The article is honest and fair, discussing mostly the issues Bernanke faces and his opinion on them. I thought, while I'm at it, I'll throw in my opinion on the matter.
The first order of business is understanding inflation. Inflation is, as the great Friedman said, always the result of increased money supply (in the long run). This is something people should remember when they discuss inflation, since the huge recent upswing in gas prices is not because of an increase in the money supply, but politics and market dynamics.
Inflation in the US is measured by the Bureau of Labor Statistics (not the Fed), and is a very involved process. It should also be noted that there are many different measures of inflation. It is not a cut-and-dry indicator. For example, there is the consumer price index (CPI) that measures inflation as experienced by the consumer. There is the Producer Price Index (PPI) that measures inflation as experienced by producers, and the Employment Cost Index (ECI), which measures inflation in labor costs. Within each of these broad categories are many different baskets and groups of inflation measurement.
Just so you know, the CPI-U is the broadest most commonly version of inflation. The U stands for Urban, and CPI-U is a broad measurement of the inflation experienced by urban consumers (which is different from poor consumers or rural consumers or wealthy consumers). To quote from the BLS website,
Which index is the "Official CPI" reported in the media?
Our broadest and most comprehensive CPI is called the All Items Consumer Price Index for All Urban Consumers (CPI-U) for the U.S. City Average, 1982-84 = 100.
In addition to the All Items CPI, BLS publishes thousands of other consumer price indexes. One such index is called "All items less food and energy". Some users of CPI data use this index because food and energy prices are relatively volatile, and these users want to focus on what they perceive to be the "core" or "underlying" rate of inflation.
That last paragraph is important. The Federal Reserve is one such user that bases its decisions on the "All items less food and energy" measure. To quote from the Federal reserve:
[P]olicymakers examine a variety of "core" inflation measures to help identify inflation trends. The most common type of core inflation measures excludes items that tend to go up and down in price dramatically or often, like food and energy items. For those items, a large price change in one period does not necessarily tend to be followed by another large change in the same direction in the following period. Although food and energy make up an important part of the budget for most households--and policymakers ultimately seek to stabilize overall consumer prices--core inflation measures that leave out items with volatile prices can be useful in assessing inflation trends.
What does that mean? It does not mean the Fed is trying to hide a multitude of sins by using an inflation measure that does not include energy and housing costs. It is a statistics argument. Let me explain.
The Fed must look into the future. Have you ever done that? It's hard. Very hard. It's even harder when what you're trying to predict doesn't even have a clear indicator. In order to look into the future, the Fed must employ elaborate statistical/econometric models that use current and historical data.
Data can vary dramatically in its usefulness both in statistical terms and in econometric terms. What do I mean? From a strictly mathematical and statistic perspective, if a variable (i.e. the inflation measure) has too much white noise (movements that are not associated with past or future values; I'm using this term loosely by the way, the statistical definition is much more... involved) then it provides less predictive information. There are several methods for trying to remove the white noise and assessing the trend, but if the white noise is too erratic, you will likely find no trend at all. If this is the case, your variable is, alas, useless. From an econometric perspective, it may make sense that the variable (gas prices) be included in the model (gas prices do, in fact, effect our living expenses), but if it makes the whole model useless, you must get rid of it.
So, my astute reader, you will quickly see that price fluctuations in gasoline are very dramatic (relatively). For example, just two months ago gas was around $2.35 a gallon (ha, Utah has its benefits), but it is now up around $3.60. That's a full 49% price increase in less than two months. If inflation was 49% in the US, I'd recommend getting as many credit cards as possible and maxing them out on cigarettes, whiskey, diet Coke, and ammo, since that's what you'll need in the ensuing barter system. However, I am not recommending such action, nor would anybody argue we are actually experiencing 49% inflation.
Some may argue that on average if we include gasoline along with all the other things, we would be closer to reality. Statistically speaking, this is not true at all. The variability of gasoline makes it a very abnormal commodity. You do not see the price of your furniture, or your tv, or your cars, or books, or dishes, etc. change by more than 2% on average. And that's the point.
