The Two-Income Tax Trap:
Yesterday I published in the Wall Street Journal a column entitled the "Two-Income Tax Trap" which was derived from my earlier post here.
I have received a few emails pointing out that in the column, it is unclear what taxes are included in the percentages listed. To clarify, in the book that provides the figures, the percentages (24% in the 1970s and 33% in the 2000s) refer to the total of all federal, state, and local taxes. But the growth in the tax burden between the two periods is largely a result of the growth in the federal income tax portion of the tax bill and the fact that the wife's income pushes the family into a more progressive tax bracket.
In any event, I agree that the tax burden is unfairly distributed.
So how much shielding has eroded in the past 30 years.
They should then divorce and participate in an informal family business (could be web-page design or landscaping) for which they adjust the joint income between them annually so as to minimize their tax bill. Nobody could possibly second-guess their division of labor and income. Of course they should buy their household furnishings, liquor and cigarettes in the state with the lowest sales taxes.
Then they need to buy a fixer-upper for the stay-at-home member of the pair to improve and then sell every two years, earning income for labor that is 100% tax-free up to $250,000 every two years. They should do their own plumbing, electrical and auto-mechanical work, realizing that they will have to earn $200 to pay a plumber $100 for $50 worth of work, assuming everyone is in a 50% tax-bracket (realistic with all the fed, state and fica tax).
Accumulation of wealth nowadays involves learning how to game the tax system. It's unfortunate that the folks least capable of mastering these simple concepts are the very ones most likely to breed.
Agree that the marriage penalty is a problem, but trying to spin it into an overarching criticism or the tax system as a whole is a stretch.
Hm. Surely the Reagan boom in military spending and the current adventure in Iraq have nothing to do with the increase. Why would taxes bear any relationship to expenditures?
Yes, SS taxes are capped, but so are the benefits. If both are uncapped, SS's financial problems get worse (and we get to read "exposes" like "Ross Perot collects $100k is SS benefits", ignoring the fact that such payments will be a poor return on his payments). If only the former is uncapped, SS becomes a significant cost with no benefits for high-income people.
For years, SS advocates have argued that SS is not a welfare system and that its popularity depends on not being a welfare system. Were they wrong? How will SS be affected when high-income people see SS a welfare system?
The portion of the taxes that we pay for our military are quite low by historic standards. The portion of our taxes that constitute robbing Peter to pay Paul have never been higher.
The public sector amounted to 28.8% of GDP in 1972 and 30.1% of GDP in 2002. Thats an increase of 4.5%. At the same time (according to Warren and Tyagi), the tax burden on the typical family grew from 24% to 33%, an increase of 37.5%. So part of the story is not that taxes have dramatically increased over the last 30 years, but that the tax burden has shifted considerably.
Another interesting tidbit is that, although the tax burden increased on typical families, marginal income tax rates for the married couples discussed by Warren and Tyagi, actually decreased substantially over the same period, from 45% in 1972 to 27% in 2002. Check out:
Except that, using the 2000 tax rates and assuming this is a family of four using the standard deduction (They would do better if they itemized given their mortgage) they are in the 28% bracket for their taxable income over $43,850 and their total federal income tax bill would be $8,089.50 (roughly 12% of their overall income) Even if you add their share of the FICA tax (6.45%) you are still under 20% of their income going to tax. Yes, you can add state sales and local property and income taxes as well, but those vary from state to state. They still would be unlikely to even approach 33%.
I hate it when people confuse tax bracket with overall tax rate.
Tax burden (GPO, not census)
http://www.gpoaccess.gov/usbudget/fy05/sheets/hist15z3.xls
Tax rates
http://www.taxfoundation.org/taxdata/show/151.html
It is true that the marginal rate is 25% at that level, but the lower tax rate on the income below that level accounts for the lower overall percentage.
Here are the federal tax rates:
http://www.irs.gov/formspubs/article/0,,id=150856,00.html
"unlikely" What in heaven's name are you talking about? Try computing for CA, NY, or NJ.
On top of that your little hand-wave about FICA being 6% is grossly misleading. FICA is effectively 13%. Its a pretty fair statement that if FICA didn't exist, we'd be getting another 11% in net pay.
Honestly: most of the kicker is in FICA which as you might recall was lower in 1970 than it is now.
I suppose this is probably true. But, realistically, we should consider not just the family's actual pay, but benefits as well, most of which *are* compensation and *are not* taxed. Health insurance would be the main benefit here, with retirement-type benefits being the next largest source. (And you need to include this for the 70's family as well, of course).
Could you show your math on the family's federal marginal income tax rate?
By my calculation, the family's marginal federal income tax rate was 15%, and their effective federal rate was about 5.6% (when you counted the standard deduction, as well as exemptions and child tax credits).
