The OECD stylised estimates of direct GDP effects

Earlier this week the OECD put out a useful little note Evaluating the initial impact of COVID-19 containment measures on economic activity.  

It is less than five pages long, including a couple of charts, and isn’t even attempting to offer definitive answers.   And what it offers isn’t an assessment of the impact of the differing containment measures being adopted in each country, but of the likely impact of the initial economic structure of each country’s economy to a set of common assumptions about the extent of containment and associated changes in economic activity.   Thus when, as local media have reported, this chart shows that New Zealand is one of the most adversely affected economies

oecd covid 2

it isn’t telling you anything about the impact or intensity of New Zealand policies, but simply about the pre-existing structure, given the (common across countries) assumptions OECD staff make about which sectors are most hit as containment measures take hold.  We don’t have many high-tech online services firms and, on the other hand, we have a lot of construction and a lot of production accounted for by tourism. (If you are wondering about Ireland –  far left –  remember that GDP in Ireland is massively inflated by some of the corporate tax distortions, and the loss of national income will be much larger than is shown here).  I was interested in the number of central European countries close to us on the chart, which appears to reflect the importance of the motor vehicle manufacturing and supply chains in those economies (and Germany), and of course car sales –  and production – have collapsed.

It is also important to remember that these are direct effects only.  That might not much affect the cross-country comparisons but will affect the absolute magnitudes of the likely falls in activity.  On the numbers in the chart, the direct effects alone crunch GDP by about 25 per cent for the median country shown.  Indirect effects, including confidence effects, credit availability effects, uncertainty effects, income effects and so on will probably amplify these numbers greatly.  I struggle to see how a fairly extensive lockdown will not reduce GDP while it is in place by up to perhaps 60 per cent (I noticed the comments yesterday by the chair of ANZ Bank in New Zealand, John Key, who was using numbers of that sort of order of magnitude).  Measurement is going to be a huge challenge –  and it may be several years down the track until, say, SNZ can offer us fairly definitive estimates (especially in the absence of quarterly income –  profits –  data) –  but reality (even if not accurately measured) is more important than whatever the published headline numbers.     Even in most industries deemed “essential” activity levels are going to be very significantly reduced (not many new mortgages being written for example, production capacity in meat works reportedly half what it was in norma times, and so on).

The OECD’s second chart –  using the expenditure approach to GDP –  is consistent with these sorts of really large numbers.  It focuses on consumer spending (rather than total GDP), and they show estimates for only a few countries.

oecd covid 3

Read the footnotes, and if anything these look as though they could be underestimates, at least in a New Zealand context (eg pretty sure car use here has dropped by more than half, and hotel/restaurant spending by more than three quarters (since all restaurants are closed and hotels must have little more than a few residual stranded tourists).

Of course, imports will be lower, which offers some offset.  But so are exports, most notably our tourism and business travel exports.  And, in any properly measured sense, a fair bit of public consumption spending is going to be affected too – teachers might be getting fully paid, but no one supposes school online is going to be anything like the extent/quality of the normal product.

And then there is investment.   Last year, almost 24 per cent of GDP was accounted for by investment spending (gross fixed capital formation).  60 per cent of that in turn was construction (residential and other).  None of that is happening this week.  Very little of any other investment spending will be occurring either, whether because it is simply illegal for firms (or bits of government) to operate or because of the extreme uncertainty of the environment, sharp reductions in forecast demand, or whatever.

It is difficult to get anything like a fully comparable estimates of the extent of the “containment measures” various countries have taken, partly because there are so many dimensions to most countries’ restrictions (and in places like the US, restrictions are at a state or county level), and because whatever the headline how the rules are applied in practice also matters (who knew, for example, the Arobake’s luxury baked goods counted as “essential”?).   But relative to at least the other Anglo countries –  with whom we often compare ourselves –  New Zealand’s partial lockdown seems to be more extensive.  If so, the immediate economic costs here are likely to be greater, especially as –  first chart – if the OECD rough and ready analysis is right then our production structure meant that for any given set of restrictions we were more exposed than the other Anglo countries.

