MNI DATA PREVIEW: US May Payrolls Forecast +185k;Downside Risk

     WASHINGTON (MNI) - The median forecast in an MNI survey shows analysts

expect May nonfarm payrolls to rise by 185,000, a more modest gain than the

surprise 263,000 gain in April. 

     Some analysts see a ramp-up of government hiring in preparation for the

2020 Census as a positive factor that could appear in the May data and the

months going forward.

     However, the incredibly soft ADP report lends significant downside risk to

forecasts. Analysts look for a 175,000 gain in private payrolls after the

236,000 increase in April.

     The Bloomberg survey consensus is for a 178,000 gain for overall payrolls

and 172,000 for private payrolls.

     Analysts in the MNI survey also expect average hourly earnings to rise 0.3%

after a disappointing 0.2% gain in April, the average workweek to rebound to

34.5 hours, and the unemployment rate to hold steady at the decades-low 3.6%.

The Bloomberg survey looks for the same readings.

     Ahead of the release on Friday, we outline important themes for particular

attention.

     --ADP reported a gain of only 27,000 in May, causing immediate concerns

that BLS private payrolls will suffer a similar fate. The ADP report tends to

deviate from the actual BLS payrolls number by a wide margin, missing by an

average of almost 65,000/month over the last year even when more accurate

estimates in March and April are included. However, even taking that into

account, the private payrolls estimates outstanding would appear to be

relatively high.

     --One reason analysts did not immediately lower their forecasts for private

payrolls after the ADP data was more positive readings from other, the most

recent of which being a solid gain in the non-manufacturing ISM employment

reading that printed shortly after ADP. Additionally, the manufacturing ISM

employment index was up, the Conference Board's jobs plentiful measure jumped,

and survey week jobless claims were only marginally higher than their April

lows.

     --In the last 20 years of data, analysts have overestimated May nonfarm

payrolls 11 times, and underestimated 9 times. While analysts overestimate only

slightly more often than they underestimate, they miss by a larger amount when

they underestimate, with an average underestimate of 96k versus an average

overestimate of 46k.

     --Analysts are calling for a payrolls gain of 185,000 while markets

anticipate a gain of 200,000. In the past year, markets have overestimated 4

times and underestimated 8 times, missing in the same direction as analysts each

time. On average, markets overestimate to a slightly higher degree than analysts

(69k vs 61k), while they underestimate much more closely (62k vs 59k). Given the

tendency of markets to miss in the same direction as analysts, markets also have

no clear directional bias to their estimates.

     --Both markets and analysts are calling for a 0.3% gain in average hourly

earnings, following a 0.2% gain in April. After April, analysts tend to miss

slightly more than markets, with 5 underestimates averaging 0.1%, and 4

overestimates averaging 0.1%. Meanwhile, markets underestimate as often as

analysts, with 5 underestimates averaging 0.1% and 3 overestimates averaging

0.2%. Given both the market and analyst tendency to underestimate, an upside

surprise to average hourly earnings is possible, but there is not a significant

risk in that direction.

--MNI Washington Bureau; tel: +1 202-371-2121; email: kevin.kastner@marketnews.com

[TOPICS: MAUPR$,M$U$$$]