Scenes from the June jobs report: a weak but improving economy

 

 - by New Deal democrat

There’s no important new economic news today, so let’s take a somewhat deeper dive into last Thursday’s employment report for June. To cut to the chase, this is of a piece of my “Big Picture” narrative from Friday in which I wrote that most of the signals suggest and economy coming out of a weak patch rather than going into one.

Let me start with a comparison of the typically leading manufacturing and residential construction sectors vs. services, which sometimes barely decline at all during recessions. Here is the historical look from 1982 to just before the pandemic:

Typically before a recession in a single month 100,000 jobs or more would be lost from manufacturing (red), and 10,000 or more lost from residential construction (orange), while service providing jobs (blue) would only start to decline once the recession had begun.

By contrast, as shown in the post-pandemic graph below, manufacturing’s biggest monthly decline was 60,000 in late 2024. That sector has recently shown more monthly *gains.* Residential construction’s biggest monthly decline has been less than 6,000 last August and this May. Service producing jobs have increased every month since January:

Not only have manufacturing jobs increased in recent months, but take a look at one of the 10 “official” leading indicators, average weekly hours worked in manufacturing jobs:

This metric has increased sharply in the past 24 months, and is now tied with its post-pandemic high. There has never been a case of a recession with manufacturing hours so strong.

And the YoY change in total payrolls bottomed last winter, and has increased gradually since:

Total YoY employment gains are still very weak, but the negative trendline appears to have been broken.

Next, there has been much comment about the decline in the employment-population ratio and the labor force participation rate in recent months. But a closer look at each indicates that in the prime age 25-54 year demographic (red in the graphs below), there has barely been any decline at all. First, here’s the labor force participation rate:

And here is the employment-population ratio:

This looks very much like the tail end of the large Boomer generation shuffling off into retirement and thus skewing the comparisons, rather than a generalized weakness. The sudden dropoff in June looks like noise that may be revised next month.

Finally, let me update two statistics that I cite almost constantly. First, almost every week I point out that jobless claims (blue in the graph below) have had a 60 year history of leading the unemployment rate. In the past few months, it appeared that the unemployment rate (orange in the graph below) had stalled. But when we carry the comparison out another decimal point, by disaggregating the unemployment rate into the number of people unemployed divided by the civilian labor force (red), the gradual decline this year becomes apparent:

I expect we will see a further decline in the unemployment rate (at least as decomposed) before it bottoms perhaps several months from now.

And last of all, here is the updated graph of aggregate nonsupervisory payrolls (red), together with its component parts: total hours worked in the economy (gold) and average hourly earnings (blue); compared with the inflation rate (black), all normed to 100 as of 12 months ago:

While total hours have barely budged since last November, average hourly pay has more of less kept even with monthly inflation, except for the Iran war spike. We won’t get June’s inflation number until next week, but the Cleveland fed estimates it will come in just below 0, rounding to -0.1%. Since last June rounded upward to +0.3%, the YoY inflation rate is likely to decline roughly -0.4% to 3.8%. Since aggregate nonsupervisory payrolls also declined -0.1% in June, this means that real aggregate nonsupervisory payrolls YoY will remain higher by 0.7% — very weak, but not recessionary.

The economically weighted ISM indexes for June show continued growth, equipoise in employment, and still strong inflation at the producer level

 

 - by New Deal democrat

The economically weighted ISM manufacturing + services indexes have become one of my favorite datapoints. In particular, in the past year they have been an excellent dashboard for the state of the US economy as a whole, both as a coincident and short leading indicator. Last summer they accurately suggested that the economy was very close to recession, while since the beginning of this year they have indicated improvement. 

To recapitulate, services are about 75% of the economy and goods production the remaining 25%; hence, their weighting. To reduce noise and amplify signal, for forecasting purposes I use the three month moving average of the weighted average. One caveat is that these are diffusion indexes, which tell us how widespread a trend is rather than its net value.

Last Wednesday we got the manufacturing report, which was very positive for new orders, and for the first time in many months showed that jobs were not contracting, and price increases were less widespread than before. This morning’s ISM services report was similar, in general showing a slightly weakening but still quite positive sector. [Note: in all the graphs below, the manufacturing number is in blue, and the services number in gray].

