The form below is the result of combing through PDF files from irs.gov to get an accurate picture of how federal payroll withholding tables changed through the years. I later tacked on tax tables to compare with the payroll tables to see how they lined up (which they do pretty well... with a couple exceptions that I'll get to). I've populated data from the years 1998 to 2013, and added a simple HTML5 interface for charting the results.
Choose a salary from the slider (if you don't see a slider control, the charts won't work with your browser - the HTML5 "range" input type currently only works in Webkit browsers and Opera), and then click married or single, to see the chart for that salary.
The code makes some basic assumptions for the sake of simplicity. First, the salary is taxable income, after other deductions such as 401k, social security, etc. Second, the tax line adjusts the salary down by the standard deduction (twice that for the "married" category, assuming "married filing jointly" is the filing status), where the withholding line does not. This appears to be the intent of the descrepencies between withholding and tax tables, that is, separating what your employer knows about you from what the government does. Your employer doesn't know about your mortgage interest, whether you'll file joint or separately, whether the child tax credit applies to you, how much you spent for college tuition, or any of the hundreds of other deductions or credits you'll talk to Uncle Sam about. They know what you contribute to 401k, how many dependants you intend to claim, and that's about it.
While putting the tables together, I noticed some pretty odd things, and with some digging on senate.gov and gpo.gov, I was able to reconcile them all with the public laws they originated from. Let's start in the present and move back in time.
Note: I tend to avoid political commentary in this blog, but in this case numbers and politics are tightly intertwined, so some of my liberal ranting is inevitable. If that bugs you, feel free to ignore it all and scroll down to the charts.
First, in 2013, there is a very narrow salary range for single filers between the 35% and 39.6% brackets. Here is the tax table in question:
Salary Base withholding % of excess
$ 0.00 $ 0.00 10%
$ 8,925.00 $ 892.50 15%
$ 36,250.00 $ 4,991.25 25%
$ 87,850.00 $ 17,891.25 28%
$ 183,250.00 $ 44,603.25 33%
$ 398,250.00 $ 115,553.25 35%
$ 400,000.00 $ 116,165.75 39.6%
The 400k figure is from the Taxpayer Relief Act of 2012, which specifies the very round figures of $400,000 for single and $450,000 for married as where the 39.6 percent tax bracket should be reintroduced. It dovetails with Obama's "pay your fair share, Mitt Romneys of the nation" reelection ads, preys on a common misunderstanding of how tax brackets work, and is completely ineffective at its stated goals.
The misconception is that, under this table, if someone makes over 400k, their tax rate is 39.6 percent. It isn't. If it was, the base withholding in the top bracket would be 39.6% of $400,000: $158,400, instead of the 116k and change listed in the table. It's only the salary after 400k that gets taxed at the higher rate. The $1,750 before that is taxed at 35%, the $215,000 before that is only taxed at 33%, etc. If you make exactly 400k, your effective tax rate under this table is a smidge over 29%, more than 10% less than the big figure you imagined was being used to stick it to the man.
It fails at its stated goal of making the rich pay a "fair share" (which we'll define as what they typically paid before the Bush tax cuts), as you can clearly see in the charts below by cranking the salary to the top, and comparing 2013 with 1998. It also fails at delivering on its name: providing taxpayer relief, as the other tax brackets remain essentially the same.
The next oddity happens in 2010, where there are two more payroll withholding brackets than there are tax brackets. This comes from the American Recovery and Reinvestment Act of 2009 (ARRA), otherwise known as the Stimulus Bill (or the Bailout, depending on your ideological leaning). The ARRA was purportedly enacted to cover the gap in private spending with more government spending. I don't want to go into whether or not that's right thinking, because I don't know. There are two basic camps about why the stimulus didn't fix anything: people who think John Keynes was a quack, and people who think that the bill didn't go far enough (most famously, Paul Krugman, who estimated the spending gap was three times what the bill was covering).
Oddly, the GPO link to the bill's text is broken, but here is a link to the (4 megabyte) senate conference report, which includes details about the extra withholding brackets. In short, they relate to the $400 Earned Income Credit introduced in 2009. From pages 517 and 518 of the report:
Income tax withholding
Taxpayers’ reduced tax liability under the provision shall be expeditiously implemented through revised income tax withholding schedules produced by the Internal Revenue Service.
...
The provision provides eligible individuals a refundable income tax credit for two years (taxable years beginning in 2009 and 2010). The credit is the lesser of (1) 6.2 percent of an individual’s earned income or (2) $400 ($800 in the case of a joint return).
So basically the tax tables remained static, and withholding was lowered to get the $400 in everyone's hands quicker than waiting for a tax return. Since you only get the credit if you have a job, then that should work out for everyone. Except for small businesses without tax lawyers and a payroll IT department on hand, who would have to spend money to get the new rules implemented in the middle of a tax year.
At any rate, the charts below reflect the intent. Until around $80,000 for single and $140,000 for married is reached, you'll see a dramatic drop in withholding in 2009 (the new tables were released after the tax year began, so this makes sense), which tapers in 2010, then lines back up with the original curve in 2011 and forward.
