One of the by-products of having so many economists (and central bankers) in ivory towers is that they operate under the implicit assumption that if it’s not measurable, it doesn’t exist.  The nice thing about living in middle America is that you get to see how “regular” people live.  One of the things that strike me from living in northern Texas is how many people have “jobs” that are far from mainstream – whether it’s engaging in some sort of part-time consulting (or part-timing in general), or multi-level marketing, or selling products and services online.  It got me to wondering about how all these people are accounted for in the labor force participation, unemployment rate and wage statistics.

You cannot tell me that people just drop out of the economy.  Yes – some people go back to school, or may be discouraged or scam the entitlement or disability system.  But what about everyone else?  A 3% drop in labor force participation is about 5 million people.  Those other categories do not add to 5 million.  They may disappear from the measurable economy, but there are various shades of grey in the harder-to-measure off-grid underground economy.  In particular, I am wondering how all of the people empowered by the opportunities afforded by the technology boom are accounted for in the official statistics.  Here are a few examples:

  • There is a 13 year old in my neighborhood that made $300K last year making YouTube videos. Apparently, spending an obscene amount of time on video games is not useless.  The top YouTube personalities can make millions (sometimes tens of millions a year).  Many internet personalities tend to be younger.
  • Etsy has not been doing well lately, but they still have 1.8 million active sellers (up from 1.6 million last year). Big Cartel has nearly a million.  I know people in each, and there are other marketplaces where people can sell their goods.
  • Let’s not forget all the storefronts on Amazon (2 million sellers), Ebay (25 million sellers), Alibaba, etc. Drop-shipping has never been easier – neither has running a business from your home, whether full-time or part-time.
  • Consider all the annoying ads and click-thru posts you see on Facebook, ZeroHedge, or any other popular web site. Someone is making those web sites and marketing them.  For every BuzzFeed there are probably hundreds of thousands (give or take an order of magnitude) of start-ups that do similar things.
  • It’s never been easier to write an app. I think it’s fairly standard at most reasonable high schools.  The process could start even earlier than in the Social Network – people get together to discuss various online projects.
  • The World Series of Poker attracted a record 107,833 entries this year. A lot of the younger professional players have a statistical background from large number of hours playing online poker.  It is technically not legal for vendors to offer games in the US(not illegal to play though, in most states), but there are plenty of work-arounds (i.e. Bitcoin).  Another related magnet for those statistically inclined is the world of fantasy sports betting.
  • Speaking of Bitcoin, I’m guessing a portion of the activity involves non-recordable income (see India). Does the surge in the price in part reflect a larger underground economy?

How exactly are all of these people being accounted for in the official statistics?  There are a few things to consider… many of these happen to be self-reporting “businesses” (at least until they make it big).  There are probably various levels of underreporting.  The fact of the matter is, the more you make, the more payroll and income taxes you have to pay, and you may be required to give benefits (like healthcare).  I’m not saying people are dishonest, but there is a massive incentive to under-report (or “creatively” report), or not report at all.  The St Louis Fed estimated in January 2015[1] that the underground economy could be on the order of 13% of GDP in developed countries.  That would be over $2.3 trillion a year for the US, but the advent of improved technologies since that report could have increased that percentage over the last few years.

I keep hearing how employers can’t find workers with the right job skill set.  Part of it may be the desire to hire cheaper H1 visa employees.  But maybe it’s because a lot of “skilled” workers find more lucrative opportunities off-the-grid.  If you were great at developing technology and couldn’t get a “dream” job at Google or Facebook, are you more likely to get a 9-to-5 at Macy’s technology department, or do something on your own (potentially in your mom’s basement)?  See the WhatsApp guys.

And for the love, can the BLS and every economist out there stop including the under-19 year old crowd in the official labor force participation statistics?  This isn’t the 1900s, where 16 year olds went to work in the coal mines.  High school graduation rates have been on the rise for decades, from 75% 30 years ago to almost 90% now.  The under 19 group also happens to be the group that has dropped the most, as you can see from the chart on the right.  Most parents are running around ragged, trying to take their kids to swim practice, band practice, and other extracurriculars.  This could also explain the appeal of more flexible, potentially off-grid jobs.

I would think the age group most likely to engage in off-grid jobs are those in the 20-24 crowd, next followed by the primary 25 to 54 group.  It’s because when you are younger, you have more time to experiment, but as you get older you want more stable income.  I’m pretty sure someone 55+ is least likely to be setting up a web site or engaging in other off-grid jobs.  So it seems reasonable to me that the former groups have had lower labor force participation in the past decade.

I would argue that we don’t have a labor force participation problem.  What we have is a problem in how to account for a potentially increasing part of the economy that cannot be measured.  When the relative appeal of an off-grid job is higher, you get people leaving the measurable labor force.  When the relative appeal of an on-grid job is higher, you get higher official labor force participation.  The appeal of off-grid jobs are more appealing to younger people, than older workers.  While the some of the labor force participation drop of high school grads could be explained by the increasing enrollment in college and graduate school, there are other forces at work.

It’s still not clear to me what the implication of the off-grid economy is.  I suspect that the lower labor force participation is a sign of its growing importance.  And I’m not so sure we can just assume an unemployment rate of X% based on some historical norm is an appropriate level of “neutral” – mostly because the composition and size of the underground economy is different now than in the past.  There could be a selection bias issue in the transfer between on-grid and off-grid jobs.  Those who are successful off-grid are probably the ones with better skills or work habits, etc.  Those who are not successful off-grid and go back to the labor force are probably the lower-skilled (or less motivated, etc) that may not translate into more productive or higher-earning workers.

I think the above corroborates what we have been seeing in the economy… lower labor participation than pre-crisis, lower wages and productivity on the margin (although the marginal effect from the underground economy is only a tiny factor), flatlining tax receipts and a resilient consumer.  I realize there isn’t a lot of concrete data to support most of my contentions.  But that’s exactly the point.  There is no concrete data.  But just because the data isn’t there doesn’t mean there aren’t other factors that exist.  So when we look at the subset of measurable data, we need to keep this in mind.

The Employment Report will always be the most important data set we get, since the majority of workers are captured in the data.  However, as we get closer to “full employment”, there may be more transfers between the underground and measurable economies and we may get more noise (like last month’s 429K drop in labor force participation and -233K drop in “employment”).  The measurable economy is much larger, but I think the off-grid economy could be important in assessing the marginal changes in the various labor data.

As a result, looking at broader measures of consumer health could start becoming more important.  This is why the recent weakness in Retail Sales figures is a little concerning – a total of 0% over the past four months in the headline is strangely weak.  For now, I’m just going to call it an auto and gas story.  But we should watch for consumer data more carefully.

This is a work-in-progress, but I just wanted to give you a non-traditional take to mull over.