In some sense when trying to analyze the difference between an ill-suited measure and that same ill-suited measure seasonally adjusted amounts to Isaac D’Israeli’s critique of Thomistic tendencies toward using otherwise useful brainpower for pointless ends (thus, how many angels can dance on the point of a needle – not the head of a pin as the modern colloquial version has it). In other words, this may all be useless drivel at the end of the day.

However, given that these are the “best” tools currently under employ, we might as well try to make sense of them particularly in the very needed search for corroboration of other accounts and data points. In this case I am speaking of the corruptive notion of “inflation” as well as its modern moniker the CPI. Neither the definition for it nor the equations under which it is derived are particularly well-adapted to what we are really trying to measure, not the least of which is due to the clumsy inclusion of monetary effects (harmful) with non-monetary productivity (we need more and more of this, not to be offset by “money printing”).

The primary use for inflation is not for the sake of scholastic argument about QE and rational expectations theory adaption, but rather the very real and powerful difference between price changes and wages. It is rather simple in that if prices rise faster than wages there will be significant economic distress (Japan is proving that mightily), to the point that wages can even rise at the same pace or even slightly better and malaise will still result. True economic advance is marked by periods where wages (a proxy for true wealth creation) rise far faster and more broadly than prices (though ideally prices should be more or less stable).

The bad news for orthodox economists hoping to tease out some kind of as-yet “recovery” from five years of this trend is that “real” wages fell in September. Worse still, real wages have been trending lower throughout this “robust” payroll expansion contrary to all expectations.

ABOOK Oct 2014 Wages Deflated NSA

That is very similar to the trend I see in consumer spending, where price changes in 2014 are very different than price changes in 2013; leading to a case of attrition beyond the nominal focus. Again, if real wages are growing more slowly or falling, that would certainly seem to fit within a lower “real” spending environment.

ABOOK Oct 2014 Retail Sales 13v14 ex autos

Given the behavior of other indications of spending, like food prices and revenue of companies in that business, there is a good probability that the CPI understates the pace of change and how that is corrosive toward economic function.

However, seasonally-adjusted real wages are trending in the opposite direction making this kind of narrow interpretation perhaps murkier in extrapolations.

ABOOK Oct 2014 Wages Deflated SA

While discrepancies like this between the two viewpoints will occasionally occur, this is actually something that is only now deepening. Unadjusted real wages have been trending lower since 2013, whereas seasonally-adjusted wages have largely been flat and now rising in the past two months.

ABOOK Oct 2014 Wages Deflated

In that respect, it would seem to falsify the dominant economic narrative anyway particularly since that should be indisputable in far more than just the Establishment Survey. That includes historical perspective, and the difference apparent in that comparison.

ABOOK Oct 2014 Wages Septembers

On an unadjusted basis, wages in September 2014 are not just below September 2013, they are about even with September 2007. That is a pattern that has been replicated, or close to it, in other indications such as Gallup’s track of consumer spending – flat in nominal terms over the past year and a half or so (meaning lower in real terms) and below 2008 and prior.

ABOOK Oct 2014 Gallup Spending

Certainly one month’s discrepancy is as likely statistical noise as anything else, but this is not simply apparent in only September. That raises the issue as to what exactly the adjusted figures are showing and why.

But since real wages even in the adjusted data have not risen all that much, only $7.50 per week across seven years, the “improvement” there might just as well be statistical noise since it amounts to such a paltry increase in the first place; only 2.2% total or about 0.6% compounded annually. Wage growth next to prices has been so small (or nonexistent) that highlighting the differences here really does amount to perhaps needless fine points.

That probably highlights best this “recovery” period, where the tiniest expressions of “growth” are debated as seemingly significant since there has been nothing like real progress through all of it anyway.