Historical Nonfarm Payroll Data – An Insight

January 27, 2014 in Historical and Research

Here is an insight of the historical non farm payroll data and the interrelationships of nonfarm employment with population growth and the strength of the U.S. dollar strength.

Nonfarm payroll data (often referred by the acronym NFP) is one of the most important economic releases to come out of the United States. It shows employment conditions and job growth in non-agricultural businesses in the U.S.

Unexpected changes in NFP data have been known to trigger very volatile moves in currency markets. Payroll numbers for December 2013, released January 10, 2014, surprised global markets with a smaller print than expected, causing the dollar to weaken sharply.

The resulting volatility again brought into sharp focus the significance of this data, particularly in these uncertain times when growth across the globe comes at a premium and markets have become so inter-linked.

This research note, inspired by the effects of the December 2013 NFP release, seeks to analyze the historical non-farm employment changes versus U.S. population growth and the strength of the dollar. The analysis covers 24 years of historical data commencing 1990

What happened on January 10, 2014?

The NFP report for December 2013 showed an increase of just 74,000 jobs compared to the addition of 241,000 positions in the previous reporting month. The consensus expectation was for an increase of 196,000 jobs.

The disappointing data caused the so-far bearish EUR/USD to jump up by 117 pips. A recovery being attempted by USD/JPY was short-circuited – instead it crashed through the psychological resistance of 105 and then extended the fall to 102.85, clocking a massive loss of 149 pips.

Similar volatility was observed in other currency pairs, too.

Part I: Historical non-farm payroll data (seasonally adjusted)

Before we go to the various charts, let’s have a look on the historical change in the seasonally adjusted nonfarm employment. The following table shows month by month change in the nonfarm payroll numbers since the year 1990.

Note 1: All the following figures are in thousands. The months mentioned are preceding months i.e. October 1990 would mean the data as on November 1, 1990. For the ready reference you may also wish to keep an eye on the up-do-date nonfarm payroll data since the year 1939 i.e. past 64 years. The non-farm payroll numbers are updated in case the originally released data is revised later on.

Yr/Mo

2014 2013 2012 2011 2010 2009 2008 2007 2006 2005 2004 2003 2002 2001 2000 1999 1998 1997 1996 1995 1994 1993 1992 1991 1990 1989 Average
Dec  – 74 219 230 95 -220 -705 93 169 158 128 119 -163 -178 136 298 344 300 168 132 272 307 212 23 -59 95 90
Nov  – 241 247 174 144 -21 -775 111 205 335 65 13 8 -295 227 293 280 300 296 149 422 261 139 -58 -146 277 116
Oct  – 200 160 166 228 -170 -472 86 -3 81 343 197 121 -331 -12 405 194 339 243 146 208 278 178 12 -159 111 102
Sep  – 175 138 225 -43 -233 -459 77 159 65 155 105 -55 -243 125 204 219 507 219 244 355 239 35 32 -85 249 96
Aug  142 238 165 132 -37 -210 -270 -24 179 192 125 -44 -11 -155 1 193 353 -16 197 272 298 160 141 17 -208 47 69
Jul  212 89 153 78 -86 -351 -216 -35 210 372 37 20 -100 -125 165 290 118 281 232 79 364 295 70 -48 -40 39 76
Jun  298 172 87 209 -130 -472 -169 80 80 246 78 -3 47 -128 -47 266 212 255 279 232 313 174 60 88 17 117 83
May  224 176 125 115 521 -352 -186 141 21 168 306 -10 -9 -44 224 214 399 256 323 -16 333 265 126 -127 150 118 129
Apr  282 199 112 304 229 -704 -215 76 181 362 247 -51 -84 -282 288 374 278 290 160 160 353 308 157 -212 40 173 110
Mar  203 142 205 205 154 -830 -79 186 280 135 333 -215 -24 -28 471 106 144 313 264 223 462 -51 52 -161 214 192 108
Feb  197 332 271 196 -40 -695 -85 90 316 240 43 -158 -146 63 122 404 188 299 424 210 201 242 -66 -302 249 258 106
Jan  129 148 311 69 -13 -794 14 234 274 130 159 89 -129 -15 248 127 273 231 -21 322 270 310 50 -122 338 262 111

(Source: U.S. Bureau of Labor Statistics)

Note 2: The yearly average shown is from the year 1989 to 2013. The average will be updated to take into account the 2014 figures at the end of the 2014.

