Robbie Andrew

scholar orcid twitter
CICERO Center for International Climate ResearchHome   |   About me   |   Contact me

Norway's CO2 emissions during a pandemic

First published: 12 June 2020.

The year 2020 will long be remembered as the year of the global pandemic, which has had extraordinary consequences in almost every imaginable way, with societies turned upside-down. In Norway it appears that the early and relatively comprehensive lockdown, combined with public compliance, government economic support and an effective health system, have led to less physical suffering than might have been otherwise. But clearly such restrictions on people's behaviour have led to economic, social, mental, and environmental effects. In this post I just want to focus on the effects of the pandemic and our responses to it on the emissions of greenhouse gases in Norway.

We do not have any official estimates yet for Norway's CO2 emissions in 2020, and won't have any until the middle of next year. What do we do in the meantime if we want to know what's going on right now?

What do we expect the consequences of Covid-19 to be on Norway's emissions? Let's look at five categories:

  1. We know that certain branches of the economy were shut down during the lockdown: kindergartens, schools, universities, restaurants, bars, cafes, gyms, cinemas, theatre, sports, and various others.
  2. We also know that these closures combined with social distancing advice and personal efforts to reduce disease spread led to reductions in people's movement.
  3. We know that the borders were effectively closed.
  4. We know that people's behaviour was affected in other ways, including greater use of home-delivery services.
  5. We know the world outside of Norway has changed, and that will have some impact both on Norway's exports and its imports.

The shut-down economy

Let's look again at that list of economic sectors forced to shut down. They are all business categories where people gather in significant numbers in close proximity and with significant throughput (people coming and going). In other words, they were considered ideal disease transmission hubs. But they're also all in the service sector, and they don't use enormous amounts of energy to produce their products, compared to, say, a steel mill. Shutting these service industries results directly in lower consumption of electricity for lighting and heating, but that doesn't amount to much in the context of the whole economy, and since most of this will be supplied by Norway's hydropower, has little effect on emissions. Heavy industry in Norway faced no orders to shut down operations, and with adjustments to reduce contact between staff, they have been able to operate as usual, except for cases where export demand has been affected (see later).

Staying at home

The second category is the one where we're likely to see the largest effects. When people stay at home, they're not commuting. We know that while not everyone had the luxury, nevertheless large numbers of people worked from home. Here we have some data to play with.

From quite early in the pandemic, Google and Apple both released summary data collated from mobile phones, with the express purpose of assisting authorities in tracking the effects of measures taken. Google's data are based on location histories for those who have opted in to that service, while Apple's data are based on requests for directions. These are now both available in machine-readable format.

The chart below shows the changes in Norway compared to Google's baseline period of 3 January to 6 February 2020. The raw data are rather noisy, largely because of signficant differences between weekdays and weekends, and holidays, and I've removed these to reveal a cleaner signal, although this is necessarily approximate.

The spike of visits to grocery stores immediately after the lockdown was announced on 12 March is about 50%, but this isn't much more than the normal pre-holiday grocery shopping in Norway, visible in the raw data. On holidays, many shops, including most supermarkets, are closed, and many people shop the day before.

Google's data also include time at parks, but since this is compared to a winter baseline, there's a significant blending of the effect that people are out more simply because it's warmer.

The restrictions have had a significant effect on air travel within Norway, as clear from the figure below. Domestic passenger numbers were down 87%, but the number of flights was down only 54% as many flights were flown with very few passengers.

More time at home translates into less time driving, and this shows up in fuel sales statistics from SSB. Total sales of petroleum products in Norway were down 21% in April. Petrol sales were down 26%, road diesel 18%, but construction diesel unchanged. May data will be available from SSB on 19 June. A survey carried out in April showed that almost 60% of petrol stations in the country expected to make a loss in 2020: they have low margins on fuel sales, but high margins on everything else they sell.

SSB estimates that this drop in fuel consumption in April led to a drop in emissions from road traffic of about 20% for the month, compared with April last year.

