Uncategorized

Jan jobs report puts RBA into neutral

Posted by

The Jan jobs report caught the market by surprise, with the unemployment rate falling ~14bps to 5.08% (mkt 5.3%). Given the size of the standard errors, we can be about 75% confident that the unemployment rate actually did decline in January, and extremely confident that it is lower now than it was in q3 (when it peaked at ~5.3%).

The jobs numbers are a joke – but for those that care it was +46.3k (mkt +10k). We can be about 75% sure that the number of (seasonally adjusted) jobs rose in January.

The errors get larger as we break the survey up, however one split I regard as worth the effort is the male full time unemployment rate. This declined 10bps, to 4.9% and is now 40bps below the August peak.

The hazard and exit rates, from the matched sample (gross flows) data show that the probability of an employed person losing their job has fallen back to pre-softpatch levels, and that the prospects of an unemployed person finding a job have improved somewhat.

So all three of the bits that are well measured – and the jobs number – confirm that the labour market is improving. This should not surprise, as it is consistent with the bulk of leading employment indicators.

On balance, the evidence suggests that the unemployment rate is at least steady at ~5.25%, and probably declining. And if the unemployment rate is not “rising modestly” before “declining modestly over the latter part of the forecast period” then the RBA is going to have to bring forward that 25bps acceleration of core inflation (to 2.75%) from H1’14.

An earlier acceleration of inflation means that they need to think about being tighter than 4.25% a little earlier. This makes further cuts unlikely – unless the unemployment rate spikes up again, or the inflation rate is lower than expected.

This conclusion fits neatly with a good rule of thumb for RBA policy – the RBA moves their policy rate when the unemployment rate is moving, and in the opposite direction.

So if the unemployment rate does not start rising again, the RBA almost certainly will not be cutting their policy rate any further.

19 comments

  1. Stephen Koukoulas was on the ball.

    I suspect we will not be seeing anymore rate cuts unless Europe implodes

  3. I’d be very cautious about drawing too many conclusions about a single months change. Especially one impacted by seasonal volitility. The movevents in the Jan report were just a rebound from last month. The peak of employment from young women in Dec never happened so the drop in Jan didnt happen either. This screwed with the seas adj and was the main reason behind the movements at the top level. I’d wait a couple of months to get a clearer picture.

    1. Broadly speaking, i agree – we need more data. But with the point estimate of the unemploymwnt rate falling, the rba is now most probably on hold while they await that data.

      1. Agreed, too much noise to draw any aonclusions either way. Having said that, demand may have turned the corner (anz job ads) which might support a strengthening in the Labour market. But again not nough data to be certain either way.

        1. i have a sense that the dynamic that had all the talking heads calling multiple rate hikes in Q1’2011 is now playing out – it’s too early to be sure, but if you squint hard you can maybe just see the spillovers from the mining boom starting to appear outside that narrow sector. ironic that the vast majority have given up on the theme just as it starts to play out.

      2. i have a sense that the dynamic that had all the talking heads calling multiple rate hikes in Q1’2011 is now playing out

        Seriously?! I see Chris Joye has already jumped to all kinds of conclusions, just as he did when we had that rogue inflation number last year.

        but if you squint hard you can maybe just see the spillovers from the mining boom starting to appear outside that narrow sector.

        I’m squinting, really squinting, but all I see is flat-as-a-tack employment growth since Q4 2010. And now you see jobs growth spreading from the mining sector just as the China construction boom hits the skids, global LNG prices are at 10 year lows, and iron ore prices are 30% off their highs and falling.

        Really, I think you guys just see what you want to see. You want to see !!MINING BOOM!! jumping out at you, and you’ll squint at data until you see it.

  4. Participation rate hasn’t recovered yet, and total hours worked dropped significantly. Plus it was January. I still think the job market is flat, maybe Feb will clarify.

    1. Agg Hours worked is a synthetic series therefore monthly changes are likely to be noisey…especially around xmas new years.

      1. What do you mean by synthetic series? Don’t they simply ask to each individual on the sample “how many hours have you worked in the reference period?” and then come out with full time / part time / unemployed / employed and total hours?

        There’s an explanation going around that since in December there was not the usual temp hiring (the number was down seasonally adjusted), then in January those jobs were not gone, so we gained, but only seasonally adjusted. But overall still flat, without the Dec jump.

        It’s funny when you hear “the economy added +46k jobs” when actually we lost jobs / worked hours as usual in January, but just less than in other years (and again it can be related to missing Dec jobs).

        Overall I think it’s wise to wait for Feb/Mar before concluding where we stand. cheers

        1. That was certainly my assessment of the Dec data, and it ‘explains’ the jump in jobs in Jan. but i think it is likely that the uptrend in unemployment is over.

          Sent from my iPad

      2. Its a syhthetic series in that they inflate the numbers for the reference week up to the whole month. During this process the try to account for things like holidays and other regular seasonal effects.

        Yeah the explanation I heard was the drop in Dec and subsequent jump in Jan was just from 15 to 24 yo women. The peak that happens every Dec wasnt as big as usual (so it was adjusted down) then in Jan because the Dec peak didnt happen the usual drop wasnt as large so it was adjusted up. Thats just an unavoidable problem with seasonal adj i suppose. Still the E to P ratio is still pretty high by historical standards so Its really pretty good.

        1. I told myself basically the same story. The non spike in dec retail hrs will also have pushed down the Jan hrs data.

          Sent from my iPad

  5. The Lorax – I actually think this blog has been very circumspect when analysing the data. You’re just in a huff and a puff because your bearish view of the Aus economy is wrong – again.

    1. Mate, go talk to anyone, anyone at all outside the mining sector and tell me I’m bearish.

      1. @the lorax – the data is what the data is. If you want to concentrate on subjective opinion well thats your problem, but subjective opinion is no match for the cold hard facts and they are the e to p ratio is pretty high by historical and international comparisons and the unemployment rate is low by the standards. Meanwhile the demand side of the equation is starting to turn the corner as well. All in all the data has been saying for a while now that the Labour Market isn’t in that bad a shape…yes ppl concentrate on the level estimates but the employment rate has been solid after recovering from the gfc as has the unemployment rate. A few wackos will point out roy morgans random number generator but it is just that, an imprecise opinion poll not designed to measure the subtle changes in the labour market over time. So I agree with Ricardo and BK there are no FACTS to support your opinion.

      2. Anecdotal evidence counts for nothing. Unless you’re sample is representative of the population (which unless you’re a statistician, then it won’t be) then I think you should stick to something more tangible

  6. Pingback: Jobs and population | ricardian ambivalence

Comments are closed.