My back-of-the-envelope calculation was that the family paid $3766 in federal income taxes, which is 17% of their total $22,374 tax burden. If you gave them a 33% cut in their marginal rate, from 15% to 10%, you would cut $1392.50 ($43500 taxable income minus $15650 times, the bottom of the current 15% bracket, times 5%), or 6%, off their total tax burden. Their effective federal income tax rate would be 3.5% ($2373.50/$67000).
So you have to cut the marginal federal income tax rate by 33% just to shave 6% off the family's total tax burden. Reducing the federal income tax rates wouldn't do much to help them.
You can argue fairly about trickle down, but reducing marginal federal tax rates is a very inefficient way of directly helping middle income and poor people. And directly helping poor and middle income families is your goal there's little reason to mess with anything above the 15% bracket.
Increasing the child tax credit allows you to target relief to the people you say you want to help most--families. That means you wouldn't inadvertantly be giving benefits to 100K-a-year DINK couples, allowing you to give even more relief to your hypothetical $67K-a-year family in the 15% bracket.
Unfortunately, I can't imagine even trying to explain it to the American populace, let alone getting them to support it.
Ah, but that would mean all of those tax layers and CPA's would be out of their jobs AND the public would better understand how much of their money is being stolen from them by the government.
But I can dream, can't I?
1. The biggest problem is with this:
Where do they and you come up with an increase in health insurance premiums for a family of four over the past 30 years of $620/year? I submit this is simply not believable in the absence of specific detailed information as to how it was arrived at.
What seems much more likely is that $1,030 in annual premiums in the 1970's became $1,650 in monthly
premiums in the 2000's. That just might have a significant effect on household finances and bankruptcy.
2. Let me point out the SS tax rates and incomes subject to those rates changed significantly over the period.
1970s rate of 4.8 - 6% on $7,800 to $23,000
2000s rate 7.65% on $76,000 to $97,000(current)
3. I don't see where you and the authors got their tax numbers, as least with respect to the federal component.
effective federal tax rates 2004
quintile pretax household income 2004
Middle Quintile $54,200 - effective tax rate all federal taxes 13.8%
Fourth Quintile $79,300 - effective tax rate all federal taxes 17.4%
effective federal tax rates 2002 and 1079
1979, constant 2002 dollars
Average household income total effective fed tax rate
Second Quintile $31,100 14.3%
Middle Quintile $46,700 18.6%
2002
Average household income total effective fed tax rate
Middle Quintile $51,100 14.4%
Fourth Quintile $75,900 18.7%
I don't see that actual data support your claims. And I say this as a person whose total tax burden for many years was greater than all other spending combined.
I speak as someone who once had a boss who earned considerably more than me, but who didn't believe in birth control and therefore had such a large family that he paid no income tax. I paid quite a bit, out of my lower salary. I do not approve of his behavior and see no reason why I should subsidize it.
People who fail to reproduce should certainly be denied social security and medicare benefits. These benefits for the older generation are financed by taxing the younger generation, and anyone in the older generation who has not raised at least one member of the productive younger generation should not be permitted to get a free ride on the system. Do your duty and you should be rewarded. Enemies of posterity who fail in their duty should get nothing!
Maybe because we need the children to pay taxes and serve in the military when we are old.
What do retirement-type benefits look like these days?
401(k) deductions are on 15% of income (up to 15k). Employer match is added to that, but all the money (including earnings) will be taxed when withdrawn.
Traditional IRA adds another 4k deferral (unless you earn too much).
And I'd say ALMOST no loopholes. A few things for the very poor, and I'm actually willing to consider a "head of household" arrangement for people with dependents (as long as it isn't generous enough to wipe out all tax liability for an upper middle class family).
I was reacting to Public Defender's comment assuming that people all want tax relief to go to families, defined as those with children still at home. It's quite possible that the 100K DINK couple mentioned so disparagingly is working two jobs each to keep elderly relatives in better conditions than Medicaid provides. Or they could be saving up to have kids in the future, or to put their grandkids through college. Or they could be blowing half their income on booze. It doesn't matter -- it's not the government's business to decide which of these activities merit an income tax break and which do not, although it's important to note that the boozehounds will end up paying more of a different kind of tax.
Why not? If the government is going to tax at all, it is certainly their business to use it as a tool to reward / punish certain behaviors. The state has a definite interest that citizens reproduce - otherwise the state would soon cease to exist - and thus this behavior should be subsidized / encouraged from a government perspective.
I suspect that this notion of parents with a prolific brood of children that wipes out all tax liabilty for an upper middle class family is as much of a straw man as the "welfare queen". In any case, your erstwhile boss with a ton of children had a crapload of child-related expenses that undoubtedly entirely negated the "enormous subsidy" he received via the child tax credit. It could even be that his disposable income after child-related expenses was less than yours, despite his much higher gross income, so enough with the tax credit envy already. Most parents will laugh in your face if you tell them that their tax credit represents an unfair subsidy! They are doing one of society's most important jobs, and all you can do is bellyache because they get a few miserable tax breaks? Tchah.
