The Mystery of Missing Inflation Weighs on Fed Rate Move | Hacker News Reader

The Mystery of Missing Inflation Weighs on Fed Rate Move

wsj.com

24 pointsprostoalex52 comments

As a complete layperson I might be out of my depth, but is that 5% unemployment realistic? If I remember correctly they don't count people who have been unemployed long-term. There's a lot of people who don't work; whether they collect disability/welfare, live off relatives, or make money through the grey/black markets I don't think the job market has really recovered nearly as much as a lot of people think. And a lot of those jobs are part-time or low wage. Perhaps that's anecdotal from the impression I get, but it doesn't seem like their model takes any of that into account.

Even without inflation it doesn't seem like the people I'm around (mostly middle to lower class) have all that much extra spending money laying around these days. If everyone was making more money I'm sure inflation would rebound, but as it is most the new money the Fed has been pumping into the economy seems to be going largely to the upper class.

No, 5.3% unemployment number is cooked. Even the BLS stats say 10.3% & that's probably gamed too.

http://www.cnbc.com/2015/09/04/-the-real-unemployment-rate.h... …

Understand the definitions of "unemployed" here (U1-U6): https://en.wikipedia.org/wiki/Unemployment#United_States_Bur...

Look up the time-series here: https://en.wikipedia.org/wiki/Unemployment#/media/File:US_Un...

It's a little silly to say that one of these measures is "cooked" while using as evidence an alternative metric that was measured and published by the same agency with the explicit understanding that the two metrics measured different things.

Not to mention the plummetting labor force participation rate, in the US at least (some other countries do better): http://charleshughsmith.blogspot.com/2015/12/why-has-labor-p...

Obviously, you are using U-6 instead of the headline rate, and using non-current numbers; U-6 is now 9.9% (with seasonal adjustment, 9.6% by the raw numbers.)

But while debating about which measure is "real" unemployment is fun, I guess, U-6 follows the same trend as the headline U-3 measure, so while the number is different, the relation to different points in the past is basically the same whichever measure you look at.

http://www.bls.gov/news.release/empsit.t15.htm

http://www.shadowstats.com/

I think he's using BLS figures, but emphasizing the U-6 figure, for instance, instead of U-3 figure. I don't have a subscription to compare numbers, but is that correct?

Here's a description of what they say they are doing:

http://www.democraticunderground.com/1002696672

It seems reasonable. Of course, there's a cynical interpretation: they are integrating a small but positively biased noise source (e.g. due to groups that seem like they ought to add up but don't) to obtain and monetize a politically convenient quadratic. Hard to tell.

EDIT: removed rant about the business model of this and similar publications

EDIT2: replaced my own analysis with a link to one which is deeper and more concise

The key thing that this shows is that U3 has come down but U6 hasn't. In other words, U3 and U6 have diverged, and only one has improved. That's a pretty good reason to doubt what U3 is telling us.

Egads, please don't spread unfounded conspiracy theories on HN.

Shadowstats is to economics as 911 truthers are to law enforcement[1].

Besides, independent attempts to measure inflation correspond closely to official numbers in the US. Witness MIT's Billion Prices Project, for instance.

[1] http://www.bloombergview.com/articles/2014-07-21/the-inflati...

[2] http://bpp.mit.edu/usa/

I don't know why you talk about independent attempts in plural. The one you cited is also the only example I've ever found.

The key here is not that BLS is faking numbers, but using a strange methodology. MIT's project used roughly the same methodology.

> If I remember correctly they don't count people who have been unemployed long-term.

Yes, they do. The thing the headline unemployment figure doesn't count are those "marginally attached to the labor force", who are no longer looking for work -- but those are counted in other measures that BLS collects besides the headline measure, including the subset of marginally attached known as "discouraged workers" bumps the rate from 5.0% to 5.4%, counting all marginally attached workers gets it up to 6.1%. (Nov. 15 data, with seasonal adjustment.)

http://www.bls.gov/news.release/empsit.t15.htm

I'm interested to hear whether you think it is wise to raise the short term interest rate with the economy the way it is now. You are clearly on top of the positive economics of the situation, but I don't recall you making many normative statements in the past. You strike me as well versed in economics and quite rational and unbiased and I would value your opinions/insights.