In our analysis we will ignore the numbers of the years 2008 and 2009 which showed drastic job cuts due to the fallout of the Lehman Brothers collapse triggering global turmoil. Interestingly, we note that job losses had actually started before the Lehman Brothers shock. Job growth embarked on a declining trend from February 2008, whereas Lehman Brothers filed for the bankruptcy on September 15, 2008.

Part II: United States population growth rate and changes in nonfarm payrolls

The payroll data should not be analyzed in isolation. We should also take into account the population and the population growth rate. The following table shows the year wise population growth of the United States since 1990.

Year wise population growth of United States

Year

Population

Growth (%)

2013

0.71%

2012

0.70%

2011

0.70%

2010

0.80%

2009

0.90%

2008

0.90%

2007

1.00%

2006

1.00%

2005

0.90%

2004

0.90%

2003

0.90%

2002

0.90%

2001

1.00%

2000

1.10%

1999

1.10%

1998

1.20%

1997

1.20%

1996

1.20%

1995

1.20%

1994

1.20%

1993

1.30%

1992

1.40%

1991

1.30%

1990

1.10%

United States Census Bureau projected the total population as 317.3 million on January 1, 2014. According to 21st United States census the population of the U.S. in 1990 was 248.71 million i.e. an increase 68.59 million people i.e. a 27.58% increase. Please note that this data is for total population and not employable workforce. During the same time i.e. from 1990 to 2013, the nonfarm payroll numbers increased by 27.942 million. In January 1990 the total nonfarm payrolls were 109.394 million and hence in percentage terms the nonfarm employment increased by 27.29%. 27.58% population increase resulting in 27.29% nonfarm employment increase, doesn’t look so bad, right? It will be difficult to get into the deeper analysis of each and every point here, otherwise this article may turn into a book, however if the data is not very bad, it is not very good either. The macro view of the possible reasons is as follows:

Urbanization

Automation in farming and urbanization should in turn result in the increase in nonfarm employment. Automation is in all industry but balance in the employment number needs to be there.

Native versus immigrant population of the U.S. (Historical view)

The approximate number of immigrant population in the U.S. was 19.8 million in 1990s and it went up to 40 million in 2010. Not the percentage increase of 102.02% but the absolute numbers are more important here. Most of the legal immigration takes place with job guarantees or because of employment. Hence logically the percentage of employment of the immigrant population needs to be quite high. Now if we look at the data from 1990 to 2010, the number of immigrant population increased exponentially by 20.2 million. If we keep the increase in employment (nonfarm) which is 27.942 million from 1990 to 2013 and increase in immigrant population from 1990 to 2010 which is 20.2 million then the employment data may look much worse.

Before we go to check the visual representation of the comparison of the employment data in the United States against the population growth rate in a graphical way, it may be interesting to draw out some more information about the bests and worsts of these results from this data.

Year % Change in Nonfarm
2013 -0.32%
2012 4.28%
2011 105.77%
2010 -120.23%
2009 39.67%
2008 -424.39%
2007 -46.16%
2006 -16.63%
2005 23.03%
2004 870.67%
2003 -138.17%
2002 -69.05%
2001 -190.40%
2000 -38.63%
1999 5.73%
1998 -10.52%
1997 20.51%
1996 29.31%
1995 -44.09%
1994 38.13%
1993 141.59%
1992 -234.50%
1991 -375.88%
1990 -83.95%

The Bests and Worsts of Nonfarm Payroll

Before we go into the further analysis, let’s try to sort this data a bit more to draw some other useful information. Here are the bests and worsts of nonfarm employment figures.

Monthly average change since 1990

  • Best month: The best month is usually May. The average change in May nonfarm employment, during 1990 through 2013, has been 129,000.
  • Worst month: The worst month has been August. The average change in nonfarm employment during 1990 to 2013 has been 69,000 in August.