For what it's worth, the Tax Policy Center (which generally has very solid data), provides a useful table here: I believe the data from the 1970s come straight from the Treasury.
The tables provide time series (1955-2006) of historical average (and marginal) tax rates at three points in the income distribution: 1/2 median income, median income, and twice median income. In all cases the data are for a family of 4, as in the WSJ column and the Warren book. The median income level ranges from 63k to 73k for the 2000-06 period, in nominal dollars, so this appears to be an appropriate comparison for your example for recent years.
Helpfully, there are two tables: one that doesn't include the employer share of payroll taxes and one that does.
(Technical side point: the figures provide in the TPC excel document show the same income regardless of whether the employer share is included. A more appropriate approach would be to add the dollars corresponding to the employer share to the income base. Of course, that would reduce the corresponding average tax rate. Thus the figures I'm citing here are overestimates of the true average tax rates in the including-employer-share case. For simplicity I'll ignore this point, as it just makes my claims even stronger.)
Let's first consider the table that does not include the employer share. For the 2000-06 period, the "Average ... Combined Federal Income and Employee Social Security and Medicare (FICA) Tax Rate" for a 4-person family with median income ranged between 13-16 percent, with a simple average of 13.84 percent.
For the 1970s, average combined federal tax rates for a median-income family of four ranged from just under 13 percent to just under 17 percent. The simple average was 14.50 percent.
Now let's turn to the estimates that include employer shares of payroll taxes. For the 1970s, the avg tax rates range from just under 16 percent to just over 23 percent, with a simple average of 19.20 percent. For the 2000s, the rates range from 20.64 to 23.32 percent. The simple average is 21.49 percent.
In sum, if we ignore the employer share of payroll taxes, the average federal tax rate for a 4-person family whose income equals the median actually fell by 2/3 of a point between the 1970s to the 2000s, from 14.50 to 13.84 percent. If we include the employer share, the average federal tax rate for such a family did increase, but only by 2.29 points, from 19.20 to 21.49 percent.
Now let's consider state and local taxes. The only way to get to both 24% for the 1970s avg total tax rate and 33% for the 2000s avg total tax rate is to have the avg state/local income tax rate rise. Focusing on the case including the employer share of payroll taxes, the avg state/local income tax rate would have had to be 4.8 percent in the 1970s and 11.84 percent in the 2000s. And that implies an increase of 7 percentage points in combined state/local income taxes, which amounts to more than three times the increase in the federal share over this period. I haven't yet found a tim series for avg state tax rates for a 4-person family with median income in each year, but I'm skeptical that there has been anything like that size increase over the 1970s-2000s.
Thus, the data suggest two basic conclusions:
* A 2000s tax rate of 33% seems too large. Even using the figures that include the employer share, we'd need avg state and local taxes to be nearly 12 percent. There are some states that have *marginal* income-plus-sales tax rates in this neighborhood, but few would assess an *average* rate of 12 percent on this family. I doubt that the *average* state and local tax liability for such a family is close to 12 percent. And obviously if we don't include the employer share of payroll taxes, the 33% figure is way overstated.
* Prof. Zywicki's claim in this post that
the growth in the tax burden between the two periods is largely a result of the growth in the federal income tax portion of the tax bill and the fact that the wife's income pushes the family into a more progressive tax bracket
seems very unlikely to be true. Even granting his/Warren's figure of 33% for the 2000s, the federal tax share would have grown by fewer than 3 percentage points. Moreover, the state share would have to have risen by 7 percentage points, about 3 times the increase in the state share.
Based only on these figures that Prof. Zywicki cites from the Warren book, I am skeptical of both the book's figures and, therefore, Prof. Zywicki's critique of it. That's not to disagree with his point concerning the silliness of talking about taxes only in percentage terms and other expenses in dollars. I certainly agree with that point.
But frankly I think it would have been a good idea for Prof. Zywicki to look into the details of the tax figures they used before simply accepting—and promoting—them as correct. Perhaps the figures I cited above are somehow inaccurate or incomplete (tho I doubt it given the source). But looking in to them—-and being able to provide a precise source—-would seem like Prof. Zywicki's obligation to his readers, not his readers' obligation to him.
Seems like a clarification might be warranted in the pages of the WSJ.
I don't know why it has to be this way, but parents (of minor children), the elderly, and homeowners are the only people whose claims for better treatment get any social recognition at all. I think this is counterproductive. For example, if tax rates on people with no children YET are too high, those people might not be able to reach a financial position where they CAN have children. Especially if they do not wish to rely on special benefits themselves. Similarly, if you subsidize mortgage interest payments but not the savings of a renter who wants to amass a hefty downpayment before buying, you send the message that debt financing is preferable to saving up for what you want. With predictable results. These are tax traps far more fundamental than the one that started off this discussion.