So it is not political maneuvering or deviousness on the Fed's part, but rather a realization of the limitations of statistics and data that has led to the use of the "core" inflation measure. The extreme variation in gasoline prices would literally make it impossible for the Fed to even attempt to predict inflation.
We can gnash our teeth at them and shake our fists as much as we like, but until somebody provides a reasonable way to filter out the white noise of erratic commodities and expenses, we cannot include them in the Fed's inflation predictions. Sorry. (as a side note, I don't think it's possible to "filter" the erratic nature of gasoline prices since they are not, strictly speaking, white noise, but rather reflections of the erratic nature of politics, markets, and many other things, maybe even butterfly wings)
So with that discussion of inflation completed, I return to the case at hand. Bernanke's issue is complex, and he is aware of the fine line he must tread (as he himself admits in the article I linked to but you probably did not read). The main issue is to unwind the $2.9 trillion in reserves on the Fed's balance sheet without killing the economy. Unwinding too fast could cause the mortgage market and other credit markets to freeze since prices would fall dramatically, among other things. Not unwinding could be disastrous if banks suddenly (as in, over a couple months) unleashed that $2.9 trillion to the world through lending, thus leading to horrific levels of inflation.
The San Francisco Fed's chairman, Dann Bowman, back in October, discussed this issue when he visited Salt Lake. Some argue that as soon as inflation starts to pick up (i.e. banks unleash a torrent of lending suddenly), the Fed cannot stop it because they would be stifling the economy. This is a misleading argument. Technically speaking, yes, the Fed would be limiting the credit supply, and thus, yes, technically speaking, they would be hindering economic expansion. But this makes it sound as though they would plunge the economy back into recession. This is false. Very false. Please don't think like this.
The current recovery is not based on an exorbitant or abnormal amount of lending. Does that make sense? What I mean is, we are currently weakly recovering, and even though interest rates are uber low, we are not actually lending out a ton of money. People are saving and reducing their debt (which is why we've had to lower interest rates so much just to get some lending going). So if lending suddenly picks up, putting the kabosh on super loose credit markets won't kill the economy again (within reason; if interest rates were hiked too fast it would be a problem, but there is a relatively wide margin to get this right). What it will do is allow lending to increase but at a more reasonable rate. As we can see, we do not need excessive lending for the economy to continue its recovery, and in fact we don't want excessive lending to fuel the recovery. We just want some lending to keep recovering.
Bowmann's point was that we can easily stop up the credit market if it begins to unwind too fast. The concern that the Fed won't do it is bordering on conspiracy theories, since the Fed has every reason and incentive to do so, and to do so properly. Furthermore, the Fed does have the tools to detect that inflation quickly. In fact, the fed has more data and more up-to-date information than ever before. In the '70s, inflation measures were less accurate and less immediate. Before World War II they were difficult at best to get publicly. I'm not saying we have it right this time, or that this time is different, but I am saying we are in a better position than any other high-inflation period.
One final note on the topic is that Ben Bernanke has been thinking about this very issue from the start. Before he even implemented the quantitative easing, he requested express permission to pay banks interest on their reserves at the Fed. This provides incentives for banks to leave their money in reserves, since it's earning interest there. Though this has complications and isn't strictly what the Fed wants to keep forever, it provides extra time for the Fed to unwind its balance sheet slowly.
But oh how easy and fun it is to blame the Fed, and to create and propagate apocalyptic predictions of inflation. Bill O'Reilly and Glenn Beck may disagree with me, but they're selling air time. They're trying to make the Fed's "failures" look like Obama's failure, which is fine I guess, but it's not genuine. There are real issues here, and the path isn't certain. We've never been in this position, but logical minds have thought through this as best as they can, I promise. I like the discourse and the concern; it keeps us all honest and informed. But I give my unsought opinion that '70s-level inflation is not on the horizon.