Almost a year ago, Zero Hedge looked at the trend in US tax refunds, and we found that last year the government was doing everything in its power to accelerate the remittance of refunds to taxpayers. Back then we said that "one of the primary reasons why consumers may have exhibited an abnormal propensity to spend in January and most of February (at least according to government, if not Gallup, data), is the much greater individual tax refunding conducted by the Treasury/IRS this year compared to the prior year." But if accelerated tax refunds was the story in 2010, in 2011, at least so far in the year, it is precisely the opposite. In fact, to date the IRS has refunded nearly $20 billion less compared to 2010, and about $14 billion less than in 2009. With consumers suddenly having far less cash, does this mean that February and March are set to be major disappointments from a retail sales perspective, and any other vertical having to do with consumer "strength"?
Using Daily Treasury Statement data, we compile the weekly data for the first 13 weeks heading into April 15 (the point by which most refunds have been issued), and find that both average weekly remittance and 2011 cumulative refund issuance is running at a nearly 20% lower run rate than 2010.
Below we chart the weekly data for 2009-2011...
For anybody but the Koolaid master at Goldman Sachs (who we are confident will have a rebuttal to our thesis within 72 hours), this is certainly a troubling development.
On the other hand it may simply indicate that US taxpayers have maxed out on their withholding exemptions, and as a result are due far less from the IRS, as we suggested early in 2010.
BNY's Nicholas Colas has some additional perspectives on why a nearly 20% cumulative shortfall has developed between 2010 and 2011 refunds, as well as why tax withholding data indicates that the NFP estimate of +185,000 for this Friday may be woefully otpimistic. His observations below:
The logical question about this shortfall is simple: “Why?” There are a few potential answers that relate to both the timing of filing/refunds as well as the total amount that may eventually be refunded.
- Incentives to file early have declined in 2011. The Homebuyer Tax Credit expired on June 30, 2010. According to the IRS’s data, some $3.6 billion of the +$200 billion in refunds last year were triggered by this incentive to purchase a first home. This credit was extended to all home purchases in 2010, but with only half a year of eligibility there are many fewer tax filers who have a strong incentive to get their paperwork in early and receive their $7,500.
- Lower Refunds to Collect in 2011. The American Recovery and Reinvestment Act (ARRA) of 2009 lowered payroll taxes by $400/person or $800/couple for two years – 2009 and 2010. The program began in early/mid-2009, so only 2010 saw the entire effect of these lower withholding amounts. Keep in mind that these were not reductions in taxes – they were simply reductions in withholding, and result in lower refunds, everything else equal.
I think it is too early to know if tax refunds will prove substantially lower in 2011, although the early data presented here is not especially promising. In truth, neither explanation above is particularly robust. The homebuyer tax credit might be a $1-2 billion shortfall this year (half of its contribution in 2009). ARRA might be a total of $13-15 billion for this whole tax season ($100/worker incrementally lower withholding in 2010). The troublesome point is that in just the first seven weeks of 2011 we are already short the already-mentioned $21.7 billion to last year’s cadence – far more than the total of these two potential contributing factors. All that is left by way of explanation is that taxpayers just haven’t gotten around to filing as quickly as in past years. But the trend in recent years has been to faster filing, not slower.
We will keep monitoring this data through April 15th, but it seems clear that consumer spending patterns will face an incremental headwind on top of higher oil prices as we roll through March and early April.
In addition to tracking the refund data, the U.S. Treasury Daily Statement is also a useful dataset for estimating labor market trends. That’s because the Treasury gives us a very granular look at personal income tax and withholding data. The majority of Americans work “on the books,” so their employers are responsible for submitting the funds required to satisfy each workers tax withholding and other payments to the government such as Medicare/Medicaid, Social Security and Unemployment Insurance.
The attached chart shows the trends through February 22 and is a useful reality check as we think about what next Friday’s Jobs Report might bring:
- February personal tax/withholding receipts are down 0.5% year on year, adjusted for 2 percentage points of reduced withholding in 2011 for lower Social Security payments. Unadjusted, those receipts are down 4%.
- This is the first negative comparison since June 2010. The data is choppy, to be sure, reflective of the sluggish labor market recovery we’ve seen over the past few months.
It therefore seems that current consensus estimates of 185,000 private sector jobs added in January is very optimistic. Yes, that number is based on a survey of just a few thousand households, but the disparity between the tax/withholding data is quite wide. Consider that the Labor Department currently estimates that there were 129,281,000 people in the workforce last January, and 130,229,000 in December 2010 (seasonally adjusted). If there were really another 180,000 people employed in February, wouldn’t tax receipts rise year over year? The math says they should. Also keep in mind that last month’s lackluster 36,000 jobs added came with a modest increase in tax/withholding receipts. Why should we expect job growth to increase in February to +180K if the actual tax/withholding amounts collected are lower than year-ago levels.