So, what we have now is people -- very often a large mass of people -- who happen to be in the "preferred" group at the moment using the political system to vote benefits for themselves. The cost of these benefits falls on people currently in the non-preferred group. It's like one of those two-tier union contracts, where the established workers get a good deal but new hires come in with fewer benefits and protections.
This does not go down well people whose position in the life cycles isn't the the culturally dominant one at any particular point. There are the people who fell between the WWII generation and the Baby Boomers, and then there are post-Boomers. People in small generations may just look at these things differently from people in the gigantic, pig-in-python cohorts. We always assume the good stuff will go away right about when we would stand to benefit.
I've looked again at the TPC table I linked to in my previous comment. If you use the CPI-U series to inflate 1970s dollars to 2003 levels, you find that median income for a family of 4 in 1973 was typically a good bit higher than the $38,700 figure provided by Warren and Tyagi, as cited in Zywicki's earlier post (their definition is "the median income for a fully employed man in 1973"). Median income for a family of 4--i.e., rather than "for a fully employed man"--was $56,816 in 2003 dollars. Thus, all of the discussion in my first comment should be considered with that in mind.
The TPC table also provides the estimated income for a family with one-half the median income level in each year. That figure for 1973 was $28,408. The federal tax rate for a family with that income was 3.33 percentage points lower than the federal tax rate for someone with the median income. Considering instead the federal rate inclusive of employer-share payroll taxes, the rate was 2.09 points higher. For the entire 1970s decade, the differences average 4.33 points (without the employer share) and 3.37 points.
Presumably the average federal tax rate for a family with $38,700 in income in 1973 was somewhere between the average for a family with $28,408 and a family with $56,816 (both in 2003 dollars). So this discrepancy probably accounts for a reduction of 2-3 points in the 1973 family's tax rate, which would make the federal change between 1973 and the "modern-day" (Warren/Tyagi's word) family somewhere between 3-5 points.
That said, I don't think this is an apples-to-apples comparison. Why compare a below-median family in 1973 to a median family in the 2000s? The proper comparison is median to median. So there's another way in which Warren and Tyagi's data are poorly presented. (For what it's worth, I'm guessing that the problem is they just added the median full-time male income to the median full-time female income for the "modern-day" family, rather than using the median income for a family of four. Because medians are nonlinear functions of the entire income distribution, the median of a sum does not equal the sum of medians. I could certainly be wrong on the guess, though.)
In sum, I think the proper comparison *is* median in the 1970s to median in the 2000s. Moreover, averaging over the entire 1970s avoids any cherry-picking resulting from a single year (i.e., 1973).
So, I think the numbers in my previous comment are more sensible to use than those that Warren and Tyagi use. (Obviously this isn't Zywicki's fault, though it makes his substantive criticisms of what they do with those number less interesting.)
I've looked again at the TPC table I linked to in my previous comment. If you use the CPI-U series to inflate 1970s dollars to 2003 levels, you find that median income for a family of 4 in 1973 was typically a good bit higher than the $38,700 figure provided by Warren and Tyagi, as cited in Zywicki's earlier post (their definition is "the median income for a fully employed man in 1973"). Median income for a family of 4--i.e., rather than "for a fully employed man"--was $56,816 in 2003 dollars. Thus, all of the discussion in my first comment should be considered with that in mind.
The TPC table also provides the estimated income for a family with one-half the median income level in each year. That figure for 1973 was $28,408. The federal tax rate for a family with that income was 3.33 percentage points lower than the federal tax rate for someone with the median income. Considering instead the federal rate inclusive of employer-share payroll taxes, the rate was 2.09 points higher. For the entire 1970s decade, the differences average 4.33 points (without the employer share) and 3.37 points.
Presumably the average federal tax rate for a family with $38,700 in income in 1973 was somewhere between the average for a family with $28,408 and a family with $56,816 (both in 2003 dollars). So this discrepancy probably accounts for a reduction of 2-3 points in the 1973 family's tax rate, which would make the federal change between 1973 and the "modern-day" (Warren/Tyagi's word) family somewhere between 3-5 points.
That said, I don't think this is an apples-to-apples comparison. Why compare a below-median family in 1973 to a median family in the 2000s? The proper comparison is median to median. So there's another way in which Warren and Tyagi's data are poorly presented. (For what it's worth, I'm guessing that the problem is they just added the median full-time male income to the median full-time female income for the "modern-day" family, rather than using the median income for a family of four. Because medians are nonlinear functions of the entire income distribution, the median of a sum does not equal the sum of medians. I could certainly be wrong on the guess, though.)
