Housing Heats Up

June 5, 2016

In parts of the country, particularly in the west, demand for housing is strong, causing higher housing prices and lower rental vacancy rates.  For the first quarter of 2016, the Census Bureau reports that vacancy rates in the western U.S. are 20% below the national average of 7.1%.  At $1100 per month, the median asking rent in the west is about 25% above the national average of $870 (spreadsheet link).

With a younger and more mobile population, home ownership rates in the west are below the national average (Census Bureau graph). Housing prices in San Francisco have surprassed their 2006 peaks while those in L.A.are near their peak.  Heavy population migration to Denver has spurred 10% annual home price gains and an apartment vacancy rate of 6% (metro area stats).

From 1982 through 2008, the Census Bureau estimates that the number of homeowners under age 35 was about 10 million. These were the “baby bust” Generation X’ers who numbered only 70% of the so-called Boomer generation that preceded them.

Shortly before the financial crisis in 2008, a new generation came of age, the Millenials, born between 1982 and 2000, and now the largest age group alive in the U.S. (Census Bureau). Based on demographics, homeownership should have increased to about 13 million in this younger age group, but the financial crisis was particularly hard on them.  Starting in 2008, homeownership in this younger demographic began to decline, reaching a historic low of 8.8 million in 2015, a 15% decline over seven years, and a gap of almost 33% from expected homeownership based on demographics.

In response to lower homeownership rates, builders cut back and built fewer homes.  I’ll repost a graph I put up last week showing the number of new homes sold each year for the past few decades.

Look at the period of overbuilding during the 2000s, what economists would euphemistically call an overinvestment in residential construction.  Then, financial crisis, Great Recession and kerplooey!, another technical term for the precipitous decline in new homes built and sold. As the economy has improved for the past two years, the demand for housing by the millennial generation, supressed for several years by the recession, has shifted upwards.  More demand, less supply = higher prices.  This younger generation prefers living closer to city amenities, culture and transportation, causing a revitalization of older neighborhoods.  In Denver, developers are buying older homes, scrapping them off, and building two housing units where there was one. Gentrification influences the rental market as well as affordable single family homes and pushes out families of more modest means in some parts of town.

The housing market really overheats when rentals and home prices escalate at the same time. During the housing boom of the 2000s, many tenants left their apartments to buy homes and cash in on the housing bonanza.  Rising vacancies put downward pressure on monthly rents.  Move-in specials abounded, announcing “No Deposit!”, “First Month Free!” or “Free cable!” to attract renters. This time it’s different.

Rising rents and home prices put extraordinary pressure on working families who find they can barely afford to live in central city neighborhoods which offered low rents and affordable transportation.  They consider moving to a satellite city with lower costs but face longer commute times and additonal transportation costs to get to work.  Demographic trends shift more slowly than building trends but neither moves quickly so we can expect that housing pressures will not abate soon until the supply of multi-family rental units and single family homes increases to meet demand.



For the past four decades, household income has declined, as Presidential contender Bernie Sanders is quick to point out.  Some economists also note that household size has declined greatly during that time as well so that comparisons should take into account the smaller household size.  A recent analysis  by Pew Research has made that adjustment and found that middle class incomes had shrunk from 62% of total income in 1970 to 43% in 2015.

But, again, comparisons are made more difficult because some categories of income, which have risen sharply in the past few decades, are not included.  Among the many items not included are “the value of income ‘in kind’ from food stamps, public housing subsidies, medical care, employer contributions for individuals (ACS data sheet).  Generally, any form of non-cash or lump sum income like inheritances or insurance payments are excluded.  There is little dispute with the exclusion of lump sum income but the exclusion of non-cash benefits is suspect.  An employer who spends $1000 a month on an employee health benefit is paying for labor services, whether it is cash to the employee or not.

The lack of valid comparison provokes debate among economists, confusion and contenton among voters.  The political class and the media that live off them thrive on confusion. Those who want the data to show a decline in middle class income cling to the current methodology regardless of its shortcomings.