I think the problems we have now are not really responsive to monetary policy -- what we are really need is better fiscal policy; unfortunately, we have a Congress that is completely disinterested in addressing economic problems by fiscal stimulus.

Qualitatively, I'd be concerned about a rate increase having a negative impact on employment (the unemployment rate isn't bad now, but wages haven't done well, and the price of capital going up in those conditions would squeeze wages and employment further.) OTOH, one consequence of increasing income inequality and the stagnant lower end during the present expansion (and the past cycle, as well) could have made it more likely to be in a position where rate hikes to combat inflation are indicated and where rate hikes will lead to negative changes (rather than just slowing progress) on the employment front.

Quantitatively, I'm not familiar enough with relevant models to have a firm opinion one way or the other on a rate hike and the relative employment vs. inflation risks.

What fiscal policies would you recommend? Where should we cut back spending and where should that spending be channeled instead?

Interesting, thanks for the info. Still seems low, but it's more definitive than my anecdotal evidence.

The unemployment numbers are a joke. Look at employment ratios for a better sense of what things look like:

https://research.stlouisfed.org/fred2/series/EMRATIO

Much of that is due to demographics (retiring boomers, people staying in school longer) which is what the other figures try to take into account.

Details: https://www.fas.org/sgp/crs/misc/R44055.pdf

See page 11 (prime age employment-population ratio) if you want an EPR that accounts for this trend.

The Federal Reserve can directly control short-term rates, but long-term rates are determined by expectations of future short-term rates.

Since the global economy is still quite weak, raising rates now increases the probability of reduced economic growth, higher unemployment and future rate cuts.

Prediction 1: If short-term rates increase tomorrow, then investors will anticipate this and drive long-term rates lower. Much of this is already priced in, as long-term rates fell today with 30-year treasurys yielding less than 3%.

Prediction 2: Know-nothing analysts on cable news will be shocked to see long-term rates fall when short-term rates rise.

The Fed's model is broken.

Employment might be up, but wages (except for executives) have been stagnant for well over a decade. Retailers have been squeezing suppliers in order to maintain low prices (hello Costco, Walmart!). Discretionary spending is low due to under-employment in many sectors.

BTW: the "unemployment" figures fail to take into account people who are no longer looking for work and even worse, the masses of underemployed, people who can barely make ends meet.

It sounds like we've been fudging the numbers to paint a rosier picture of the economy in hopes that would encourage spending by consumers and, in turn, investors. But it seems the public has not been fooled, probably because that rosy picture doesn't match their bank statements.

My question is: how do we encourage wage growth when it doesn't appear to respond to the cost of living? Has free trade undermined the American worker's ability to negotiate a more favorable salary?

Are we living in 'take what you can get' economy?

You encourage wage growth by increasing employee productivity.

The more customer value a worker creates, the more an employer can pay them. Trying to increase wages without increasing the value created leads to unstable employment conditions and eventually unemployment if the situation is not profitable for the business on a risk-adjusted basis.

"It is not the employer who pays the wages. Employers only handle the money. It is the customer who pays the wages." - Henry Ford

Productivity is not directly linked to wages. The cash register didn't raise wages for cashiers, it increased productivity, but it also lowered skill requirements which increased the supply of potential cashiers and put a downward pressure on their wages.

For the person who is now able to be a cashier and previously unable to be a cashier, the cash register likely put upward pressure on that individual's wage earning ability.

Or and more likely, the cash register allowed more people who could not previously be cashiers due to poor math skills or whatever else, become cashiers. This in-turn increases the supply of cashiers, while making the job easier, ultimately leading to downward pressure on cashiers' wages.

