I’m usually not paying too much attention to nonfarm payroll index. It is not a leading indicator and it’s usually not proving a good insight on how economy will be doing in the next several months. The only thing it tells is how the economy is doing right now.
However, from this month and until the recession starts (which timing I can’t predict, but it’s not too far away), I will start paying at least some attention to NFP.
The coming Friday report, based on just released ADP report, is likely to come at 95k to 100k, which is below consensus. If that happens, it could be another warning that good times are coming to end. However, if the report comes around 115k, as the consensus tells, then we still have time to enjoy.
Addition: The report came at 88k, weaker than my guess.
I saw the notion on the web and in the comments to my blog that strength of labor market is the strongest indication that we are not in danger of recession.
Nothing could be more wrong than that. First, the source states:
The most significant potential drawback to this or any model-based approach is that time series modeling assumes a predictable continuation of historical patterns and relationships and therefore is likely to have some difficulty producing reliable estimates at economic turning points or during periods when there are sudden changes in trend
Second, even after revisions we can observe that payroll data is a coincident indicator, not a leading one. The only thing we can conclude from today report is that we were not in recession in April and the chances to enter into recession in May are low.
But it sais nothing suggesting that recession cannot start in June
12 Months Net Change
Series Id: CES0000000001
Seasonally Adjusted
Super Sector: Total nonfarm
Industry: Total nonfarm
NAICS Code: N/A
Data Type: ALL EMPLOYEES, THOUSANDSYear Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Annual
1997 3040 2899 2961 3087 3020 2995 3043 2833 3121 3222 3237 3358
1998 3397 3288 3119 3106 3245 3204 3049 3411 3126 2979 2957 3003
1999 2848 3059 3035 3139 2952 3004 3178 3016 2982 3212 3222 3172
2000 3304 3028 3380 3283 3289 2989 2870 2666 2597 2163 2084 1948
2001 1690 1650 1127 547 295 208 -85 -228 -601 -919 -1443 -1763
2002 -1877 -2075 -2074 -1862 -1853 -1676 -1640 -1501 -1315 -867 -549 -530
2003 -338 -373 -523 -488 -498 -521 -430 -457 -307 -249 -204 107
2004 156 340 879 1214 1516 1594 1643 1792 1881 2055 2065 2065
2005 2046 2250 2031 2084 2003 2166 2434 2529 2464 2221 2507 2541
2006 2652 2717 2842 2646 2561 2420 2315 2299 2392 2394 2239 2263
2007 2219 2009 1937(p) 1881(p) p : preliminary
The raw 12-month trailing changes are tabulated above. As you can see the past 4 months (from January to April) map very well to similar data set of 5 month from September of 2000 to January of 2001.
I would say that this table suggests that we are now like in mid-January of 2001. As recession started in March of 2001 we can expect that this time recession can start somewhere in late June or early July, which is about 6 weeks from now
Addition: Holy crap! Indeed, the birth/death model added 317k jobs to this report. If, instead of 317k we will add the average for previous 6 months, which is 46k – today report would be -182k (yes, negative) NFP change!!!
May 3, 2007 at 2:51 pm
It is impossible to assign that level of precision to a number like non-farm payrolls. The difference between 90k and 115k is complete noise. With jobless claims averaging ~330k on a 4-week average and the 3rd week in a row with a lowered read, the labor market is looking plenty solid to dispel any ideas that the consumer is at risk right now.
May 3, 2007 at 9:54 pm
Actually the inaccuracy in payroll data is an argument against using “strong labor” to support the idea of “strong economy”.
Most likely we will notice first significant weakness in employment at he moment we’ll be already in recession.
It’s like using Sunday newspapers to track stock indexes in the past week. If Dow crashes in Tuesday you won’t know it for another 5 days
May 4, 2007 at 10:34 am
I completely agree with your second post. I didn’t mean to indicate the strength of the labor market is not a good barometer for strength of the economy. Quite the opposite I would guess. 70% of our economy is driven by the consumer and the strenght of consumption should be very highly correlated to the strength of the labor market. It’s the NFP series I don’t like. As you mentioned, it’s a lagging indicator (as evidenced by the 4th monthly reading of 3.9% unemployment in Dec, 2000 right before the Fed embarked on a 550 bps easing cycle). I think jobless claims are the best co-incident indicator. Claims ticked slightly higher in early April, but have since corrected. I think that is the key labor indicator to watch.
May 4, 2007 at 6:10 pm
[...] Jobs are now pointing that recession is possible very soon [...]