When it comes to measuring Americans’ income and wages, one can use the brute, seasonally-adjusted shotgun approach applied by the BLS which takes aggregate data and applies statistical analysis to infer monthly data, or one can actually go bottoms-up and look at credit and debit card spending data coupled with changes in bank deposits to determine net changes in household wealth with near Swiss-watch precision.
The latter is precisely what Bank of America has done: in its retail sales preview, BofA’s US economist, Michelle Meyer has used “anonymized and aggregated payroll direct deposit data” from Bank of America customer deposit accounts in the US to capture the turn higher in after-tax wages, if any.
Well, as BofA’s “chart of the month” below shows, there was a whopping 2.0% increase in the yoy pace of after-tax wages and salaries growth – rising over 5% – and was roughly 2x higher than what the BLS reported for that month.
So all else equal, Americans are suddenly earning a lot more in after-tax income – arguably thanks to the Trump tax cuts which have substantially reduced tax withholdings. And, simply extrapolating historical income-spending correlations, one would assume recent spending has similarly soared as well.
Well, one would be wrong, because the same Bank of America, this time using aggregated credit and debit car data, has found that retail sales ex-autos actually declined 0.2% in February. Strangely, this marks three consecutive months of sluggish retail sales readings after the spectacular performance in November. What is also notable is that while official Census data has tracked the BAC figures, it has recently run above the BAC data, which according to BofA suggests that the market is vulnerable for some softening in the Census data this month as well, as seasonally adjusted approximation slide back to reality.
So what can explain this divergence between a surge in income and a slowdown in spending? Here BofA makes several argument, suggesting notable cross currents for the consumer in February – namely delayed tax refunds for lower income households but take-home pay was higher as withholding schedules were adjusted to reflect the tax legislation.
Here is what BofA found: for taxpayers who claim the Earned Income Tax Credit (EITC) or the Additional Child Tax Credit (ACTC), the IRS instituted a tax refund hold which delayed the payment of refunds until late in February, similar to last year but in contrast to the trend in prior years. This is consistent with the IRS notice that households receiving the Earned Income Tax Credit or Alternative Child Tax Credit would see a delay. Here BofA did observe a sharp gain in refunds during the last few days of the month, but this is unlikely sufficient to underpin spending in the month. This likely means that “seasonal factors are therefore looking for relative strength which may be creating a downward bias this month.”
The delay in tax refunds disproportionately impacted lower income households given that the limitation is based on those earning the EITC and ACTC. This shows up as weaker spending for the lower income cohort, particularly for debit cards which are a proxy for cash. Moreover, some households receive the tax refund right to their debit card which therefore directly impacts the ability to spend. Meanwhile the lower income cohort actually saw a pickup in credit card spending which could be a sign that the constraint on disposable cash led to greater leverage for the month.
Another way of observing the hit to lower-income, refund-dependent households is the distribution of spending between debit and credit cards by income cohort. To no surprise, there is an inverse relationship with income level and debit card usage. For those households earning less than $50K, about 70% of spending is done on debit cards. In contrast, households earning more than $125K, debit cards make up only 45% of spend. As such, the decline in debit card spending among the lower income cohort which we show in Chart 3 more than offsets the increase in credit card spending for the same cohort.
At the same time, as IRS refunds slowed, employers changed withholding schedules in February, and according to the CBO, employers are likely to withhold $10-15bn less from employees each month. This is precisely the boost to after-tax wages and salaries shown in the chart on top.
Putting it all together, as delays in tax refunds weighed on spending, especially among lower-income households – those earnings $50K and lower, on aggregate income rose disproportionately for higher income Americans ($75K and more) as consumer saw a offset through a boost to their paychecks.
So what did they spend money on? Not surprisingly in a world in which the rich got richer, and the poor did not, consumer spending on cruises saw the biggest gain in February, increasing 4.2% mom after a 1.0% drop in January according to BofA card data. Similarly, there was impressive spend on lodging.
Yet oddly, luxury spending declined in the month, showing a pullback from January’s solid number, whiuch BofA believes was helped by timing of Chinese New Year.
Finally, as shown in the charts below, only two sectors showed yoy declines: home goods and sporting goods. Note that the prior 5 year % yoy average for sporting goods is also negative, at -1.5%, indicating a continued structural decline.
Looking ahead, BofA expects the delayed payout of tax refunds coupled with the increase in tax-home pay to provide a boost to spending. And while BofA expects a disappointing retail sales report tomorrow, it believes this will be the transition month and is more optimistic about future spending.
Finally, if BofA is correct about the magnitude of the aggregate after-tax wage acceleration, it is only a matter of time before aggregate prices spike and set the Fed even more behind the curve.