The headline services number declined -0.5 to 54.0. The three month average was unchanged at 54.0 as well. This compares with the manufacturing three month average of 53.3. Thus the weighted average is 53.8:

New orders also declined -2.2 to 55.1, and the three month average declined -1.8 to 55.3. Since the manufacturing average was 55.6, the weighted average is 55.4:

If new orders decelerated, there was more good news about employment, the index of which rose 3.3 to 51.2, returning to expansion after several months of contraction. The three month average also rose 3.0 to 50.0. Since manufacturing rose to 49.7, and its three month average was 48.2, the weighted average rose to 50.5 - the first positive number in four months:

Finally, as with manufacturing, inflationary pressures in services abated somewhat in June, as the prices paid component declined -3.6 to 67.7. The three month average declined -1.3 to 69.6 Since the three month average for manufacturing was 79.9, the economically weighted average declined -4.4 to 69.2:

Note that the above graph, unlike the first three, goes back five years to show the comparison with the post-pandemic inflationary spike.

Let’s recapitulate:

•  Both the headline and new orders components of the services index showed deceleration, but were solidly positive, as were the economically weighted manuacturing+services average.
•  The employment situation has improved from contraction to equipoise or slight gains
•  The prices component still shows widespread inflationary pressures at the producer level, but not as strong as the previous few months.

The economically weighted ISM averages indicate the economy is expanding now, and can be expected to continue expanding for at least a few more months, but that the inflationary spike primarily driven by the Iran war, but also the computer chip shortage, continues to be a serious problem.

The United States of America at 250: the enemy of Republics is entrenched power

 

 - by New Deal democrat

On America’s 250th anniversary, the state of the Republic is not good.

Ten years ago, after the 2016 election, I embarked upon a reading journey, to seek an answer to the questions: are Republics fragile? Are they doomed to fail, or fall into tyranny? And so I began reading books about every Republic, ancient, medieval, and modern that I could find, covering the time period from the emergence of civilization to the present, including the city-states of ancient Greece, Rome, Venice, Genoa, Geneva, Switzerland, the Netherlands, and the emerging modern UK (and others I’ve forgotten at the moment).

And I came to a surprising answer: despite the notoriety of the fall of the Roman Republic, the fact is that republics are probably more stable than despotic or monarchic forms of government. The reason is that despotic rulers fear those around them who are *too* competent, and thus threats, and so succession falls away very quickly to incompetent and relatively easily overthrown rulers, while the competent contenders are frequently assassinated by the ruler. Similarly, monarchies seldom last more than a century without a dynastic war breaking out, in part for the same reason that by the time you get to the great-grandchildren of the founder, the competence of the monarch is very much in doubt. 

So while the *form* of despotism or monarchy might last for centuries, the direct line of succession without violent dynastic war is usually much shorter. For example, the line of succession that began with Julius Caesar only lasted until the assassination of Nero in A.D. 68, and internal to that line was the assassination of Caligula and the choice of his successor, Claudius, by the praetorian guard.

By contrast, rulers in a republic do not have to fear the competence of their likely successors; and so competent leadership tends to continue. Even the Roman Republic itself lasted 500 years, hardly a transient state. And there are other examples of long-lived republics, such as Venice (over 1000 years), the Swiss confederation (800 years and counting) and the post-Glorious Revolution UK, in which Parliament is supreme (335 years and counting). And the form of succession in most republics is by some sort of “democratic” vote. In the Roman Republic, for example, each tribe got one vote, and that vote was determined by an internal majority of each tribe. In many other republics, voting was restricted to an oligarchy or aristocracy, but leadership was chosen by that vote. Hence, in the Venetian Republic, only once in its long history was there ever an attempt at an autogolpe.

Republics have historically engendered great loyalty by their citizens, because those citizens see that their own security and chances of success are bound up in the security and success of the republic itself.

But it is only because wide groups of citizens see the chance of success that they continue to be loyal to the republic.

David Frum remarked about 20 years ago, about reactionary conservatives, that if they do not see the possibility of success in achieving their goals in the democratic process, they will not change their values, but rather jettison democracy. That observation should not be limited just to right-wingers. If any large group of citizens does not see the possibility of ever being able to enact their goals into law, they will not abandon those values or goals, but rather abandon a commitment to the process of selecting a government.

Which brings me to the crux of this piece: the enemy of Republics is entrenched power, that cannot be displaced by the voting process of that republic, even if that entrenched power has only minority support. To go back to ancient Rome again, it was the inability by the ordinary plebeians to ever displace patrician power that led them to abandon the norms and forms of that Republic’s government. 