What the IRS came up with in 2010 to line up payroll withholding with the $400 credits is nothing short of bizarre. Here are those tables:
Single
Salary Base withholding % of excess
$ 6,050.00 $ 0.00 10%
$ 10,425.00 $ 437.50 15%
$ 36,050.00 $ 4,281.25 25%
$ 67,700.00 $ 12,193.75 27%
$ 84,450.00 $ 16,716.25 30%
$ 87,700.00 $ 17,691.25 28%
$ 173,900.00 $ 41,827.25 33%
$ 375,700.00 $ 108,421.25 35%
Married
Salary Base withholding % of excess
$ 13,750.00 $ 0.00 10%
$ 24,500.00 $ 1,075.00 15%
$ 75,750.00 $ 8,762.50 25%
$ 94,050.00 $ 13,337.50 27%
$ 124,050.00 $ 21,437.50 25%
$ 145,050.00 $ 26,687.50 28%
$ 217,000.00 $ 46,833.50 33%
$ 381,400.00 $ 101,085.50 35%
The new withholding numbers were inserted between the 25 and 28 percent brackets. Look closely at those percentages. Single brackets go up from 25, to 27, to 30, then back down to 28. Married brackets go from 25 to 27, back to 25, then 28. Madness. There must have been a more mathematically sensible way to accomplish the same thing. Anyway, moving on.
In 2001 and 2003, withholding and tax tables had the same number of brackets, but their percentages don't match. After the 15% bracket, 2001's rates have a half percent difference, and in '03 that rate jumps to 2%. By 2004, there is an extra tax bracket, and the maximum rate drops from 39.6% to 35%. This is a result of the Bush tax cuts, introduced as two separate bills in 2001 and 2003.
2001 brought us the Economic Growth and Tax Relief Reconciliation Act, which laid out this graduated change in tax rates:
In the case of taxable years The corresponding percentages
beginning during calendar year: shall be substituted for
the following percentages:
28.0% 31.0% 36.0% 39.6%
-----------------------------------------------------------------------
2001 27.5% 30.5% 35.5% 39.1%
2002 and 2003 27.0% 30.0% 35.0% 38.6%
2004 and 2005 26.0% 29.0% 34.0% 37.6%
2006 and thereafter 25.0% 28.0% 33.0% 35.0%
If I understand correctly, the intent was to give everyone more cash to go buy stuff, which would, despite there being no net change in the GDP, somehow grow the economy faster, make everyone richer in a few years, and give the government a net gain in revenue, which would have the national debt paid off by 2010.
The bill was signed into law in June of 2001, and before it could be put to the test, 9/11 happened, which, as you well know, had a large, negative effect on the financial health of the United States.
Rather than re-assess, Bush doubled down with the Jobs and Growth Tax Relief Reconciliation Act of 2003, which sped up the transition to the new tax rates, changing the text of the original law to contain this graduated table:
In the case of taxable years The corresponding percentages
beginning during calendar year: shall be substituted for
the following percentages:
28.0% 31.0% 36.0% 39.6%
-----------------------------------------------------------------------
2001 27.5% 30.5% 35.5% 39.1%
2002 27.0% 30.0% 35.0% 38.6%
2003 and thereafter 25.0% 28.0% 33.0% 35.0%
If you look at the charts below, playing with the salary and filing status, you will see one unmistakable fact: After 2001, the government collected less money from everyone. Much less. The current estimate is that close to $1.4 trillion in lost government revenue was caused by these cuts, on par with the cost of the Iraq and Afghanistan wars.
The original bill in 2001 was lauded by the Heritage Foundation in this report, peppered with language like "phase out the death tax", and including the following defense of the idea, an apologist response to a prior study by the CBPP (emphasis mine):
The Center on Budget and Policy Priorities (CBPP) estimates that the Bush plan would reduce tax revenue by up to $2.1 trillion and increase interest payments by $400 billion over 10 years. This estimate is very misleading because it includes a number of provisions that are not in the Bush tax plan and because it does not recognize that economic growth and the tax base will increase at a faster pace when tax rates are reduced.
They had a separate report to expand on that idea called, simply enough, "Why the Center on Budget and Policy Priorities Is Wrong About the Cost of Bush's Tax Plan". As it turns out, they were both wrong, but the CBPP was at least estimating in the right direction.
Bush and those supporting his direction were pretty bad at numbers, it seems. During his first term, the Congressional Budget Office was still estimating the Iraq and Afghanistan actions as having a final tab in the low billions, as shown in these two reports:
http://www.cbo.gov/sites/default/files/cbofiles/ftpdocs/46xx/doc4683/10-28-spratt.pdf
http://www.cbo.gov/sites/default/files/cbofiles/ftpdocs/33xx/doc3362/afghanistan.pdf
My conclusion is that it must be pretty hard to think in the midst of all that leadership.
I'll leave you with this last observation: the tax table brackets are inflation-adjusted every year. The effect is, taking out the anomalies of the Bush tax cuts and the $400 credits in '09 and '10, a downward curve in tax revenue. Not only does the money the IRS collects every year have less buying power than the year before due to normal inflation, but they're deliberately collecting less each year.
In addition, with the exception of the newly re-added 39.6% bracket, the Bush tax cuts are now permanent, costing the Federal government billions each year in lost revenue, which if not creatively repealed under the Byrd rule, will eventually cause government services to be reduced (increasing unemployment), and possibly scuttling Social Security. Me, I'd rather lose some disposable income and have the government continue to operate.
No comments:
Post a Comment