In sum, I think the proper comparison *is* median in the 1970s to median in the 2000s. Moreover, averaging over the entire 1970s avoids any cherry-picking resulting from a single year (i.e., 1973).
So, I think the numbers in my previous comment are more sensible to use than those that Warren and Tyagi use. (Obviously this isn't Zywicki's fault, though it makes his substantive criticisms of what they do with those number less interesting.)
I've looked again at the TPC table I linked to in my previous comment. If you use the CPI-U series to inflate 1970s dollars to 2003 levels, you find that median income for a family of 4 in 1973 was typically a good bit higher than the $38,700 figure provided by Warren and Tyagi, as cited in Zywicki's earlier post (their definition is "the median income for a fully employed man in 1973"). Median income for a family of 4--i.e., rather than "for a fully employed man"--was $56,816 in 2003 dollars. Thus, all of the discussion in my first comment should be considered with that in mind.
The TPC table also provides the estimated income for a family with one-half the median income level in each year. That figure for 1973 was $28,408. The federal tax rate for a family with that income was 3.33 percentage points lower than the federal tax rate for someone with the median income. Considering instead the federal rate inclusive of employer-share payroll taxes, the rate was 2.09 points higher. For the entire 1970s decade, the differences average 4.33 points (without the employer share) and 3.37 points.
Presumably the average federal tax rate for a family with $38,700 in income in 1973 was somewhere between the average for a family with $28,408 and a family with $56,816 (both in 2003 dollars). So this discrepancy probably accounts for a reduction of 2-3 points in the 1973 family's tax rate, which would make the federal change between 1973 and the "modern-day" (Warren/Tyagi's word) family somewhere between 3-5 points.
That said, I don't think this is an apples-to-apples comparison. Why compare a below-median family in 1973 to a median family in the 2000s? The proper comparison is median to median. So there's another way in which Warren and Tyagi's data are poorly presented. (For what it's worth, I'm guessing that the problem is they just added the median full-time male income to the median full-time female income for the "modern-day" family, rather than using the median income for a family of four. Because medians are nonlinear functions of the entire income distribution, the median of a sum does not equal the sum of medians. I could certainly be wrong on the guess, though.)
In sum, I think the proper comparison *is* median in the 1970s to median in the 2000s. Moreover, averaging over the entire 1970s avoids any cherry-picking resulting from a single year (i.e., 1973).
So, I think the numbers in my previous comment are more sensible to use than those that Warren and Tyagi use. (Obviously this isn't Zywicki's fault, though it makes his substantive criticisms of what they do with those number less interesting.)
I've looked again at the TPC table I linked to in my previous comment. If you use the CPI-U series to inflate 1970s dollars to 2003 levels, you find that median income for a family of 4 in 1973 was typically a good bit higher than the $38,700 figure provided by Warren and Tyagi, as cited in Zywicki's earlier post (their definition is "the median income for a fully employed man in 1973"). Median income for a family of 4--i.e., rather than "for a fully employed man"--was $56,816 in 2003 dollars. Thus, all of the discussion in my first comment should be considered with that in mind.
The TPC table also provides the estimated income for a family with one-half the median income level in each year. That figure for 1973 was $28,408. The federal tax rate for a family with that income was 3.33 percentage points lower than the federal tax rate for someone with the median income. Considering instead the federal rate inclusive of employer-share payroll taxes, the rate was 2.09 points higher. For the entire 1970s decade, the differences average 4.33 points (without the employer share) and 3.37 points.
Presumably the average federal tax rate for a family with $38,700 in income in 1973 was somewhere between the average for a family with $28,408 and a family with $56,816 (both in 2003 dollars). So this discrepancy probably accounts for a reduction of 2-3 points in the 1973 family's tax rate, which would make the federal change between 1973 and the "modern-day" (Warren/Tyagi's word) family somewhere between 3-5 points.
That said, I don't think this is an apples-to-apples comparison. Why compare a below-median family in 1973 to a median family in the 2000s? The proper comparison is median to median. So there's another way in which Warren and Tyagi's data are poorly presented. (For what it's worth, I'm guessing that the problem is they just added the median full-time male income to the median full-time female income for the "modern-day" family, rather than using the median income for a family of four. Because medians are nonlinear functions of the entire income distribution, the median of a sum does not equal the sum of medians. I could certainly be wrong on the guess, though.)
In sum, I think the proper comparison *is* median in the 1970s to median in the 2000s. Moreover, averaging over the entire 1970s avoids any cherry-picking resulting from a single year (i.e., 1973).
So, I think the numbers in my previous comment are more sensible to use than those that Warren and Tyagi use. (Obviously this isn't Zywicki's fault, though it makes his substantive criticisms of what they do with those number less interesting.)