In the US today, we see two wellsprings of entrenched power. On the one hand, we see the old evangelical WASP majority, especially in areas like the Old South that until recent decades never saw a big influx of immigration, determined not to let go of power by any means. On the other hand, we see increasingly entrenched wealth that receives ever more tax cuts every time the GOP is in power, to the point where one man now is worth 1/32nd of the entire nation’s GDP.  That entrenched wealth is never going to give up its power voluntarily. If that power cannot be displaced by a majority of the citizens of the country, if it uses every lever of power to hold on to its control, eventually the forms of the Republic will give way.

Just this past week, we saw a corrupt Supreme Court relegate Congress to the sidelines, holding that once Congress creates an agency, it no longer has any control over it. And only a few days later, we saw that 4 of the 9 justices were ready to torture even the clearest of phrases, that “All persons born or naturalized in the United States, and subject to the jurisdiction thereof, are citizens of the United States ….” into an interpretation that is nearly the opposite of what has been believed for well over 100 years. If there is no way to overcome the fiat of the Supreme Court, if even legislation and Constitutional Amendments are not sufficient to negate such sophistry, where indeed is the “rule of law” which is the defining feature of republic, at all? If there is no President too is above the law, then what exactly is worth of the “law” itself?

Have we, to mix metaphors in this essay, already crossed the Rubicon in to an era of entrenched power that cannot be displaced by the tools available in this Republic? I have an opinion, but I leave the answer to this question to each reader.

Weekly Indicators for June 39 - July 3 at Seeking Alpha

 

 - by New Deal democrat

My “Weekly Indicators” post is up at Seeking Alpha. 

Two cross-currrents I mentioned in my long piece yesterday were very much in evidence in the high frequency data this week: huge gains in aggrgate YoY consumer spending, and very weak withholding tax collections. My take on this is that the top end of the income distribution is doing very well, thank you, while the lower levels are struggling.

A usual, clicking over and reading will bring you up to the virtual moment as to the state of the economy, and reward me a little bit for my efforts.

My Big Picture nowcast and forecast at mid-year 2026 (it may surprise you!)

 

 - by New Deal democrat

Earlier this week I promised an updated Big Picture forecast and nowcast for the US economy, plus a note about how broad or narrow was the participation. 

Most commentators either have a financial or ideological agenda - getting you to buy their stock recommendations, or selling you on DOOOM, cherry picking the data du jour that buttresses their opinion - or else they simply project current trends forward, and thus miss turning points. By contrast, for over 20 years I have doggedly followed a variety of the same (growing list of) economic indicators, following their historical records as to how soon or late they telegraph turning points. No single indicator or system is perfect, but if you go by the rule “it’s not different this time,” you will be right much more often than you will be wrong.

Only once in that entire period have I forecast a recession: at the end of 2006, I said one would likely follow in a year or so. And boy howdy did it! Several times since (2018, 2022) I have gone on Recession Watch, and a few indicators even warranted a Recession Warning, but never beyond that. Of course I did not foresee the Giant Flaming Meteor of Death in March 2020, but I think that can be excused!

One important lesson of all this is that it takes an enormous disruption (an oil shock, COVID, a speculative credit implosion in the housing market) to cause the US economy to go into reverse. Otherwise, a disruption on one side is likely to be offset by continued growth elsewhere. A second is that you need to leave your political ideology at the door. A prime example is T—-p’s first term of office, where he was gifted with an economy entering a good growth period, and did not know where the levers of economic power were enough to disrupt it. And so far in his second term, it has withstood both Tariff-palooza! last year, and - so far - the Iran war this year.

So where do we stand now? To cut to the chase, unless the continued closure of the Strait of Hormuz causes gas prices to go to about $6 a gallon or more, there is no imminent danger on the horizon to the economy as a whole. But the benefits of the expansion are very skewed.

A good place to start here is my alternative “consumer nowcast” system. This posits that increased consumer spending can be fueled by several sources: increased real incomes, a decline in financing interest rates, or increased asset prices. Only if all of those fail will consumers pull back and cause a recession.

Here is what real incomes (dark blue), real wages (light blue), house prices (gold, /10 for scale), stock prices (red, /10 for scale) and mortgage interest rates (orange, inverted so that “down”=“bad”, right scale), all look like together:

Both real incomes and wages have turned down this year with the inflation caused by the Iran war. Mortgage rates remain elevated, and so a negative. House prices are flat, so cashing out increased equity is not an option for most homeowners. 

But stock prices have increased almost 20% in the past year, creating a huge asset class that can be tapped for spending by the uppermost income tiers. So long as that remains the case, the “consumer nowcast” is not forecasting an imminent recession. That’s a major caveat, because if the AI data center construction boom is in fact a Bubble, when it pops it will take down the stock market with it; and if the other indicators remain negative, then the nowcast would suggest a recession would follow in short order.

