美国劳动力市场（3月）：经济数据反应

Economics-Data Reactions United States Whipsawpattern ◆Nonfarm payrolls rose 178,000 in Marchfollowing a 133,000 decline inFebruary;however, webelievetheunderlyingtrend for hiring remains modest Ryan WangUS EconomistHSBC Securities (USA) Inc. ◆The unemployment rate fell to 4.3% in March from 4.4% inFebruary and isnearlyunchanged compared to a year ago +1 212 525 3181 ◆Thelaborforceparticipation rate fell to 61.9% in March, the lowest since late2021; slowernetimmigrationmaybeweighing ontheparticipationrate Facts Nonfarm payrollsrose 178,000 in March (consensus+65,000). February was revisedlowerto-133,000from-92,000, whileJanuary was revisedhigherto +160,000 from +126,000.Therefore, thenet revisiontoJanuary and Februarywas-7,000. Average hourly earningsincreased0.2% m-o-m (consensus +0.3%) and 3.5% y-o-y(consensus+3.7%)inMarch. The unemployment ratefell to 4.3% in March (consensus 4.4%). Thelabor force participation rate fell to61.9% (consensus62.0%) from 62.0%, the lowest since November 2021. Implications We wrote last monththat badweatherduring the mid-January through mid-February periodmay haveweighed onreported nonfarmpayrollsforFebruarybydisruptingthe typical hiringactivity that would haveotherwise occurred (US Labor Market(Feb): Nojobs growth, 6 March 2026). If so,we thoughtemployerswould havesome need for catch-up hiring in March andbeyond.Moreover,theBureau of Labor Statistics Current Employment Statistics (CES) “strike report” indicates that a strike in Disclosures & Disclaimer:This report must be read with the disclosures and the analyst certifications in theDisclosure appendix, and with the Disclaimer, which forms part of it. Issuer of report:HSBC Securities (USA) Inc. the health care industry involved 31,000 workers from 26 January through 23 February 2026. Thetimingmeansthe strikesubtractedfrom February nonfarm payrolls,butthe end of the strike added to March nonfarm payrolls. Giventhese factors as well as the inherentvolatilityof the data, we would nottake much signalfrom the whipsaw pattern of nonfarmpayrolls sofar this year(strong January, weak February, strong March).One reasonable approach would be to average the results;for example, the six-month averagechangeintotalpayrolls stands at +15,000, indicating a very mutedrate ofoverallemploymentgrowth.The six-month average change in private sector payrollsisstronger, at +53,000. The main reason for the gap is an ongoingdecline in federal government employment(one that continued in Marchwith a drop of 18,000). Wecommentedlast month that we viewthe stability ofaverage hourly earnings growth(figure 2)asan indication that the overalllabor market is in a balanced state–one where neither firms nor workers hold an outsized degree of bargaining power.In March, themonthly increase in average hourly earnings came inslightlybelowconsensus estimates. But we still view theunderlying trendasconsistentwith wage growth of around 0.3% per month (and perhaps 3.5% or a bit higher on a y-o-y basis). From a historical standpoint,it ishighlyunusualto think that an environmentof near-zero employment growth could be one wherethe labor market is broadlyinbalance.Yet there are reasons to think this may be an accurate characterization ofcurrent labor marketconditions. Indeed,the unemployment rate fell to 4.3% in March (figure 3). The unemployment rate isclearlyhighercompared totheApril 2023 low of 3.4%, whichoccurredamidstintense hiring needs when broad sectors of the economy were reopening afterpandemicshutdowns.Butcompared with 12 months ago,the unemployment rate is nearly unchanged. Thecurrentunemploymentrate also sits very close to the FOMC’s median estimate of 4.2% for the“longer-run”unemployment rate.Arecent note from FederalReserve Board staff1noted that“labor force growth has been slowing and could be near-zero this year, driven by weakpopulationgrowth reflecting low net immigration and bydeclininglabor force participation reflecting population aging.”Theauthors said thatnegative job growth could be“almost as likely as positive job growth in any given month”butalso thatbreakeven employment growthwouldbe near-zero (hence, verylow employment growth would not necessarilylead toa higher unemployment rate). The labor force participation rate (figure 4) hasfallennotably over recentmonths, but wehesitateto draw a firm conclusion aboutwhy this hasoccurred(we had thought the March rate might rebound a bit,perhaps inpart reflectingbetterweather). We still think itpossiblethat data noise isplaying a roleand that the very recent declines areexaggerated.That said, it isalsopossible thata steepslowdown innetimmigrationsince the middle of 2024is beginning toweighonlabor force participation (in addition to thedemographicheadwindfor labor forceparticipationfromanagingpopulation)in a more pronounced fashion. This could be a logicaldevelopment sincenewand recentimmigrantsmay skewtoward prime-age cohorts with higher participation rates.We will be lookingclosely a