Best and worst years in absolute terms

  • Best Year: The best year was 1994 during which 3,851,000 jobs were added in the nonfarm category. The year 1994 saw net job additions during every month, right from January to December.
  • Worst Year: Unsurprisingly, the worst year was 2009, the year after Lehman Brothers filed for bankruptcy and the global financial crisis hit. The United States saw a fall of 5,052,000 job positions in the defined nonfarm sector in that fateful year. During 2009 no single month went without net job cuts. The second worst year was the worst year itself in the recent economic history. It was 2008 when the only month which saw an addition of 14,000 jobs and all other months witnessed job cuts. The other year which saw most months with job cuts was 2001. During 2001 the only growth in the nonfarm employment was seen in the month of February. The worst terrorist attack in the history was 9/11 which had major economic impacts. However the negative growth in the nonfarm payrolls had started much before September 2001. Year 2000 was a good year with an addition of 1,948,000 nonfarm payrolls but 2001 saw 11 months going with net job cuts. Was it because of the job losses due to the end of Y2K issue? Probably not but let’s analyze that also and we will cover that soon

Best and worst years as percentage over the previous year

  • Best year: 2004 is the clear winner here. That year witnessed a growth in employment in the nonfarm category by 870.67% as compared to 2003. In the year 2003, jobs grew by only 62,000, but during 2004 2,019,000 jobs were added by goods, construction and manufacturing companies in the U.S.
  • Worst year: The worst year was hands-down 2008, which witnessed a fall of 424.39% in non-farm employment. About 3,617,000 jobs were slashed this year, whereas 1,115,000 jobs were added in 2007.

Nonfarm Payroll Additions Versus Change in Population – Historical Chart

The following chart represents the year wise historical changes of the population of the United states and the change in the nonfarm employment in the respective year.

Year wise historical chart of US population growth versus nonfarm payroll additions

Now last not but least, let’s also see how the U.S. dollar performed against the performance of nonfarm payrolls historically. The charts below compare the US dollar index which measures the strength of the US Dollar against a basket of currencies 6 major currencies i.e. the euro (EUR), Japanese yen (JPY), British pound (GBP), Canadian dollar (CAD), Swiss franc (CHF) and the Swedish krona (SEK). The data ranges from the year 1990 to 2013

Historical chart of % change in nonfarm employment versus the US dollar index

The following chart shows the comparison of the historical changes in the US dollar index and the nonfarm payroll numbers since the year 1990.

Chart of historical comparison of nonfarm payroll data and the US dollar index.

Has there been any correlation?

Let’s just have a look if there has been any correlation between the strength of the US dollar against the performance of the nonfarm payroll data historically. Let’s check the above charts after joining some of the peaks and valleys. The following chart represents the correlation between the year wise change in the non-farm employment versus the US dollar index since 1990.

Correlation of nonfarm payroll and historical US dollar index

As we can see from the charts above that there has not really been any strong correlation between the strength of the US dollar index with the nonfarm employment. Interestingly, however, we can see that the correlation has been more on negative side i.e. when the performance of the nonfarm payroll was not good, the US dollar index strengthened and vice versa. The following chart shows this correlation between the non-farm payroll employment and the strength of the US dollar index, historically:
Year wise correlation between nonfarm payroll and US dollar strength.

3 responses to Historical Nonfarm Payroll Data – An Insight

  1. Thank you Himanshu you show a good information that is nice for forex trader.I think that every trader should to read this article for good information and gather more knowledge aboutfundamentals.

    Wish to have more article like that.

  2. Thanks Carol. A nice comment always helps after you put is some days work on a single post. There was a mistake (not in data) and that has been corrected. For ready reference we have posted all historical data of nonfarm payroll on another post and that will be kept up-to-date for ready reference. Planning some more on other topics as well.

    As far as Gold is concerned, have been planning something on that too but wish to write something which is really in-depth and not shallow.

    -Himanshu (Tokyo)

  3. Well Himanshu, you have quite out done yourself. I loved this information and it will prove very informative in the future. I’m going to print it off.
    I don’t suppose you analyse Gold by any chance? MT4 Charting is not very good for working ABCD’s or E.W. No matter if you don’t.
    Keep up the good work

    Best wishes
    Carol, from the UK