Now let’s turn to my other system of long leading, short leading, and coincident indicators. I don’t plan on being exhaustive here, but want to give you a representative look.

The long leading indicators include corporate interest rates, the bond yield curve, real money supply, housing permits, corporate profits, and real retail sales per capita. Several weeks ago I indicated that government expenditures (i.e., any stimulus measures in effect) should also be added. We’ve already seen above that interest rates remain elevated, although they have been at these levels for about three years, so they are now a neutral. And most people know that the yield curve, which had been inverted, have normalized. Here are the two most commonly followed metrics, the 10 year minus 3 month, and 10 year minus 2 year, comparisons; as well as real M2 money supply YoY:

The former regularized about 9 months ago, and the latter over 18 months ago. Historically, by the time this long a period of time has passed since regularization has passed, recessions have been over and a recovery has begun. Additionally, real money supply turned positive almost 2 years ago. In short, these are positive.

Next, here are what corporate profits (/2 for scale) look like in comparison with real retail sales per capita:

While real retail sales per capita declined in 2024, and then were generally flat in 2025, they picked up in the past few months. Corporate profits, meanwhile, with the exception of the first half of last year, have remained very positive. Since these are long leading indicators, this means that the producer side is a positive, while the consumer side remains a neutral (since the trend one year ago was moribund).

Finally, while I won’t post a graph of government expenditures, note that the “Big Beautiful Bill” gave the wealthy lots of more money to play with in the form of tax cuts.

In short, there is a neutral interest rate background, a rebound in the bond yield curve and money supply, and a positive trend in corporate profits. Only real retail sales per capita are troublesome. Absent a shock (like what might happen if the Strait of Hormuz remains closed), these are not forecasting recession in the next 12 months.

Next, let’s examine a general trend in the short leading indicators. Last week I wrote that the “quick and dirty” system of stock prices and jobless claims had had a stellar record in the last 10 years, and was very positive now. Additionally, real retail sales in the aggregate (rather than per capita) were not negative YoY either. Here is a repeat of that graph:

While housing (as shown by units under construction) and motor vehicle sales (as shown by truck sales) have both been recessionary, both have stabilized or rebounded in the last few months:

And although I won’t repeat the graphs, I have written many times in the past few months about the rebound in manufacturing data, like core capital goods orders, the regional Fed new orders indexes, and manufacturing employment and hours. All of these are positive. The only real soft spot has been construction employment and real spending, which has been flat to trending slightly lower.

In other words, the bulk of the short leading indicators have been trending higher as well.

The one big caveat is the real aggregate payrolls, which tend to turn shortly before a recession, peaked in January and have been lower since, although they are not negative YoY at this point:

These are worth a yellow flag. Much there depends on what happens going forward with inflation. But again, the bulk of the short leading indicators are positive.

Finally, in terms of coincident indicators, we did look at real personal income above, noting that it had turned down. Real manufacturing and trade sales turned down in the most recently reported month, but the trend has been higher. Industrial production has also been positive. And finally, employment, growth of which was basically nonexistent for most of 2025, has been mildly positive in the past few months:

In other words, the bulk of the coincident indicators suggest current growth, but that the bulk of average American households may not be participating.

So let me sum up this Big Picture nowcast and forecast of the American economy at mid-year 2026:

1. Almost certainly, we are not currently in a recession, although the bulk of the gains are among the top tier of income earners and the wealthy.

2. The large majority of signs indicate that the economy is not going to tip into contraction in the next few months, although again the real payrolls for the sum of nonsupervisory workers have contracted this year, an important caution.

3. The long leading indicators are most consistent with a rebound going into 2027.

Finally, let me note that all of this can be tossed into the trash can if there is a geopolitical shock. The most likely candidate for such a shock is that the Strait of Hormuz remains closed, and the gamblers in the oil futures markets, who have been betting that it will reopen before the Strategic Oil Reserve is drained to crisis levels, are wrong, causing gas prices to rocket to $6/gallon or higher. Or the Chief Chaos Agent in Washington DC does something else that completely derails the above trends. A second major caveat, as indicated above, is if the AI data center construction boom is a Bubble that pops, the short leading indicators on the producer side are likely to turn negative in a hurry.