The BLS reported job gains of only 38,000 in May, far below the gain of 173,000 private jobs reported by the payroll processor ADP and below all – yes, all – the estimates of 82 labor economists. The weak report caused traders to reverse bets on a small rate increase from the Fed later this month.

Almost 40,000 Verizon employees have been on strike since mid-April and just returned to work this past week. On the presumption that a company will hire temporary workers to replace striking workers, the BLS does not adjust their employment numbers for striking workers.  However, most employers of striking employees hire only as many employees as they need to, relying on salaried employees to fill in.  Do strikes contribute to the spikes in the BLS numbers?  A difficult answer to tease out of the data. In the graph below we notice the erratic data set of the BLS private job gains (blue line; spikes circled in red) compared to the ADP numbers (red line; spike circled in blue).

Each month I average the BLS and ADP estimates of job gains to get a less erratic data swing.  The 112,000 average for May follows an average of 140,000 job gains in April – two months of gains below the 150,000 new jobs needed to keep up with population growth.  Let’s put this one in the wait and see column.  If June is weak, then I will start to worry.

Labor and Purchasing Managers Index

September 7, 2014

Labor Report

The Bureau of Labor Statistics (BLS) reported net job gains of 142K in August, much lower than the 200K+ expected.  The private payroll processor ADP reported 204K net private job gains earlier this week.  Some economists predicted that the number will be revised upwards in the next month.  Some point to the difficulties of the seasonal adjustment factor in August.  Below is the monthly net change in jobs with and without seasonal adjustments.

As usual, I average the private net job gains reported by BLS and the payroll processor ADP to come up with net job gains of 169K, add in the 8K job gains in the government sector to get a total of 177K. Another approach to take out the variability is to use the year-over-year change or percent change in employment.  As you can see in the chart below, the monthly seasonal adjustment (in red, overlayed on the blue non-seasonally adjusted figures) attempt to replicate this year over year change on a monthly basis.

As the year-over-year job gains topped the 2 million mark at the start of 2012, the “Golden Cross” – when the 50 day average of the SP500 crosses above the 200 day average – occurred shortly thereafter.  Zooming in on the past year, we can see that the difference between the two series is relatively slight.  In fact, the economy is nearing the levels of late 2005 to 2006 when the labor market was a bit overheated in some regions of the U.S.  The difference between now and then is that workers have relatively weak pricing power.  The average wage has increased just 2.1% in the past year.

A comparison of the monthly growth in jobs, as reported by the BLS, to the Employment index of the ISM Non-Manufacturing Survey shows that the ISM number charts a less erratic path through the variability of the employment data.  The index has been positive and rising since the hard winter dip.

The unemployment rate ticked down slightly in August, but the more significant trend is the decreasing number of involuntary part timers, those who are working part time because they can’t find full time work.

The widest measure of unemployment, which includes both these part time workers and those who have become discouraged and stopped looking for work, finally touched the 12% mark this month.

In short, this month’s employment report was good enough but not so good that it would shorten the period before the Federal Reserve begins to hike interest rates.


Constant Weighted Purchasing Index (CWPI)

Each month for the past year, I have been doing a little spreadsheet magic on the Purchasing Managers Index published by ISM to weight the employment and new orders components of this index more heavily.  This has proven to be a reliable and less erratic guide to the economic health of the country.

The manufacturing component of the ISM Purchasing Managers Index was particularly strong in August.  Because the CWPI weights new orders and employment heavily in its composition, the manufacturing component of the CWPI is at levels rarely seen in the past 34 years.  Levels greater than this have occurred only twice before – in November and December 1983 and December 2003.  Both of these previous periods marked the end of a multi-year malaise.

The services sector, which comprises most of the economic activity in the country, is strong and rising as well. New orders declined slightly but are still robust and employment is growing.  The composite of these two components is near robust levels.