That isn’t a commentary (at all) on the merits of the extent of the New Zealand lockdown. It also isn’t a commentary on whether a tougher approach now will, or won’t, reduce the overall economic costs across time (there are a couple of papers around on that latter issue which I will try to write about next week).  It is simply to make the point that many of the estimates that are around for the extent of the fall in New Zealand’s GDP –  including the sorts of numbers the Ministe of Finance alluded to at the select committee on Thursday – still seem on the light side.   We don’t have monthly GDP data (they have such estimates in Canada and the UK) but if we did then for the period of the current partial lockdown I’d have thought you would have to struggle to produce estimates of a fall as small as 50 per cent.


33 thoughts on “The OECD stylised estimates of direct GDP effects

  1. When I was a programmer we had a reply “How long is a piece of string” when we were asked for answers to similar impossible questions. GDP: the monetary value of all finished goods and services made within a country during a specific period.
    So what period? For the period of last week Mr Key’s 60% reduction sounds about right; whatever happens in this epidemic eventually we move out of lockdown if only for psychological reasons. My guess for ‘eventually’ is three months maximum so a GDP reduction of 60% during lockdown could equate to only 15% reduction as an annual figure. Unlikely given the probable future state of tourism and possible distaste for crowded venues.
    Given the number of foreign workers leaving and Kiwi emigres returning to NZ who knows what our population will be? Isn’t GPD per capita more significant to the individual than the national total?
    It is the term ‘monetary value’ that keeps me returning to this website. I know the state of my finances and can estimate changes to my property values but what will my disposable money buy at Xmas?


    • My, slightly glib, answer to your final question is probably (a) more than you expected (inflation will be v low), and (b) but less will be available than you hoped (supply disruptions, esp around non-essentials from non-core countries abroad (so when i was stocking up pre lockdown I focused not on staples v much, but on, eg obscure food items from abroad that I routinely buy and use slowly, and which i worry might not be around reliably for the next few years.


      • My rents on 11 properties all paid fully up to 31 March 2020. My wage for the last 2 weeks fully paid less 1 day due to the lockdown. I stocked up on most day to day requirements which will last me for 4 weeks. I finally got my 145 litre freezer from the Warehouse which was ordered 4 days before the lockdown but only received when the government loosened the criteria that whiteware was an essential. So I went shopping for more frozen goods. Queued for 15 minutes with 1 metre spacing at countdown at 7pm in 3 Kings after driving past Mt Roskill New World and Mt Roskill Countdown which looked like 30 minute queues.

        Other than bananas everything else was full priced. Usually you would see large 20% or 30% discounts. This does mean that prices for groceries is up 20% to 30% now compared to before lockdown. But because I am not driving and not buying $13 lunches. The increased grocery costs matches off not spending on retail priced lunches and lower petrol costs.

        It feels like a holiday at the moment. The calm before the storm. The wage subsidy has certainly taken the panic from employers because it is around $7,000 paid upfront which will cover employees wages at 80% plus some fixed costs. But it also means that cash subsidy will burn up in 4 weeks. The real panic will start kicking the guts out of businesses after 4 weeks.


  2. So Michael we may find ourselves on the other side of the pandemic with an economy reduced by at least a third and possibly by as much as 50 percent? Meanwhile we have a population touching on five million people that has increased by more than 25 percent in the last 20 years (it was 3.85 million in 2000) mainly through mass, low skilled immigration? This is a recipe for social and economic disaster.


    • I think that would be to overstate things. While the full lockdown is on GDP would be lower by 50 per cent or more. Once it was eased there would be some bounce back, but as discussed in yesterday’s post. Also important to remember that we could eliminate the virus here but would probably still have GDP perhaps (but big margins of error round this number) 15 per cent lower ( no tourism, lots of uncertainty etc etc).


  3. Curious on your thinking about the magnitude of hit to long term ‘wealth’. A few make the point that (aggregate) physical and human capital has effectively been mothballed rather than permanently impaired. I guess the hotel buildings are still standing but will be underutilised for some time – albeit, if they have an economic life of 30yrs, perhaps the long term value hasn’t reduced that much. Ditto for premises that house SMEs – some of which will regrettably fail. I guess a stock v flow query and to what extent physical/human capital is being permanently impaired?


