes of social change is resulting in a situation where compromise has become impossible because only one side is ever able to get the results it wants whereas the other side finds that the results it gets are always in the opposite direction that it wants, despite winning policy concessions.  Only long term and sustained pressure for cultural change to properly attribute the causal forces for these problems to the market can have any hope of breaking this spiral.  As it is now, compromise for piecemeal policy changes will only drive polarization as one side's favored policies actively work against its favored outcomes.

[This post got very long so I'm ending it here, I never even got to stable societies and interest groups.  He's mostly wrong here too, look at Italy and its basic political instability as a partial cause of its interest group based pension scheme.  Instability can often drive the creation of interest groups while long term stability can erode their position as people push for more broad based social policies rather than targeted ones.  This often results in higher social spending but it tends to be less targeted to specific interest groups.  Also, interest group formation usually precedes government policy rather than vice versa, though instances of the opposite kind do also exist.]

Sources not linked above: Alesina, Alberto and Glaeser, Edward.  Fighting Poverty in the US and Europe.  Oxford: Oxford University Press. 2004.

Patterson, James T.  America's Struggle against Poverty in the Twentieth Century.  Cambridge: Harvard University Press. 2000.

[edited slightly for clarity and to add references]

[Edit:  Something I wanted to add some detail on, but didn't because the post was getting too long, is that there are very good grounds for criticizing the government's response to all three of the issues mentioned above.  But dealing with them still requires properly attributing causes and the government is clearly responding rather than causing these factors.  Markets as a cause fit the evidence far better.]