My main point is that his solution is a very inefficient way of directly helping his hypothetical $67K family of four. But, respectfully, I don't think that's his real goal. I think his real goal is a marginal tax reduction for people in 25% and higher tax brackets. And plenty of respectable trickle-down arguments support that position. But Zywicki's $67K-family-of-four hypothetical is misleading rhetoric, not serious argument.
As to the suggestions that we should increase the tax burden on middle and lower class families with kids, just try making that your platform. Remember, a flat tax would have to be 5.6% just to keep this family at the same level they have with the current system.
You could exempt the first $30K or so of income, but then the rate would have to be around 11% for this family to break even. For singles with $155K+ taxable income, an 11% rate would be a 67% decrease in their marginal rate (from 33% to 11%).
You try selling to voters a 67% tax cut for $155K singles while giving nothing to Zywicki's hypothetical $67K family of four.
I have been waiting for someone to address the confiscatory taxes paid by married couples for a very long time, and I am particularly surprised that no one has identified it as a huge issue for women. When you look at what happens to a married woman's income when it is perched atop that of well-paid husband's, it essentially takes a 50% whack.
This is just plain wrong for the vast majority of married couples. Look at the tax rates for singles and married couples. Up to $123K joint taxable income, the tax brackets for married couples are exactly double that for singles. That means that unless each partner makes exactly the same income, married couples with less than $123K joint taxable income get a marriage bonus.
Even above $123K taxable income, married couples where one partner makes more than another can still get the marriage bonus.
Now for my second follow-up comment, in which I'll focus on median income for a family of four for the 1970s and median income for a family of four for the 2000s. The Tax Policy Center has another useful table.
This one is titled "State and Local Tax Revenue as a Percentage of Personal Income". It does have a couple drawbacks, in that it concerns the average state and local tax revenue share for everyone in the economy, rather than the taxes that would be paid by the relevant family of 4 with median income. The figures here are a bit higher than that family would face in one respect, due to some states' use of progressive income taxation. On the other hand, very high-income households will have considerable savings in a typical year and thus lower-than-average sales tax obligations. On the third hand, higher-income folks generally have more expensive houses and thus higher property taxes relative to income.
So it's not clear exactly how to adjust the figures in this table. For the sake of argument, I'll take them as correct as given.
The second disadvantage of this table is that it provides data only for selected years between 1977-2005 (I assume this has to do with the relatively infrequent fielding of the census of governments).
With those caveats in mind, the table shows that total state and local tax revenues were 10.81% of personal income in 1977 and 10.73% in 2005. (For completeness, the other figures are: 10.41% (2004), 10.20% (2002), 10.54% (1997), 10.47% (1992), 10.29% (1987), 9.63% (1982). These figures are a bit higher than I'd have guessed; my bad.
Adding the 1977 figure of 10.81% to the 1970s figures that include the employer share, as reported in my first comment, one would get a total combined 1970s tax rate of 19.2% + 10.81% = 30.1%. (As I noted in my first post, the figures I cited from TPC include the employer share of taxes but don't add those dollars to the income base. As above, I'll ignore that issue here.)
The average state-plus-local tax as a share of personal income for the years 2002, 2004, and 2005 is 10.45%. Adding this figure to the average federal rate of 21.49% in my first comment, one gets a 2000s tax rate of 31.94%.
Two observations, then:
1. The 2000s tax rate of 31.94% is quite close to the 33% figure provided by Warren and thus used by Zywicki.
2. The 1970s combined fed/state/local rate of 30.1% is within 2 percentage points of the 2000s-era tax rate.
As such, unless there is some other, better source of data that shows radically different trends in state/local taxes for a family with median income, the increase of 9 percentage points discussed by Warren and Tyagi and discussed in great detail by Zywicki in the WSJ piece is just not a very useful number.
In this earlier post, Zywicki uses Warren and Tyagi's figures and writes
Instead of using Zywicki's/Warren-Tyagi's 24% and 33% figures and below-median income of $38,700 for 1973, let's rewrite this text using the data discussed above and in my first two comments. That is, we use the average level over the 1970s of median income for a family of four, expressed in 2003 dollars, which is $55,908, and we use 30.1% for the 1970s family's tax rate and 31.94% for the 2000s family's tax rate. We also use the average level over 2000-06 of median income for a family of four, expressed in 2003 dollars, which is $65,497 (slightly lower than Warren and Tyagi use note that their number is quite cose, contrary to my previous comment's guessed criticism of it; my bad). Here's what we get:
It's worth noting the following points:
1. The total increase in taxes is less than 1/3 of the number that Zywicki finds. Of course, some of this difference is due to the fact that my 1970s family has greater income than Warren/Tyagi's, and thus Zywicki's. Holding the tax rate constant, the family would of course pay more *dollars* in taxes. Which brings us to....
2. If one insists on using the $38,700 example provided by Warren and Tyagi for the 1970s family, then a substantial chunk of the increase in taxes paid will be due to the fact that the family's position in the income distribution has improved, moving from below the median to the median. (In this case, Zywicki's comment regarding progressivity would be correct.)
