“Never as Dominant” a Partisan Effect

In his comments concerning the disparity between the sentiment of Democrats and Republicans in early 2017, the University of Michigan’s Richard Curtin wrote that “The surge in consumer confidence was based on political promises, and not, as yet, on economic outcomes.” He also pointed out that in 50 years, the surveys “never recorded as dominant” an effect of partisanship on economic expectations. Although at that time Democrats had become much more pessimistic over the last six months, and Republicans much more optimistic, Dr. Curtin himself expected the divergence to converge. Why? He believes consumers themselves find their own expectations useful guides, and in order to regain that tool, both parties would have to “temper their extreme views.”

That has not happened. While Independents tend to track all households, the outlook among Democrats and Republicans has instead widened slightly since Trump took office, as the outlook has risen 4 points among Democrats, and close to 7 among Republicans. Small beer, perhaps, but against Dr. Curtin’s expectations.

UMich collects data by political party only sporadically, so we don’t have a full history, but look at the spike in the gap between the assessments of Republicans and Democrats. It went straight up, by close to 100 points, in February of 2017:

In 2008–2009, all agreed that the outlook was grim, with sentiment among Democrats and Republicans hitting lows, and series lows, of 31 in June 2008 among Democrats, and 48 in March of 2009 among Republicans. Both parties also had low levels in June 1980, and both parties ended 2008 at 59.

The raw average for the entire intermittent series is 80 for the Democrats, and 89 for the Republicans. The raw range for Democrats is the 55 points between 51 and 106, and for Republicans the 77 points between 48 and 124 (rounding). Taking out the lows of the Great Recession narrows that to 32 points for the Democrats, but just to 70 for the Republicans. o Republicans are clearly more excitable and optimistic when their party is in office than the more dour Democrats

And it’s all on the exuberant side of the scale. Both parties have a floor around fifty, but the Republicans top out at 124, almost 20 points higher than the Democrat’s 106.

In the surveys taken during the Reagan and GW Bush years, Republican sentiment averaged 20 points above that of Democrats, if you exclude the 3.9 and 0 for November and December 2008 after Obama’s election, that is. During the Reagan years Republican sentiment averaged 108, Democratic, 88, and during the HW Bush years Republican sentiment ran at 80, and Democratic at 64.

In those polls taken while Obama was President the Republicans averaged 19 points below the Democrats, but averaged just 9 points below during the recession years of greater agreement. (So as the expansion took hold, the split widened.) Whoever thought we’d remember anything comforting about those two years?

During the years Obama was president, Democratic sentiment averaged 87, and 93 if you take out the recession years. During those same years Republican sentiment averaged 69, and 72 if you take out the recession years.

During the years Trump has been President, Republican sentiment has averaged 117.6; while that of Democrats averaged 80, meaning Republican sentiment is running 46 points above the Obama years, while sentiment among Democrats is running just 13 points below (both calculations exclude the recession years). Democratic sentiment is running near where it has in prior Republican administrations, while Republican sentiment is off the chart.

It’s no surprise that consumers feel more confident when the White House is occupied by people whose policies they tend to support. But currently, the Republican outlook, which peaks above the Democratic outlook, is in the cat bird’s seat, while the Democrats are more restrained on the downside. (Dr. Curtin noted that all of April’s gains were driven by Democrats and Independents.) It’s not illogical to wonder if the survey is becoming biased to the upside during this Republican administration, given these differences in character.

The gap is even greater for the expectations component of the index. Dr. Curtin points out in September 2017 comments that over the then-last seven months, Republican expectations exceeded the series peaks for all households, while Democratic sentiment was about level with the start of the recessions of the last half-century. In October 2016, the Republican outlook was 61 while the Democratic outlook was 95. As you can see where blue fades to light red on the right of the lower chart, by February 2017 the Republican outlook had doubled, to 120, while the Democratic outlook fell to 56, rocketing the gap between the two by close to 100 points, from -34 to +65. That unprecedented disparity narrowed only to 48 in September, and has since widened to 55, considerably larger than the gap for current conditions. During the final months of the Obama administration, a gap also widened, but it averaged 22 points for the 5 months, and only extended to 34 points.

A striking point Curtin makes: the demographics really changed in 2017. Through 2016, generally, assessments of economic possibilities rose with income and education, and fell with age. That changed in 2017. Overall expectations for those with less than a high school degree had risen from 68 in October 2016 to 82 by August 2017, but fell for those with a college degree from 87 to 80. For those 65+ they rose from 2016’s 69 to 83 in 2017, while slipping a bit, to 86, for those 18 to 34 years old.