Which brings us to where we are today, the lack of need for any cashier, please tell me you've used a You-Scan before. What do the wages look like for the cashiers who have been replaced by robot and self-service?

Of course I have, and in many cases, I prefer the You-Scan.

This is just an application of the general principal that if the most valuable skills you can bring to a job are trivially replaced by an inexpensive self-service kiosk, you have terrible wage prospects. I don't see any reasonable way around that.

Raise the minimum wage and you'll see a lot more robots and kiosks.

>Raise the minimum wage and you'll see a lot more robots and kiosks.

Good. Humans have no business doing what robots can do anyway.

The question remains: How do we adapt the humans who lost their jobs? And before you write them off as 'unskilled,' remember that your job could just as easily be next.

With a minimum wage high enough that there is nothing they can do that's worth that amount in the employment market, their options are to be unemployed or to work as an entrepreneur, independent contractor, or some other arrangement (possible quasi or not legal) that works around the minimum wage law.

In that way, they can use the skills and effort they have to a productive end, even if no one will bid the minimum wage rate for those services.

If the job I'm doing now is "next" (I don't think the job is "mine" in any possessive sense), I would have to step back and figure out what my next best option is. I'm quite sure that absent the utter collapse of civilization, that I'll figure something out and be fine.

Take a look at that graph: http://www.theatlantic.com/business/archive/2015/02/why-the-...

Productivity is not the problem, productivity is one metric the US economy has been doing great on.

The BLS publishes six unemployment indexes [1] that each include and exclude different groups for which employment status is debatable. However, each index has a largely consistent definition over time. They all follow essentially the same trend [2]. Could you be more specific as to the type of definition manipulation you believe is afoot?

[1] https://en.wikipedia.org/wiki/Unemployment#United_States_Bur...

[2] https://en.wikipedia.org/wiki/Unemployment#/media/File:US_Un...

The total labor force participation rate is total number of people who worked last month over total number of civilians over 16 years old. It is at historic lows but somehow "unemployment" is 5%

http://data.bls.gov/timeseries/LNS11300000

Not really surprising, Boomers keep retiring, and retirees not interested in work are part of the denominator for labor force participation rate, but not in any reasonable sense of the word "unemployed".

Ever-expanding lifespans and a demographic bulge hitting retirement age naturally combine to make labor force participation historically low.

Except that is the opposite of what is happening, Boomers and others who should be retiring are not and those who should be working are not. The only reason the LFP is not lower than it is is because more people can't afford to retire

The issue is and continues to be the LFPR which has been forecast to decline due to demographics, the problem is its been declining much faster than forecast. Which is because those within the peak working age population are not participating at the historical average, while those who are supposed to be retired continue to work to make ends meet. https://www.stlouisfed.org/publications/regional-economist/o...

You think economists at the Federal Reserve don't know all of those things?

Isn't that the point of the article ? They had a nice curve that didn't need to include those elements because everything else followed so they ended up with a nice equation with only money supply and interest rate in it that controlled the key to heaven.

Now they need a new model, probably one or more extra variable in the existing one. However we are at macro-economic level, that's not obvious how to put parent observation in a number.

But the Bank of Japan former director is more or less saying something like that: not all money is equal, money from saving does not trigger the same boost to the economy as money from wage. Or maybe it is just a moral stuff like they suggest too - inflation is a measure of optimism - pessimistic player money just drive inflation down.

It's naive to think that the Fed just uses some simple Taylor rule or Philips curve model in their policy decisions. Along with having a full staff of economists and research assistants doing econometrics, they have sophisticated microfoundational computational models like the NY Fed's DSGE model[1] (that they recently open-sourced in Julia[2]). There's a lot more to macroeconomic prediction than just "adding another variable."

[1] http://libertystreeteconomics.newyorkfed.org/2014/09/forecas... [2] https://github.com/FRBNY-DSGE/DSGE.jl

> BTW: the "unemployment" figures fail to take into account people who are no longer looking for work and even worse, the masses of underemployed, people who can barely make ends meet.