This month the CWPI composite of manufacturing and service industries topped the previous high of 66.7 set in December 2003 and is now at an all time high in the 17 years that ISM has been publishing the non-manufacturing index. If the pattern of the past few years continues, this overall composite will probably decline in the next month or two.

Strong economic activity was muted somewhat by a lower than expected monthly labor report.

Labor’s Journey

January 12th, 2014

A dramatic decrease in new orders, mostly for export, for the non-manufacturing sector of the economy offset other positives in the December ISM report.  The composite non-manufacturing index dropped slightly but is still growing.  A blend of the manufacturing and non-manufacturing indexes, what I call the CWI, declined from its peak as expected. A month ago I noted the cyclic pattern in this index, and the shorter time between peaks as the economy has formed a stronger base of growth. Most businesses are reporting expansion, or strong growth.  Some respondents to the survey noted that the severe winter weather in December had an impact on their business.


Ringing in the New Year, the private payroll firm ADP issued a strong report of employment growth before the release of the BLS figures on Friday.  The reported gain in jobs was above the best of expectations.  In the past few months,  several reports in production and now in employment have exceeded expectations or come in at the upper bounds of estimates.


Wells Fargo announced that they will be offering non-conforming mortgages to selected buyers who present a low risk.  Non-conforming mortgages may be interest only, or have loan to values that don’t meet guidelines. Reminiscent of the “old days,” Wells Fargo intends to hold onto the mortgages instead of selling the paper in the secondary market.

The Gallup organization announced their monthy percentage of adults who are working full time, what Gallup calls the P2P.  I call this the “Carry the Load” folks, those people whose taxes are supporting the rest of the population.  At 42.9%, it is down a percentage point or two from previous winters.

The 4 week average of new unemployment claims is still below 350,000 but 20,000 higher than a month ago.  As I mentioned last week, this metric will be watched closely by traders in the coming weeks.  Although there is little statistical significance between a 349,000 average and a 355,000 average, for example, there is a psychological boundary marked in 50,000 increments.

Friday I woke up and found that somebody stole the ‘1’s at the Bureau of Labor Statistics.  The BLS reported net job gains were 74,000 and I thought that there was a smudge on my computer screen blocking the ‘1’ of 174,000 and reached out to wipe it off.  There was no smudge.  It is difficult to interpret the discrepancy between the ADP report and the BLS report.  Some say that the particularly harsh winter weather in the midwest and east caused many people to stop looking for work or that many businesses returned their BLS survey late.  If so, we may see some healthy upward revisions to the employment data when the February report comes out. Here’s a look at total private employment as reported by BLS and ADP.

As you can see there is a growing divergence between the two series.  As a percentage of 120 million or so employed in private industry, the divergence of a few hundred thousand is slight.  The BLS assumes a statistical error estimate of 100,000.  But people closely watch the monthly change in employment as a forecast of developing trends in the overall economy, changes in corporate profits and consequently the price of stocks.  Here is a chart of the difference in private employment as measured by the BLS and that measured by ADP.  A positive number means that the BLS is reporting more employment than ADP.

As with any estimates, I tend to average the estimates to get what I feel is a more accurate estimate.  This averaging works well when bidding construction jobs and some statistical experiments have proven the method reliable.  Averaging the two estimates for private payrolls gives us an estimate of job growth that is still above the replacement threshold of about 150,000 net job gains per month needed to keep up with population growth.

The figures above do not include 22 million government employees, or about 14% of total employment.  Flat or declining employment in this sector has dragged down the headline job gains each month.  Adding in net job gains or losses in the government sector gives us a net job gain of about 150,000 in December.

For those of you interested in more analysis of the employment report, Robert Oak at the Economic Populist presents a number of employment charts similar to the ones I have been doing in past months.