Sentiment for all three education terciles tracked each other closely throughout the series, all peaking around 2000, high school at 100 and the two college groups at 120, before falling raggedly to about 60 in the recession. The current divergence is led by an increase in the outlooks of the two lower attainment levels while those with a college degree are basically flat.

Since 1978, the gap between those with a college degree and those with some college has averaged 4.8. It has widened over time, and varies a lot. It’s averaged -0.4 since January 2017, and is currently -1.4. The gap between those with a college degree and those with high school or less has averaged 12, and had never been below the water line through 2016. It did slip to 1% in February 2009, maybe there’s something about the aftermath of an inauguration, but recovered to the average and even spiked into the 20s in early 2013, a record high, although it hit 20 in early 1985. In October 2016 the gap was 14 points, slipped to 6 in January, and then ran slightly negative for 3 months. It has now recovered to 4.2, and has averaged 3 since January 2017.

This is a tricky issue. We want people to get the education they need to have job and economic security, and we doubt that less than high school fills the ticket, but people with low educational attainment shouldn’t be punished, and we want them to do well. Still, such a shift is unsettling.

It really is a drag that we don’t have a full series. Maybe one of the subgroups, perhaps those with some college in the north central region? always gets it right. We’ll never know.

by admin· · 0 comments · Current Affairs, Uncategorized

Protected: Flow of Funds: Corporate Imprudence in a Languid Field

This content is password protected. To view it please enter your password below:

by admin· · Enter your password to view comments. · Uncategorized

Just around the Corner? Maybe Not.

A very smart analyst joked yesterday that the Consumer Price Index is the new Nonfarm Payroll. Since the FOMC has made it clear that they are waiting for 2% inflationor perhaps Godotall eyes are searching for signs we might be getting there, so all price indexes are good fodder, especially the PCE, the Fed’s favored index.

The FOMC’s models have continually demonstrated a “just around the corner,” feature where, despite currently weak PCE measures, the expectation is that higher inflation would occur in the near future.
Unfortunately, a historical look at the broad components of this inflation measure belies their optimistic projections:


Above is a logarithmic graph of the three components of the PCE price index: durable goods prices (in blue), non-durable goods prices (in red) and service prices (in green). Two key visual elements are apparent: durable goods prices have been consistently declining since the mid-1990s. Non-durable goods prices have been stagnant to slightly lower for the last 5 years. That leaves services as the only PCE price index component that can exert upward pressure.

Let’s look at the same data from a Y/Y percentage change perspective:


Durable goods prices (in blue) have been subtracting from price growth for the last 20 years. Non-durable goods prices (in read) have been declining since 2011-2012; they subtracted from PCE price growth for most of 2015 and only recently turned positive. Only service prices (in green) have increased PCE price pressures on a consistent basis.

There are several important lessons to draw from this data. First, the PCE price index looks at prices from a business perspective. According to the Cleveland Fed, “the PCE is based on surveys of what businesses are selling.” The above charts indicate that neither durable goods nor non-durable goods companies have any pricing power. Second, the Y/Y percentage change in service prices has been declining. Third, price growth for 31% of PCEs are either negative or very weak. That means the remaining 70% of prices would have to increase at a faster Y/Y rate to hit the Fed’s 2% PCE Y/Y inflation target. This runs counter to the second conclusion regarding service prices, that the y/y rate of change is narrowing.

Those trends mean the Fed cannot hit its 2% y/y target under the current circumstances.

by admin· · 1 comment · Red Flags, Uncategorized

Around the Web in a Couple of Clicks


September 11

Hurricanes and economic data

Weak inflation may force the Fed to pause its rate setting policy

After appearing weakened a year ago, Merkel is now the German front-runner.
Subscription required

The Atlanta Fed’s GDP Nowcast is trending lower

NY Fed’s GDP nowcast at 2.06%

Gavyn Davies: Global Nowcasts are improving
Subscription required

Gavyn Davies: The Global Upswing Could Be Stronger This Time
Subscription required

Cleveland Fed’s Yield Curve Recession Predictor: 12.9%

See also: A recent look at term spreads

Scott Grannis: A better PE ratio

What to watch for in the Census’ earnings, income, and poverty release

Are weaker oil prices responsible for stronger global growth?

See also, James Hamilton’s research at the NBER

September 8–Focus: Income Inequality

Those of us who don’t want to think about this as an ethical issue might want to give it some thought as a pragmatic economic bummer. The details matter a great deal for an economy based on consumption, as is ours here in the U.S., especially since those who spend the largest percentage of their incomes are the very ones getting shorted.

