The Covid Crash payback

A common pattern is emerging over the last couple of months, in Europe, the USA, Australia, and other places too.  Manufacturing PMIs are plunging, after a brief levelling off.  But service PMIs are rising sharply.  Here is the commentary from the UK news release from S&P Global:

The latest survey indicated a robust and accelerated increase in service sector output (index at 54.9), with growth the highest for one year. In contrast, manufacturing production (index at 48.5) decreased for the second month running and at the fastest pace since January. 

The contrasting trends for business performance in April largely reflected divergent demand patterns. New order growth hit a 13-month high in the service economy amid rising spending on travel, leisure and entertainment. Meanwhile, manufacturers attributed a renewed fall in new work to customer destocking, elevated energy costs and subdued demand for big ticket consumer goods. Similarly, export sales increased at a solid pace across the service sector, but manufacturers experienced a decline for the fifteenth consecutive month. 

I think the difference is due to a recovery from the Covid pandemic.  Manufacturing, directly impacted by rising rates, is struggling.  But people have been starved of travel and holidays and shows (travel, leisure and entertainment) by lockdowns.  And it's taken time for their plans to enjoy themselves to be realised.  So now they're flying away on holiday, staying at hotels and resorts, and going to see music and plays and bands once again, after a prolonged drought.

As this chart for the US shows, mostly the services and manufacturing sides of the economy move in sync, but manufacturing  (the blue line) tends to lead services (the orange one):

It is not often that services lead the business cycle---except with the Covid pandemic, because it was lockdowns (not monetary policy) which crushed services.  

In the chart below, note how the gap between manufacturing and services expands during recessions, with manufacturing falling faster than services.   Except, that is, during the Covid crash, when services fell much faster than manufacturing (the spike at the beginning of 2020).   The gap the other way between manufacturing and services now is "payback" for the gaps when services were below industry, evident since the beginning of the Covid crisis.

Click on chart to see a clearer image.
Chart shows gap between manuf and services PMIs
Shading shows US recessions

Having explained this anomaly to my satisfaction, and I hope yours, the obvious question is:  when will this end?  Is the post-Covid recovery in services over yet?  

It won't be until unemployment starts rising.  The problem is that, if overall GDP remains robust, because services are strong, and price increases in services remain high, Central Banks will go on tightening.  And because of the lags involved, they risk tightening too much.   By the time they realise their mistake, it will be too late.   But at that point, unemployment will be rising fast, and services will have followed their manufacturing brethren into recession.  

How many months away is that?  I don't know.  Any ideas?  Comment below.