Closed borders

Air traffic was down dramatically, with international passengers through Avinor's airports down 98.8% in April, and much the same in May. However, the number of international flights was down 92.8%, partly because freight flights (e.g. salmon exports) were only down 13.2%.

While emissions from international air traffic are not included in Norway's international reporting, and are not part of the country's international responsibilities, these emissions are nevertheless at least partly the result of actions by Norwegians and Norwegian companies, and form part of the larger picture of Norway's contribution to global emissions.

Daytrips across the border to purchase goods and services (known in Norwegian as harrytur or grensehandel) declined substantially already in the first quarter of 2020, as did the amount spent. Norwegians are attracted by the substantially lower prices across the border. We'd expect this to decline even further for the second quarter, as the border was effectively closed and anyone returning to Norway faced a two-week home quarantine as the cost for making the daytrip. This reduction in expenditure on groceries (almost 90% of which is in Sweden), because it represents a reasonable share of total expenditure on groceries by Norwegians, is one possible reason for seeing more time spent in Norwegian grocery stores in the Google data, above: instead of shopping in Sweden, we're now shopping more at home.

I haven't seen an estimate of the CO2 emissions associated with 3 million daytrips to Sweden, the vast majority by private vehicle. Departures are concentrated in areas close to the border, and are probably replaced by shorter-distance trips in private vehicles to local supermarkets and liquor stores.

Some share of those who regularly purchase groceries in Sweden will also purchase their fuel there, so fuel consumption in Norway might have declined even further than the Norwegian sales statistics suggest.

Our changed habits

I'm sure there will be many theses written on the events of 2020, but those are yet to come. I don't pretend to have full oversight of how our behaviour has changed during this pandemic, but point to some specific examples.

Despite grocery stores remaining open, a desire to prevent catching or spreading the disease led to a surge in the use of online grocery delivery services, and in fact this started several weeks before lockdown was announced. This does mean more vans driving around, with higher emissions, but this must be offset against the fact that some share of people normally drive their private vehicle to the supermaket (unless that vehicle is electric). Vans delivering food will have lower emissions per kilogram of groceries than someone driving a car to purchase food only for their family. So there's a potential here for a reduction in emissions, although it's not likely to be large.

As discussed above, daytrips across the border to buy goods and services have declined sharply. Most of this expenditure is on food, drink and tobacco, with only 15% other goods. While the effective closure of the border has led to reduced purchasing of tobacco products by Norwegians in Sweden, the import of tobacco products has increased by 50% by value, although some share of that increase is a result of the lower value of the Norwegian krone. People haven't stopped smoking, but higher prices in Norway have caused some to curb their habit.

But when it comes to food, the effect has been that already in March grocery sales increased by 13% in Norway, and alcohol sales were up by 15%. This was especially pronounced in the municipalities in the southeast bordering Sweden, but restaurants being closed has meant more people preparing meals at home across the country.

If this translates into more food production in Norway, then the emissions associated with food production will also increase, although these are effectively transferred from Sweden's accounts to Norway's, assuming similar production methods.

One interesting change in food consumption patterns is that sales of root vegetables surged. Various potential explanations spring to mind: a return to safe, traditional food types; foods with long shelf-life; foods that can be simply washed and peeled to remove any infection risk; foods that are easy to cook; and so on.

Research by NINA (the Norwegian Institute for Nature Research), using opt-in data from training app STRAVA, has showed that outdoor activity by users in Oslo increased by close to 300% following the lockdown, which - unlike in some other countries - did not restrict such activities. In the figure below, the blue line at top-right shows the increased outdoor activity following lockdown.

The world around us

With signficant reductions in electricity demand in the EU, Norway's exports of liquefied natural gas (LNG) have faced a steep drop in prices. Combined with the low value of the Norwegian krone, this means significant drops in export income. Whether this has translated into lower production of natural gas (and indeed crude oil), I haven't yet discovered, but if it did, then emissions associated with that production would have declined. Emissions from the production of oil and gas in Norway amount to over a quarter of total emissions.