3. In fact, of the $4,092 increase found by comparing medians to medians, only (0.3194-0.301)*$5,5908 = $1,029 is due to the increase in the family's effective tax rate. The rest, which amounts to 74.86% of the total increase in taxes paid, is due to income growth at the same point in the income distribution.
4. Zywicki's conclusion that "the increase in tax obligations is over three times as large as the increase in the mortgage" (his bold) might no longer be correct. No doubt my median income family of the 1970s would spend more on housing than did Warren/Tyagi's below-median family. The question is how much.
In sum, based on the data I can find, both the Warren and Tyagi example and the Zywicki takedown seem to be unilluminating at best, and made of straw at worst.
Huh? No one says the woman has to stay home with the kids. If the wife makes enough money, the husband can stay home. I have a friend who lost his job just about the time his school-teacher wife had their baby. He decided to stay home and raise the kid. They get by on her school-teacher pay.
Feeling Pinched in NJ, do you make more than a school teacher? If so, your husband can stay home and take care of your family. You may have to live in a smaller house, eat out less, and own only one modest car, but I bet many people successfully raise a family on far less than your pay.
Regardless of the upper income marginal tax rate is 35% or 25%, raising children will be very stressful if both parents want to keep demanding, high-pay jobs.
Don't blame the tax code for your choices.
Pinched in New Jersey here. You assume that I have a demanding, high paying job, but in my case, I did want to be with my kids as much as possible. This is a second marriage for me and my husband, with my 2 and his 3 kids being products of the first marriages. With the guidelines in the New Jersey being quite generous, his kids want for nothing. I had to go back to work when mine were young, my first husband having a more modest salary, and paying relatively little in the way of support. It came as an enormous shock to me to have 52% of my salary withheld, and not to qualify for any deductions. Now that my kid are college age, the feds look at my husband's salary, even though he is not responsible for their college education, and they are considered not to have need. If we are so wealthy, why is it that all we do is work? Looking at what we make, I would have assumed we could afford a mansion, but we live in a 3 bedroom house between two houses that are in such bad shape they will probably be demolished. I realize my situation may be somewhat different from the norm, but then again, maybe not. All I know i that after 20 years of part-time work that I gladly took so I could be with my kids, I could make twice as much and still be only marginally closer to being able to pay tuition. The tax code is like a thresher, it brings in the wheat, but does not see the nests of young it demolishes.
But I bet he pays less tax on his salary than he would if he were single. As a unit, do you two pay more in joint taxes than you would pay if you both filed individual federal income tax returns? Even at high incomes, you can still get a marriage bonus if the income disparity is high enough. And even if you have a marriage penalty, is it more than the economies of scale of combining two households? Finally, if you want to be treated as individuals for income tax purposes, do you agree to pay estate taxes if he dies first like you would if you were single? Are you willing to be treated as singles for health insurance purposes?
The college aid problem is not a problem of marginal federal income tax rates. College aid officials see marriage as a partnership, and justifiably so. I don't see why students who have no parents with six-figure incomes should get less financial aid just because one parent, who voluntarily chose marriage knowing that parenthood was part of the deal, also chose to be "not responsible" for the college education of the children in his new family. But again, even if you disagree, this is not a problem of marginal federal income tax rates.
Your post also shows what's likely the real problem--divorce imposes enormous financial penalties on the ex-spouses. Again, that's not a problem with marginal federal income tax rates.
And going back to the topic of this post, if there's any "problem" in the marginal rates, it's for people making substantially more than Professor Zywicki's hypothetical $67K family of four.
Leaving aside the specifics of Warren &Tyagi's argument as characterized by Zywicki, there are some nontrivial conceptual issues in thinking about what the proper income levels to compare are. Above, I said apples-to-apples meant median-to-median. But perhaps things are more complicated.
Economists (I plead guilty) are used to thinking of voluntary choices as indicating preference. So, the fact that so many families have added 2nd earners between the 1970s and now is evidence that they think they are better off with 2 earners than one. Note that this does *not* necessarily mean that people feel better off with two earners in the 2000s than they felt with 1 earner in the 1970s. Rather it means simply that they feel better off with 2 earners in the 2000s than they would with 1 earner in the 2000s, and better off with 1 earner in the 1970s than with 2 earners in the 1970s. Comparisons across time are where the difficulty arises.
Now, Zywicki writes in his first post that
The first sentence is certainly true. Of course there are other things that "conventional economic theory" might say. For example, one reason why 2nd earners enter the labor force might be that relative wages for women appear to have risen. Economists call this a positive substitution effect, and it's a reason to be happy about 2nd earners' labor market entrance.