For the past 5 – 10 years, much has been written about the growth in income inequality during the past 30 to 40 years. I’ll call income inequality “Aye-Aye” because the abbreviation  “II” looks like the Roman numeral for “2” and because Ricky Ricardo used to exclaim “Aye, Aye, Lucy!” on that much loved comedy series.  Those on the left blame former President Reagan,  British Prime Minister Thatcher, and deregulation for Aye-Aye.  Those on the right blame increasing regulation that disincentivises businesses from taking chances, from making capital and people investments to pursue robust growth. The expansion of social welfare programs makes people ever more dependent on government and less likely to take jobs that they don’t want.  Economists cite the aging of the population as a cause of the growth of Aye-Aye.  Few I know of seriously challenge the idea that Aye-Aye has been happening.  The argument is over the causes and the solutions.

Thomas Piketty’s Capital in the 21st Century will add to the debate.  The English translation will be published in March.  A book review in the Economist outlines some of the ideas in the book.  Piketty’s analysis of almost 150 years of data from several countries indicates that the slower an economy grows, the more unequal the distribution of income.  One might think that the U.S. would have the most unequal income distribution, but Piketty reveals that it is France that tops the list.

Piketty’s rule of thumb is that the savings rate divided by a country’s growth rate will approximate the ratio of capital wealth to gross income.  As this ratio increases, more of the national income goes to those with capital wealth. So, if the savings rate is 8% and the growth rate is 2%, then capital wealth will be about four times gross national income.  Furthermore, he finds that population growth accounts for about half of economic growth over the past century and half.  Slowing population growth in the developed nations therefore leads to greater inequality of income.  If this rule of thumb is fairly accurate, stronger economic growth is the only way to lessen the inequality of income that has grown steadily over the past thirty to forty years.

If you are familiar enough with French, you can read a preview here or pre-order the English version here.  The book is sure to spark some lively discussion between those in the economic growth camp and those in the demographic camp.  The topic has long been a topic of discussion in emerging economies.  I will quote from an Asian Pacific policy journal published in 2003, “The most important determinant of inequality is not [emphasis mine] economic growth, however, but rather changes in demographic age structure.”

BLS Estimates

Every month the Bureau of Labor Statistics (BLS) releases their initial unemployment estimates and these figures headline many newspapers and news broadcasts. The unemployment percentage moves markets and provides endless opportunity for comment and analysis.

TrimTabs, an investment research company, argues in a concise four page report that the employment numbers from the BLS are inaccurate. There have been many critics of the BLS methods in the past but they have generally focused on the “birth/death” adjustment that the BLS uses to account for the creation of jobs by new businesses. In the past several years, these BLS adjustments have proven to be fairly accurate. But the BLS does not appear on CNBC to refute their critics.

This TrimTabs report focuses on several other flaws in BLS methodology. The methods that the BLS uses were developed many years ago when the majority of employment in the U.S. was in manufacturing. Today, manufacturing represents a small portion of the U.S. economy. Yet the BLS continues to survey mostly manufacturing companies and government institutions to come up with their unemployment number. Of the private companies surveyed by the BLS, most of them are large despite the fact that smaller companies, those with 500 employees or less, make up 50% of the economy and most of the economic growth in the U.S.

The BLS reports a statistical 90% confidence in their estimate numbers, resulting in an error of + or – 100,000 jobs lost to their monthly estimate, a relatively small error out of a workforce of 140 million. But market trading is based on pre-estimates of monthly jobs lost by “analysts” as well as competing estimates like that of the payroll processing company ADP. If the BLS figure of jobs lost is 300,000 and concensus pre-estimates were 400,000 jobs lost, the stock market often rebounds. Yet, statistically, the BLS estimate could be the same as pre-estimates. The market conveniently forgets the sampling error of both analyst estimates and the BLS estimates and trades on the estimates as though they were hard data.

The BLS does not report on the hard data, actual state unemployment insurance claims, till it has received and compiled all the state reports. This is done about 12 months later.
In a 12 month period during 2005-06, TrimTabs research shows that the BLS had underestimated job growth by 750 million when the estimates were compared with actual data. While the BLS does a remarkable job gathering data from almost 400,000 companies in a month to arrive at their estimates, the problem of accurately assessing the employment activity of this country is enormous.