UPDATE 16 June 2020: While the export value of crude oil exports was substantially down in March-May, the volume is virtually unchanged, see SSB Figur 3 here; the export volume of natural gas in May was 14% down compared to May 2019.

The following figure shows the LNG price in the EU (in US dollars per thousand cubic feet) from Fitch Ratings.

SSB's weekly international trade statistics show little effect of the crisis on the value of imported goods, but a marked reduction in the value of exports. The data show significant drops in the export value of oil products (down 29% in weeks 20-23 compared to the same period last year in constant-price terms), iron and steel (down 28%) and non-ferrous metals (down 22%). The category "selected chemical products" has actually increased by 10%. Unfortunately these data don't include crude oil and natural gas.

If these export demand reductions are reflected in reductions in production, then this also suggests big drops in emissions from these energy-intensive industries in Norway. The drop in sales is likely a combination of the lower value of the Norwegian krone and lower industrial activity in the EU, activity that would normally require these intermediate goods in production.

Tourist ship traffic normally starts to pick up in May, but this May estimated emissions from passenger ships in Norwegian waters were down 59% on last year, or about 80 kt CO2, according to data from Note there's little apparent change in emissions from cargo vessels, fishing boats, or the oil and gas industry.

All up?

In many countries, power generation is one of the largest contributors to CO2 emissions, but in Norway this is almost entirely hydropower. This is why I have focused here on other emission sources. But what has electricity generation done in Norway during the pandemic period?

The figure below shows that, apart from lower consumption in January because of a milder winter, there hasn't really been a strong divergence in demand for electricity in Norway. This should be corrected for temperature effects, but it does suggest that we're heating and lighting and running industrial processes that rely on electricity (e.g. aluminium smelting) much as normal.

With no change in emissions from power generation, about 20% drop in combustion of liquid oil products, some reduction in metal production, little change in agriculture, emissions could be down in the order of 10% in April and May compared to last year. Already in June we expect emissions to return close to normal levels again as activity returns to normal in Norway.

See also: SSB has put together an excellent dedicated page of statistics and articles about the effects of the pandemic in Norway.

Why don't we have official emissions estimates already?

The world is full of high-frequency financial data, with enormous sums of money invested in producing these data. Better, faster data means higher profits: there's money in monitoring money. But the same is not true for the environment, and estimates of emissions of greenhouse gases suffer from both infrequency and long delays. In Norway, the first preliminary estimates of 2019's emissions were released by SSB only this week, and these come with little detail, which won't come until November.

Part of the problem is one of collation: while individual sources might have their data in almost real time, those are checked, rechecked, filled into forms or otherwise transmitted to other clearing houses, with a deadline set according to expectations of capacity to deliver, then more checks and rechecks before being sent onwards, collated with other data, and so on, with iteration when errors are suspected. And when we're talking about official estimates, there is a natural aversion to releasing data that might be incorrect. Interestingly, GDP estimates are often revised even in developed countries as additional data come in, and this early release of uncertain preliminary data points to a recognition of the high societal (read: business and political) need for timely economic data.

While high-frequency environmental statistics are of considerable interest to the general public and researchers like me, the benefit they give to policy-makers compared to annual estimates isn't clear, and often that benefit is what drives funding for the considerable effort required to assemble a new batch of statistics.

Looking ahead: Norway's CO2 in 2019

blog post

Early indications of Norway's CO2 emissions in 2019, before the year has finished.Learn more »

Norway's emissions exports

blog post

Norway exports substantial emissions to other countries via oil and gas.Learn more »

Norway's agricultural emissions

blog post

Norway's agricultural emissions have declined, but the reasons are interesting.Learn more »

Norway's GHG emissions

blog post

A quick rundown on where Norway's greenhouse emissions actually come from.Learn more »