Data released this week by the Australian Bureau of Statistics in conjunction with the Australian Taxation Office reveals that at least 6% of workers have lost their job over the past month, with the accommodation and food and arts and recreation industries absolutely smashed by the impact of the coronavirus.

One of the problems with the standard data releases is that while they are nice and regular, during times of big shocks, they can also be somewhat out of date. The GDP figures for example only come out every three months – and the next lot released on 3 June will only cover the months January to March.

That doesn’t mean they are unimportant, but it makes them less valuable when attempting to formulate policy in a fast-changing economy.

We saw another example of this when the latest unemployment figures for March only showed a small rise from 5.1% to 5.2%. The survey for each month’s labour force figures covers the first two weeks of each month, and in March the real employment carnage happened in the last two weeks.

That is why the ABS should be congratulated for the work they have been doing to provide more up-to-date information on a range of areas such as retail sales and jobs.

This week the bureau, together with the ATO, has been able to present the weekly change in employment and wages since the start of the year.

We normally don’t care about such “original” data, because it is too noisy:

“Original” data, for example, doesn’t take into account that jobs usually dry up in January and then come back in February as firms start hiring after the summer breaks. Usually we look at the trend or the seasonally adjusted figures which takes what has happened in the past into account.

But nothing like what is happening now has happened in the past.

This is why original data now is very useful because it shows just how quickly employment began to fall in the second half of March:

While the fall in jobs was pretty similar for both women and men, there were large differences across age groups.

In the last two weeks of March, employment fell 6% for all workers, but for those under 20 and over 70 it dropped by 9%.

The most secure were those in their 50s, and yet they even experienced a stunning 4% fall in jobs:

The likelihood of losing a job was also very much dependent upon what industry you worked in.

Education is the only industry to employ more people in the first week of April than it did in the first week of March. But accommodation and food services, and the arts and recreation industries are clearly the ones that have suffered the biggest falls:

A quarter of those in the hospitality industry and one in five of those in the arts and recreation industry who were working in March no longer have a job.

The loss of jobs in those two industries directly echoes the timing of the lockdown and forced closure of restaurants and arts and sporting events:

We can see how out of whack these overall figures are when we compare this original data with what has happened over the past 15 years.

Usually February sees a strong increase in jobs, followed by small increases or even a dip in March and then generally a bit of an increase in April (besides February, the best months to get a job are September and December).

So far this year we were seeing that play out – but if the numbers released this week continue, we will see April drop completely out of the scope of what occurred even during the GFC:

What this will mean for unemployment is hard to gauge.

How the jobseeker numbers will affect the “unemployed” numbers is mostly guesswork at this stage. Also on a month-to-month basis there is not much correlation between employment growth and changes in unemployment.

But if we look at annual changes we can get some idea of what is going to happen.

Usually we need employment each year to grow around 1.7% to keep the unemployment rate stable:

In the early 1980s and 1990s recessions when employment fell by 3.5%, unemployment rose by between 2.5 percentage points and 3.9 percentage points.

A fall of 6%? That would have us looking at best at a four percentage point rise – to around 9.2%. And that’s if all goes well.

We await the full data to show the economic damage of the coronavirus. But what we have seen already suggests we are experiencing something that has never happened before, and it is very scary indeed.

• Greg Jericho writes on economics for Guardian Australia