The Facts:
Striking it Richer: The Evolution of Top Incomes in the United States

Mobility Report Cards: The Role of Colleges in Intergenerational Mobility

Economic Inequality: It’s Worse than You Think

A deep dive into the history, with those Gini Coefficients you wanted for your birthday:
Income Inequality

Distributional National Accounts: Methods and Estimates for the United States

And finally, apologies for the “self link,” but here’s our summary of an un-linkable paper outlining public opinion on the matter. The facts were very different from what we often hear.

by admin· · 0 comments · Uncategorized

Spidery Views on Inflation and Employment

We are happy that our interpretation of Fed Governor Lael Brainard’s speech on Tuesday, May 30th, to the New York Association of Business Economists seems to have had an effect on how analysts are viewing the current economy. As we wrote here, her talk was not the shedding of her dovish feathers, but a continuation of her focus on ongoing weakness in inflation despite, as she put it, the growing number of forecasters who believe the U.S. economy is at or very close to full employment. We decided to let the spiders do the talking on the state of US employment, and on inflationary pressure, or lack thereof, the now ubiquitous tag.

First employment. Here’s an updated employment spider inspired by one put together by James Bullard, the president of the St. Louis Fed, when he was feeling “Halloweenish” back in 2015.


The graph uses an integer measure showing the percentage of recession losses that have been regained in recovery and expansion. (For example, an 0.5 reading means that 50% of the losses have been regained; 1.5, 150% have been regained. In most of what follows, the numbers are well above 1, meaning that they’ve recovered and then some.) In a phrase, we’re well beyond the recovery phase and firmly into expansion.

Bullard sorted ten major employment indicators into four groups: employer activity (payroll employment, job openings and hires from JOLTS), confidence (“jobs plentiful” from the Conference Board’s confidence survey, quits from JOLTS), slack (narrow and broad unemployment, share working part-time for economic reasons), and leading (initial claims and temp employment). The measures of slack are the least recovered, with the share working part-time for economic reasons the only one of the 10 still in a hole. Hires are lagging both openings and employment. The leading indicators are looking good, though with the growing casualization of employment, temporary firms may have lost some of their leading characteristics. And the confidence measures have also picked up nicely.

Which takes us to the next spider (or radar, if you prefer) graph. With the job market relatively strong, how would a such a graph look for inflation? Much, much weaker.


In this graph, we’re not looking at losses recovered but how the inflation-related indicators look compared to their levels at the onset of recession in December 2007. Following Bullard’s model, we sorted twelve inflation-related indicators into five groups: CPI measures, GDP measures, earnings, prices, and expectations. Headline year-to-year inflation is 60% below its pre-recession level; core, 30% below. Even medical inflation is down by half. That’s true even though real GDP growth in 2017Q2 was a little above where it was in 2007Q4. Underscoring the point about the lack of inflation, nominal GDP growth is well below end-2007 rates, unlike its real counterpart. That speaks to calls by Fed advisor Michael Woodford to track nominal GDP as a way of getting around problems with inflation adjustments Average hourly earnings growth is more than a sixth lower than it was ten years ago, and the employment cost index, over a third lower. Price indexes in the ISM surveys, especially services, are also below where they were at the end of 2007—and the oil price is down by half. And inflation expectations—measured by the Michigan survey of consumers, and what is implied by the spread between 10-year Treasuries and comparable TIPS—are about a quarter below where they were then.

With monetary policy gently lifting and fiscal policy in great disarray for now, deflation looks like more of a risk than inflation.

Spotlight on shelter inflation

It’s been nearly a year since we looked at the outsized contribution of shelter to the overall inflation. It’s still outsized.

For the year ending in July 2017, both headline and core inflation were up 1.7%. That’s about where the headline number was a year ago; core was half a point higher. But inflation in shelter was 3.2%, down only 0.1 from a year ago. Take shelter out of core inflation, and prices are up 0.6% for the year, less than half the rate twelve months earlier.

Now we realize that taking food, shelter, and energy out of the CPI takes slightly more than half out of the market basket. Our dear John Liscio used to joke that some analysts wanted to take so many things out of the price indexes that poor Hampton Pearson of CNBC would have to report the numbers while standing in front of the BLS in the dark, wearing only a barrel. But inflation that’s largely concentrated in shelter and (from time to time) energy doesn’t have much to do with the shape of the labor market.


Tire kicking

Quite the buzz arose on the release of the June Job Opening and Labor Turnover Survey (JOLTS) numbers, showing record-tying openings in the private sector. (There wasn’t much else of note in the report; the hires rate was flat, and the layoff rate was up 0.1pp from May and the prior year.) The inability to find qualified employees is the standard explanation; a creative variation that’s arisen recently is that too many fail their drug tests, or would if they had the nerve to take them.

We’re flipping that argument. How can the high level of openings coexist with a less exuberant quit rate (which, by the way, slipped a notch in June)? In 2006, openings ran 0.9 points above quits, the spread tightened into 2009, and then widened again. It’s averaged 1.7 since 2015, and in June was 1.9. Why has the gap widened so?