Another reason 2nd earners might have entered is that they might not have felt able to make it on just one income. Economists call this a negative income effect, and it's a reason to lament 2nd earners' labor market entrance.
Sorting out how much of the change is due to positive substitution effects and how much to negative income effects ain't easy. And, I suspect that much of the debate over whether the economy has been "good" or "bad" for "typical" people boils down to this issue.
Frankly I don't think the within-couple insurance issue to which Zywicki points in his second sentence quoted above is as big a deal. That's not to say that 2 (probabilistically independent) sources of total income X aren't better than 1 source of income X for risk-averse people. It's just that I'm not sure that's the big issue about the growth in two-income families. (Moreover, the point as Zywicki states it requires showing that savings actually increase when people move from 1 to 2 earners; if the only change is that consumption increases, then there is no precautionary savings to discuss.)
Returning to our main issue of finding a 1970s apple to compare to a 2000s apple, it's not at all clear how to compare one-income families in the 1970s to two-income families in the 2000s. Should we focus on families that have chosen lower income and one earner in the 1970s, or those that have chosen higher income and two earners? If we choose the first kind of family, we are likely to end up with something like Warren and Tyagi's family of four with total 1973 income of $38,700, below the family median. But if we choose the second kind of family, we are likely to end up with something closer to the overall median of $55k+. If we go with the first case, taxes will increase more than discussed in my previous comments simply because total income has increased, and also to some extent because of federal income tax progressivity (though I'm guessing W&T's tax-rate figures, and thus Zywicki's, are understated a bit for this 1970s case). If we go with the latter case, then the right comparison is what I offered in the comments above.
How you answer this question really depends on whether you see the change from one- to two-earner families as a response to wonderful new facts--higher wages for women--or unfortunate new facts--slow growth in male wages and rising relative prices of health care, housing, etc.
In other words, it's difficult to answer the factual questions without making a choice about proper counterfactuals that is self-reinforcing. Definitely a tricky question.
A.C., you may not like it, but many of us think it is pathetic and sad that you regard having children as a "lifestyle choice" like renting vs. buying a house or driving a hybrid vs. an SUV. Taxes exist to fund the government? Well, what is the government FOR? Ultimately, the purpose of government is to ensure a stable environment in which families can successfully raise children. Without children (and properly raised ones at that), society and the government would quickly cease to exist, and would be purposeless and irrelevant anyway. The government has a great moral and practical interest in helping people have children and raise them properly, to include tax breaks. The "neutral" or default government policy should be to assist families, not to give them "equal" treatment with the childless, who can and should look out for themselves.
I don't know why it has to be this way, but parents (of minor children), the elderly, and homeowners are the only people whose claims for better treatment get any social recognition at all.
That is exactly as it should be in a civilized society! That is what civilization does - it helps children and the elderly. I am astonished that anyone could regard these attitudes as baffling.
Your examples of "tax traps" are rubbish, and proceed from faulty premises.
So, what we have now is people -- very often a large mass of people -- who happen to be in the "preferred" group at the moment using the political system to vote benefits for themselves. The cost of these benefits falls on people currently in the non-preferred group. It's like one of those two-tier union contracts, where the established workers get a good deal but new hires come in with fewer benefits and protections.
The government wants and should want people to join the "preferred group"! There is nothing morally illegitimate about the government encouraging people to reproduce and buy homes. It is also a sound pragmatic policy - it is something the government should be doing. The government has no business at all subsidizing or encouraging those who refuse to reproduce (i.e. selfish hedonists) or leveling the playing field between those who reproduce and those who do not.
This does not go down well people whose position in the life cycles isn't the the culturally dominant one at any particular point.
I love the assumption here that wanting to have and raise children, to own homes, and to grow old securely are mere social constructs and transient artifacts of the currently "dominant culture." These are not epiphenomena peculiar to the USA in 2007, but basic human impulses, and good luck creating some sort of "neutral" social regime that does not favor them. If you did create such a regime, in which (at minimum) having and raising children was not encouraged, it would be a disaster. However, such an anti-natalist regime would be relatively short-lived, because a pro-natalist society would eventually move in and take over. A society that did not require the young to assist the old would be odious. (Plus one has to ask, you may not plan to have children, but do you not plan to get old at some point?)
There are the people who fell between the WWII generation and the Baby Boomers, and then there are post-Boomers. People in small generations may just look at these things differently from people in the gigantic, pig-in-python cohorts. We always assume the good stuff will go away right about when we would stand to benefit.
I am not in one of those "big demographic cohorts," and the way I look at it, government policy absolutely should favor parents with minor children, homeowners, and the elderly. Anyone who does not successfully raise children has fundamentally failed as a human being. Anyone who does not think society needs to help parents with children and the elderly isn't much of a human being.
What we have is not a syustem to raise revenues. What we have is a tax system design to change/control behavior.
Abolish the income tax.