Arnold Kling  


Shlaes on the New Deal... Medicare, the U.S., and France...

Robert E. Rector, Christine Kim and Shanea Watkins write

Overall, households headed by persons without a high school diploma (or low-skill households) received an average of $32,138 per household in direct benefits, means-tested benefits, education, and population-based serĀ­vices in FY 2004. If expenditures for interest and other financial obligations relating to past government activities are added to the count, expenditures rise to $36,989 per household. If the cost of public goods is added, annual total expenditures on benefits and services come to $43,084 per low-skill household.

...By contrast, low-skill households paid only $9,689 in taxes.

The rest of households pay an average of $34,629 in taxes and receive $30,819 in benefits.

It is difficult to trace through their methodology, so I do not know whether the following quibbles are relevant:

1. It is possible that the authors over-estimate expenditures for schooling for low-skill households. The expenditures on schools for low-skill households are probably low, because they probably tend to live in school districts with low expenditures levels.

2. It is possible that the authors over-estimate Social Security and Medicare benefits for low-skill households, in the sense that the cohort of 70-year-olds today includes more people without a high-school diploma than the cohort of young workers today. If low-skill and high-age are correlated, then part of what the analysis is picking up is the fact that old folks are tax-eaters. In fact, what this whole exercise cries out for is a lifetime analysis, not an annual snapshot analysis.

Having raised the quibbles, my guess is that the main thrust of the analysis is almost surely correct. The distribution of benefits and taxes is such that low-skilled people are net winners and high-skill people are net losers. That's a progressive system for you--what would be shocking would be to find the opposite. What is more interesting is looking underneath the totals to see where the progressivity comes from. On the benefit side-means-tested aid is important (low-skilled workers get a lot of Social Security and Medicare, but presumably highly-skilled workers do as well). On the tax side, it would seem that income and property taxes are most important.

The pointer is from Byron York, who says that the data support taking a dim view of unskilled immigrants. However, that is where the annual snapshot is inadequate as a tool of analysis. The fact that 70-year-olds without high school diplomas are receiving Social Security benefits does not tell me to keep 35-year-olds without high school diplomas out of the country.

Comments and Sharing

CATEGORIES: Income Distribution

COMMENTS (5 to date)
Viscus writes:

Who cares how many dollars someone pays. That doesn't really go to their sacrifice. Imagine a high-earner earns $850 per hour consulting.

The real measure of contribution is how many hours does it take you to earn the money that you pay.

Let us say that a relatively low-income household payed $9,500 in taxes. Let us say that such a household, the salary of those working was $10 per hour. It would take 950 hours for that low-income household to pay their taxes.

Compare to the high-earner. He had regular consulting work, about as much as wanted. Let us say that he pays four times as much in taxes, say $40,000. At his pay rate, that would require only 48 hours of work.

So, even though the high earner is paying four times as much, he only has to work 1 hour for every 19 hours that the less well off family does to pay their taxes. So, not only does the less well off family make less money, they have to work much harder to pay their taxes. Also, the money that the lower income family pays is much more valuable to them than the money that the high income individual pays. It is clear that the low-income family is making a much bigger sacrifice than the high-income individual in this example.

I am not saying that this example is correct in terms of the taxes paid. It is conceptual, not empirical. But it does illustrate the flaw of measuring contribution via money. For some people, $850 is earned in 1 hour. For others, that amount is earned after 100 hours of what is probably much less pleasant work.

The better measure of relative contribution is hours worked to pay taxes. Not amounts.

Floccina writes:

I find the $32,138 interesting in that I believe that the poor get far less value than that out of the transfer. I think that most would be far better off with an annual check for $10,000. When nationalized pre-school was a hot topic I was working in a restaurant and having an economics background, I polled the single mothers who worked in restaurant about what the would rather have X dollars extra a week or free day care, since most had a relative or swapped with a friend the average would rather have an extra $20/week than free day care. Now the state would probably spend about $100/week per child for daycare and get $20/week value for it.

We pay a high price for not trusting the poor with money.

When it comes to education:

1. The poor tend to get little out of it.
2. Some people those who now send their children to private schools find government education worthless at 0 cost.

We pay a high price for not trusting the poor with money.

John Thacker writes:

The better measure of relative contribution is hours worked to pay taxes. Not amounts.

OK fine, Viscus. I first question whether you read the post, since in the first place the poor get back more in benefits than they pay in taxes, unlike other people. That means that even if they paid more hours of income in taxes, they would get back even more hours of income in benefits.

However, in any case the poor pay a smaller amount of taxes by the hours worked measurement as well. This is because the tax system is progressive, and part of what it means. Consider:

1) The rich pay a higher percentage of their income in taxes, and

2) The rich work more hours on average than the poor. (This is a relatively new development in the history of civilization, the rise of the often-highly educated rich workaholics instead of the idle rich landowners. For example, you mentioned consultants in your post. The consultants I know work tremendously ridiculously long hours for their enormous salaries.)

1) implies that, if people actually worked the same number of hours, the rich would obviously be paying more hours of work worth of taxes as well.

2) implies that even if they actually paid the same percentage of their income in taxes, the rich would be paying more hours of work because they'd be paying that same percentage on a greater number of hours.

Combine facts 1) and 2), and you see that the rich certainly pay a higher number of hours of their labor worth of taxes.

Steve Sailer writes:

This was studied before, by the National Academy of Sciences in 1997, and they concluded that an immigrant with no high school diploma costs the taxpayers about $100k lifetime over what he pays in.

TGGP writes:

Arnold, you seem to be hoping that immigrants will assimilate to the average. In fact, for Mexican immigrants things get worse in some ways. Crime rates and illegitimacy are both higher for second than first generation, and higher for third than second. Education levels do increase from first to second, but hit a plateau below the national average. The problem is not per-pupil expenditure, which has doubled accounting for inflation in 30 years and is far higher in D.C than Wyoming or Montana (the latter of course have higher test scores). What is being created is a new permanent underclass, and judging by the interstate migration rates of native-born americans, it is certainly not desirable to live near. The policies the government has enacted in response to the previous underclass did an awfully poor job of moving them up (Bryan Caplan and Charles Murray both argue they made things worse).

What I am very concerned about is the political impact immigration will have. Bryan Caplan sees public opinion as vitally important to public policy. He also notes that people who are less educated, poorer and less educated tend to be more statist. The political culture is Mexico is awful. We are beginning to see it seep into the U.S. The new underclass will be motivated to vote for more entitlements for themselves (and polls show they want to). The Democrats will likely move away from courting well-off "knowledge workers" and the Republicans will try to convince latinos they represent "family values" and further shunt aside a small-government focus. Perhaps the middle-class will react to dehomegenization by rejecting "Swedish model" type policies that seem to rely on homogeneity for support, but demographic growth rates are not on their side. It seems to me more immigration means the U.S becoming more like Mexico, and like the immigrants themselves I prefer to live in the United States. I asked Bryan before how he could reconcile these two beliefs of his and I am still waiting for an answer.

Viscus, it's a good thing nothing can ever be proven morally, otherwise you wouldn't be able to freely yammer on about "sacrifice" and "contribution" sans any reference to any data without it being possible to determine if what you say has any truth. I don't know if you were trying to state that we should permit more low-skilled migration because it would be good if A: the United States got more hours "sacrificed" or B: it would decrease how many hours they would have to "sacrifice" in Mexico, but it almost seemed to me as if you didn't bother to read Arnold's post (which is norm-free until Byron York comes in) before you spouted off.

Comments for this entry have been closed
Return to top