The Institute for Supply Management’s Non-Manufacturing, or Service, survey fell in March to 57.3 from 59.7 in February. Like its venerable brother, the Manufacturing survey, this is a “magic-50” index where any reading above 50 indicates that the economy is expanding and anything below 50 represents an economic contraction.

This means that the Service side of the economy, which is far larger than the Manufacturing side, is still growing, though the rate of growth decelerated in March relative to February.

This makes it 16 straight months that it has been over the magic-50 level. While that’s still solid, the number came in well below than the 59.5 consensus expectation, and is therefore a big disappointment.

Like the Manufacturing survey, this index is made up of ten sub-indexes that roughly correspond to the manufacturing sub-indexes. In this recovery, the Manufacturing side of the economy has been stronger than the Service side. That is still the case, as the Manufacturing index (released on Friday, and thus overshadowed by the jobs report) slipped to 61.2, from 61.4, a better-than-expected number.

However, it looks like both sides of the economy are doing very well based on the ISM data. The slippage on the Service side is a bit disconcerting, but the Manufacturing side is consistent with absolutely boom-like conditions.

Seven of the sub-indexes fell and Just two increased on the service side this month. One was unchanged. All but the index tracking import levels are above the magic-50 level (it was right at 50.0).

Business Activity Down

The performance of the more important sub-indexes were mixed, but mostly negative. The most important measure of current business activity is, well, the business activity sub-index. It plunged 7.2 points this month, but remains at a very healthy level of 59.7.

Fifteen industries reported higher business activity and just two reported a slowdown in activity. The business activity is subject to seasonal adjustment, and that appears to have been the primary reason for the decline; on a not-seasonally-adjusted basis, the index rose to 63.0 from 62.5.

Backlog and New Orders

The most important index for the very short-term future is the backlog of orders index. That rose by 4.0 points to 56.0. There were eight industries reporting an increase in backlog, and three with a decline.

As businesses work off their existing backlog of orders, they need to replace them with new orders. There the news is not quite as good, with a fall of 0.3 points to 64.1. That still indicates that new orders are rising at a very healthy clip, just not quite as fast as last month.

There were 16 industries reporting higher new orders and only one reporting a decline. It thus appears that business activity is picking up — some of that is just seasonal, but the activity will be sustainable since businesses are getting enough new orders that their order backlog is growing, and not just being consumed by the current activity.

Employment

The employment sub-index is also very important. The 1.9 point decline is disappointing, as it fell to 53.7. That is still well below the Manufacturing employment sub-index level of 63.3, which fell 4.7 points, but from the highest level since 1973.

Eleven service industries reported higher employment while three reported trimming payrolls. In Friday’s report, though, the vast bulk of the job creation appeared to be coming from the Service side, not Manufacturing. Manufacturing only accounted for 17,000 of the 230,000 private-sector jobs added, while Services added 199,000. Then again, there is a much smaller base of manufacturing jobs, just 10.7% of the total.

The employment sub-index has been above the magic-50 level for seven straight months now. The employment index is also subject to seasonal adjustment, but the affect was not as big as for the business activity index. On an unadjusted basis, it rose to 54.5 from 54.0.

One thing to keep in mind is that these are diffusion indexes (the reading is the number of positive responses, plus one half of negative responses). As such, they measure the number of companies that are adding or trimming jobs, not the number of jobs they are creating or losing. Thus it is not exactly an apples-to-apples comparison. Still, over time, the ISM employment indexes, both Manufacturing and Non-Manufacturing, have tended to track employment well.

The graph below shows the history of the key sub-indexes. Unfortunately, the St. Louis Fed has not yet updated its FRED database, so the numbers are only through February, but the graph does show the longer-term picture.

Very Disappointing

Overall, this was very disappointing report. It was far below expectations, and the drop in the composite index is the largest since November of 2008. However, the overall level is still quite strong.

It is highly unlikely that the economy will slip back into a double-dip recession if the ISM Service and Manufacturing indexes are both not only above the 50 mark, but one is above 60 and the other above 57.

As the report was below expectations, the stock market should not like it. The deceleration in growth is not welcome. The business activity index, in particular, dropped very sharply, but it was at its highest level since December 2004. While it was the sharpest drop in the business activity index since November 2008, there is a big difference between dropping from 66.9 to 59.7 as we did this month, and dropping from 43.6 to 34.5 as we did back then.

Even though it looks like seasonal adjustment was a big part of the drop, it is not at all what we want to see. The Non-Manufacturing index does not have the same sort of long history as does the Manufacturing index, but the level of the ISM manufacturing index is consistent with an economic boom in the making.

Paul Krugman (from this source) last month produced this graph between the quarterly average of the manufacturing PMI and the rate of economic growth (the PMI manufacturing data goes way back, so those quarters of 15%+ economic growth are actually from the 1930’s). While the fit is far from perfect, it is pretty solid. Plugging the numbers in the regression would predict economic growth of 6.32%. While I’m not sure that the data from the 30’s and 40’s that went into that calculation is still relevant, so I would not want to base my economic forecast only on the PMI data.

The table below comes from the ISM report and shows the sub-indexes for the Service index, as well as the corresponding sub-indexes on the Manufacturing side.



* Non-Manufacturing ISM Report On Business® data is seasonally adjusted for Business Activity, New Orders, Prices and Employment. Manufacturing ISM Report On Business® data is seasonally adjusted for New Orders, Production, Employment, Supplier Deliveries and Inventories.
 
Zacks Investment Research