  4. I suspect it will be a little bit like the debates about whether there were permanent effects of the financial crisis and great recession of a decade or so ago. I’m pretty sceptical, at an economywide level, altho clearly the recovery was pretty protracted in many places and some cohorts of individuals will have been permanently scarred, not just financially. It will probably be very difficult to tell this time as well, including because the crisis is likely to precipitate changes in approaches to policy in various areas that are v hard to guess at this point (in addition, eg, to higher taxes).

    My starting point is that with say a 10 year horizon in mind this virus itself won’t leave longer-term per capita GDP materially altered. From my quick look at what literature there is on 1918, I’d say that is probably a fair reading of that episode, which did kill many more people in the West (than we are counting on current policies saving us from).


  5. I have just completed a model run looking at the costs and benefits of a one month close down in the building and construction sector, The exercise was purely illustrative( and i am suffering from flu like symptoms ( 1:10000 chance of cV?) so the results a may not be entirely robust. the cost (250,000 jobs at an average cost of $3,000 a week) where $3000 million and the benefits ( lower deaths and illnesses over the next year ) around $150 million. A benefit/cost ratio of .05 which sadly is not too bad by government standards
    The model seems to be misbehaving a bit so i am not too confident in the detail – it looks like the reduction in deaths -mostly elderly – will be around 30 or so.


    • A model should include the cost in Doctors and Nurses (over 50 Italian doctors have died). Hypothetically if this disease was precisely the same but without the symptom of breathing difficulties – just the same number of almost random deaths but no hospital stay then it would not impact the health service and we would quickly learn to live with it – in fact it would be a minor concern to the elderly compared with fears of the far more common dementia.

      I wrote that paragraph and realised I am wrong. It is the number afflicted in a short period that frightens us. Compare with the public reaction to air accidents and road accidents – nobody has proposed an Erebus like memorial to those who died on our roads in the last 40 years.

      Liked by 1 person

  6. Best case scenario – it’s only 19 days to go

    The objective of the lockdown is to break the transmission threads. 28 days is double the incubation period. If everything has gone to “plan” it should be possible to lift the lockdown. To protect that, all off-shore arrivals should be immediately quarantined. No mucking around. No self-isolation stuff. NZ should be able to get back to normal activities. The measure of that will be the number of positive tests still occurring in the final 5 days of the lockdown period

    Of course tourism, associated tours, accommodation, Hotels, Motels, Airb’n’B’s, Universities, and ethnic restaurants will all be seriously diminished.

    On the matter of tourism, the government is still sitting on $53 million derived from the passenger levy that was to fund tourist facilities in the regions as in campervan parks and toilet and ablution infrastructure. Not a dollar has been spent. Now will be the time to use the break to get in and do it.

    Liked by 1 person

    • Yes. And I’m hoping there is a specific initiative to push out our winter seasonal flu shots in the weeks prior to the lock down being lifted. I’ve been told from 15 April our local surgery will open them to the general population. Seems a bit tight, as I think they have a 7 day period before they become effective and an individual is protected? Our health services get over-loaded every winter as it is. I dread thinking about dealing with COVID plus normal seasonal flu at the same time.


      • It would have been cheaper for the government to take control of the Auckland International airport and the Air NZ building which are empty for the foreseeable future and convert to ICU beds than to burn all the governments budget and wreck $150 billion or half the NZ economy to feed starving and jobless people. Idiocy at the highest levels of government and their advisets.


      • I don’t agree with your policy bottom line, but am not going to spend the time debating it. But on your specific, setting up new beds would not generate new capable doctors etc.


      • Michael, that is wrong. You do not need more doctors. You need ICU ventilators because it is mainly a lung infection and breathing issue. Death mainly is from the lack of oxygen in the body. Basic level nurses ie how to ensure ventilators are working is what is required.


      • Ok (will take your word for it) but even if so don’t overlook the heightened infection risk to medical staff from really intense peaks, when tiredness, equipment shortages etc can mean lots more deaths among that group (as in Italy).


      • The medical process of intubation first requires an anaesthetist, and then a person specifically trained in intubation technique. The latter to my knowledge in NZ is all doctors/junior doctors and a small number of post-graduate (specialist) nurses trained in intensive care who are members of the New Zealand College of Critical Care Nurses.