This is strictly true -- the one measure called "unemployment" does not capture these.

However, the official unemployment figure is one of six "alternative measures of labor underutilization" published by BLS [0], which while they have different numbers tend to follow the same overall trend, and those measures do track things like discouraged workers, underemployment (at least, in some forms), etc.

[0] see, e.g., http://www.bls.gov/news.release/empsit.t15.htm

> This is strictly true

the best kind of true.

Perhaps you're only referring to the headline unemployment figure that typically is reported on? The exclusion of people who are no longer looking for work is intentional: they are no longer part of the workforce, for a variety of reasons (there is no need to include someone who lives in the wilderness on passive income or savings as "unemployed"). But the BLS also provides a detailed breakdown of unemployment[1], including 6 alternate unemployment measures[2]. These alternate measures not only include "discouraged workers"[3][4] but you can also find numbers for people who are "marginally connected to the workforce" and people who have to work more than one full or part-time job[5].

[1] http://www.bls.gov/news.release/empsit.toc.htm

[2] http://www.bls.gov/webapps/legacy/cpsatab15.htm

[3] http://www.bls.gov/news.release/empsit.t15.htm

[4] https://en.wikipedia.org/wiki/Discouraged_worker

[5] http://www.bls.gov/news.release/empsit.t16.htm

edit: wow - a lot of people posted essentially the same thing.

Exactly. Not sure what the faux mystery is all about.

In the search for new models I've always been partial to the inclusion of credit (marked to market) as a factor in predicting inflation, instead of just money supply. This at least correctly predicted low-to-no inflation post-financial-crisis despite massive increases in the money supply.

http://www.economicsjunkie.com/inflation-deflation-revisited... has some good explanations of why.

There's also what you measure: consumer prices do not reflect increases in asset prices, and the latter have grown very quickly in the past few years (as credit has recovered).

Related, the Federal Reserve Bank of New York recently open sourced one of its models (coded in Julia)

https://github.com/FRBNY-DSGE/DSGE.jl/blob/master/doc/Data.m...

Discussion: https://news.ycombinator.com/item?id=10670138

Software is leading to much more leverage in capital efficiency. This is driving a reduction in CapEx spending in the US despite low interest rates. It's driving a reduced cost in exploration and extraction, which is in turn increasing supply for commodities and energy. We do more today with almost every conceivable good than we ever have before. More with less is the theme and it's getting much more powerful with software. IBM: $100bn, VMWare: $10bn, Docker/Containers: $1bn. All the same stuff. Our efficiency is increasing rapidly, and it's very difficult to predict with historical figures where this goes.

Companies seem to prefer buying back stock rather than spending on CapEx lately, but that's ending: http://imgur.com/SnbPXDC

Oh I get it, the Dockers lead to the multi-year global commodity glut, well that's a stunner..

Truthfully: it does amaze how the lens we use to view the world, shapes the picture we see.

Missing? There is definitely some inflation in college tuition, housing, healthcare costs and food.

3 out of the 4 you mention have all seen the Federal government silently take over, the US Mortgage Industry, and Student Loans are now majority run by the Feds and healthcare is on the same road, waiting for those savings any day now

I really don't think any of those were silent takeovers.

They basically have one valve they can turn. The model is too simple.

The model is nonsense - not even wrong.

But then the data has been massaged into irrelevance anyway, so it's pretty much a twofer.

How these people have the gall to pretend they're doing anything that resembles mathematical analysis is astonishing.

The problem for the Fed is that they have zero knobs. They have a balance sheet that is now much larger than before. Some people might say "reduce the balance sheet before raising rates." But that misses the point. The Fed just wants their knob back. The normal one. It's not about raising rates, but having a rate other than zero, even a symbolic low rate. As long as it isn't zero, the knob is there, ready to turn later.

Reply on news.ycombinator.com