June employment report: weakly positive headlines, and important negatives in hours and payrolls

 

 - by New Deal democrat

My Big Theme for the past few months has been that the AI Boom (or possibly bubble) is counterbalancing a stagnant or even shallowly recessionary rest of the economy. After three good reports in a row, the June employment report was very weak, and in some respects even negative.

Below is my in depth synopsis.

But first, because this report was release on a Thursday, let me note that initial jobless claims continued very low, unchanged at 215,000. The four week moving average declined -2,500 to 222,000. With the typical one week delay, continuing claims rose 2,000 to 1.814 million. More importantly, on a YoY basis, initial claims were down -6.9%, the four week average down -7.1%, and continuing claims down -7.2%. This series continues to be one of the two most positive short leading indicators for the economy, along with the stock market.

HEADLINES:

• 57,000 jobs gained, Private sector jobs increased 49,000, while government jobs added 8,000. The three month average rose declined to 111,000, about average for this year, and better than last year.
• The pattern of downward revisions to previous months completely resumed this month. April was revised lower by -31,000, and May was revised lower by -43,000, for a total decline of -74,000.
• The alternate, and more volatile measure in the household report, declined sharply, by -507,000 jobs. On a YoY basis, this series was negative for the fifth month in a row, now sharply down by -963,000 jobs, or over -80,000 per month
• The U3 unemployment rate declined -0.1% to 4.2%. 
• The U6 underemployment rate declined -0.2% to 7.9%.
• Further out on the spectrum, those who are not in the labor force but want a job now declined -142,000 to 6.085 million, in the average range for the past 12 months..

Leading employment indicators of a slowdown or recession

These are leading sectors for the economy overall, and help us gauge how much the post-pandemic employment boom is shading towards a downturn vs. rebounding. These were mixed.

• The average manufacturing workweek, one of the 10 components of the Index of Leading Indicators, was unchanged at 41.6 hours, tied for the highest number in 5 years, equalling its 2021 peak.

• Manufacturing jobs rose 3,000, only the 3rd increase in the last 12 months.
• Truck driving continued its decline, by -1,300.
• Construction jobs rose +11,000.
• But Residential construction jobs, which are even more leading, declined -2,900, taking out their interim low from last April, and setting a new 3 year low.
• Goods producing jobs as a whole rose +10,000. 

• Temporary jobs, which declined by over -650,000 since late 2022, rose by 9,300, continuing to improve from their post-pandemic low set last October.
• The number of people unemployed for 5 weeks or less declined -28,000 to 2.182 million, about average for the past 12 months.

Wages of non-managerial workers 

• Average Hourly Earnings for Production and Nonsupervisory Personnel increased $.07, or +0.2%, to $32.38, for a YoY gain of +3.4%, the lowest in 5 years low. This is also substantially lower than the 4.2% YoY inflation rate as of May.

Aggregate hours and wages: 

• The index of aggregate hours worked for non-managerial workers *declined* -0.4%, and is up 1.0% YoY, still a good comparison with the past 2 years.

• The index of aggregate payrolls for non-managerial workers also *declined* -0.1%, and is up 4.5% YoY, about average for the past 2 years, but only 0.3% above the YoY inflation rate through May.

Other significant data:

• Professional and business employment rose for the third month in a row, by +36,000. These tend to be well-paying jobs. This remains above its October low, and has finally turned higher YoY as well.
• The employment population ratio declined another -0.2% to 59.0%, vs. 61.1% in February 2020, and its lowest since October 2021.
• The Labor Force Participation Rate declilned -0.3% to 61.5% , vs. 63.4% in February 2020, and the lowest since February 2021. IMPORTANT: both the EPOP and LFPR are greatly affected by the retiring Boomer population. In the prime age 25-54 demographic, they are virtually unchanged.

SUMMARY

After three  good monthly reports in a row, this was a big stumble, although it remained mainly positive, with the major exception of hours and payrolls.

The positives included the declines in the unemployment and underemployment rates, the headline employment number, plus increases in the leading manufacturing, construction, and general goods producing sectors. Temporary help jobs and professional and business jobs increased. 

But there were important negatives as well, including the volatile monthly Household Survey employment number, as well as the leading sectors of residential construction and truck driving. But the most significant negatives were the YoY change in nonsupervisory hourly wages and the actual declines in both aggregate nonsupervisory hours and payrolls. The first two are almost certainly negative in real terms, once we find out what the June inflation number is, and the last one may have edged closer to flat YoY as well. This would mean that this very important short leading indicator will remain below its January peak for the 5th month in a row, which in the past has strongly suggested a recession is near.