      • Sea-Long Medical Systems has drawn huge interest from hospitals and foreign countries seeking alternatives in treating COVID-19 patients amid a ventilator shortage.The Sea-Long device resembles a crude spacesuit helmet, with a transparent hood sealed at the neck and two tubes extending from its base. The helmet was originally designed to supply oxygen to patients receiving treatment in hyperbaric chambers.
        But doctors in Italy, where a version of the helmet has long been used to treat people experiencing breathing problems, found it to be effective in helping some COVID-19 patients.
        Dr. Bhakti Patel, who has been studying the devices for six years, said they hold promise as an early intervention that could spare respiratory patients the need to be put on the more traditional — and costly and invasive — ventilators.

        Patel believes the device could lead to a sea change in intensive care units that have long relied on traditional ventilators.


  7. If the number of positive Covid-19 tests continue to erupt causing the lockdown to be kept in place, then disadvantaged businesses can blame these twerps. They don’t care.

    Surfers defy lockdown ban at Muriwai. Surfers are still out in the waves on Auckland’s west coast this morning despite all water-based activities being explicitly banned during lockdown. Surfers are still out in the waves on Auckland’s west coast this morning despite all water-based activities being explicitly banned during lockdown

    Rodney Local Board chair Phelan Pirrie said there was a “steady stream” of cars heading to the beach this morning with surfboards attached. He had been told about confrontations between angry residents and surfers who were flouting the lockdown rules. They’re just being defiant. People have contacted me and said they’re aware of the [rules] but they’re just not interested.”

    If they were in Australia they would get instant fines of $1000


    • That is another Jacinda Ardern lie. Blame someone else is her motto operandi. Successful viruses like all parasites try to live within the host. This virus will hibernate if conditions are not right. They flare up when conditions are ripe. If covid19 behave like the flu then even with a vaccine, last year we got 500 deaths from 500,000 infections. The depth of winter in NZ and we get peak infections. The flu virus activates in winter. Covid19 looks the same. Don’t blame others. Expect a flareup as winter cold sets in.

      This Level 4 emergency lockdown is a waste of resources. The weather is still too warm and the UV radiation too high. Transmission and multiplication rates in NZ is still lower than the ordinary flu which means that more than likely we will have a winter flareup which will require another lockdown or we continue under lockdown indefinitely which is really not an option. We are not testing enough. Easier to blame others.


  8. It’s hard to know what will be left of the economy after lock down. I wouldn’t want to be the person balancing health & lives vs the economy right now. What are we talking here like a 1% death rate vs a 50% hit to the economy? It’s ugly either way.

    Looking forward to seeing the Govt’s plan post lock down.


    • So far 1% death rate is an exaggeration. We currently have as of Saturday, 1039 infected with only 1 death and only 3 in ICU. This officially makes Covid19 less deadly than the ordinary flu in NZ which killed 500 from 500,000 infections last year.

      This level 4 alert emergency lockdown is starting to look like a unnecessary and costly joke.


      • 50% hit to the economy equates to $150 billion. Bye bye NZ. Expect 20% to 30% inflation rates as our NZD collapses. Actually it is already here. Check your latest grocery bill. You are already paying much more than before lockdown, the reality of a collapsing NZD.


      • Don’t forget that the comparator here is what would happen if we had no lock down – which would be many more cases and many more deaths. Expect NZ’s death rate to climb as the disease matures or if it hits pockets of older patients.

        Liked by 1 person

      • There are many countries like Singapore as example that have not gone full lockdown. They show only a marginally 4 out of 1000 deaths which is not significantly higher than NZ, 1 out of 1000. Singapore yesterday reversed and eased its lockdown with the emphasis of getting people back to work.


    • Probably time and age differences. Our cases are ramping up much later and our population is younger than Italy’s.


      • Or our NZ Covid19 is still in hibernation unable to multiple quickly under our specific conditions and will flare up in a vengeance in the colder depths of our winter like the normal flu. The evidence suggests slow transmission and only 1 death with 3 in ICU in our current warmer and higher UV climate conditions. The real test would be the cold wet winter not long after we come out of our 4weeks lockdown.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s