Presumably the already-employed are qualified and capable of passing a drug test. It seems odd that employers with all those openings don’t bid them away. Could it be that employers just don’t want to pay up, or can’t with productivity so weak?

Perhaps, also, the already employed are aware that their own employers are putting out feelers but reluctant to follow up with actual hires.

We’re not even talking about drawing the unemployed or marginally attached in from the sidelines. If there is real information in the openings rate, we are confident there would be more job-switching.

July sales tax collections

In July, 54% of the states in our survey met or exceeded their forecasted sales tax collections, down from June’s 70%, and 97% reported growth over the year. That’s better than June’s 82%, but was padded by the fact that July 2016 ended on a weekend, forcing payments into August last year. The average over-the-year rate of change moved to 4.7% from June’s 52%, and the margin from forecast rounded down to 1% having rounded up to 1% in June.


July collections tend to be a “throw away” month in revenue departments. Some of the collections accrue back to June, and July is what’s left, so trend watching resumes in earnest in August. The positive comments were subdued. Our contact in a large western state noted that they are continuing their “trend of beating forecast,” but that’s a conservative 1.2% for the current quarter.

The energy extraction states had another strong month, but recent state detail on job gains in those states does not suggest that extraction employment is leading the way. No surprise there, it’s a tiny sector. For example, in Texas, logging and mining work constitutes just under 2% of total employment, so its 7% contribution to the over-the-year gain of 313,000 jobs is impressive. But that contribution is eclipsed by professional & business services, and government work’s contributions of 20%, or 60,000 jobs each. Each of those sectors comprise about 14% of total employment.

In some states collections soared toward the end of the month and faded in the early days of August, reflecting the calendar effect. Our contact in one such state noted that his conversations with industry leaders does not suggest a consistent “more jobs=more spending” scenario, but rather that perhaps the better job gains among college students this summer is contributing to decent withholding and quite strong sales revenues. He’s also concerned that debt may be creeping up.

State-level GDP: Down on the farm with energy rising

During first quarter of 2017, economic growth declined in five of the seven Plains states. Within this region, only Missouri and North Dakota saw their economies grow during the quarter. North Dakota’s growth came from a reviving energy sector that contributed 1.7pp and offset almost the entire 1.9pp drop in its agriculture sector. Missouri’s agriculture sector declined by 0.8pp, but this was offset by relatively strong growth in durable goods manufacturing and wholesale trade.

The states that experienced the greatest declines in output were Nebraska (-4.0%), South Dakota (-3.8%), and Iowa (-3.2%). Lower prices for row crops likely explain the decline. During the first quarter of 2017 corn prices averaged 4.4% below the prior year and wheat prices were down 11.7%. Soybean prices were up somewhat compared to 2016 but were trending lower.

Nationwide the agriculture, forestry, fishing and hunting sector subtracted 0.41pp from the overall annualized growth, which totaled 1.1%. As the accompanying chart shows, absent the agriculture sector national GDP grew by 1.6% during the first quarter.

The fasted growing sector in Q1 was mining, which includes oil extraction. It was up 0.3%. As shown in the accompanying chart, excluding the mining sector national GDP would have only increased by 0.8%.


Texas 3.9% gain was the strongest. Its mining sector increased by 2.1pp. West Virginia experienced the second highest rate of overall growth at 3.0% and its mining sector contributed 3.2pp. New Mexico had the third highest growth rate at 2.8% with its mining sector contributing 1.8pp. Other energy states also experienced increased growth. Oklahoma’s 1.9% growth in output ranked 11th and noisy Alaska’s 1.8% growth ranked 14th. (A revenue estimator from the state told us they basically don’t even try to forecast revenue growth in the state because the economy is so volatile.)

But other energy states did not fare as well. Lousiana’s economy grew by only 1.0% (ranked 28th). Growth in Wyoming clocked at just 0.9% (32nd), and Montana saw a 0.5% decline (45th)

Finally, a couple of standout states for the quarter were Washington and Wisconsin. Washington continued a pattern of strong performance from 2016 when it took the blue ribbon. Its economy grew at an annualized rate of 2.7% during the first quarter and by 3.8% for all of 2016. This growth was driven by its information sector, which accounted for 1.7pp, and durable goods manufacturing, which accounted for 0.7pp, of the overall growth. Wisconsin’s 2.1% annualized growth for the first quarter represents a reversal of fortune from 2016 during, when its economy grew by only 0.9%. The sectors that contributed the most to its first quarter 2017 growth were real estate (0.6pp), durable goods manufacturing (0.4pp) and nondurable goods manufacturing (0.4pp).

by admin· · 0 comments · Uncategorized