Tuesday, December 29, 2020

No, Facebook has not moved IP out of Ireland

There are a number of reports that Facebook has moved intellectual property out of Ireland:

The story is now in wide circulation but it is wrong.  While it is true that Facebook has relocated some IP licenses to the United States these licenses were not in Ireland to begin with; they were in the Cayman Islands.

The Facebook subsidiary at the centre of these stories is Facebook Internation Holdings I Unlimited.  And as this extract from its most recent (and ultimately last) filing with the Company Registration Office states the subsidiary is “domiciled in the Cayman Islands.”

Facebook Ireland Holdings Unlimited Note

And Facebook itself have been clear that the licenses for the international rights to sell advertising on the Facebook platform were located in the Cayman Islands.  One example of this was when spokesperson Delphine Reyre appeared become the TAXE Committee of the European Parliament a few years ago (video here).

Here are some extracts from Reyre’s contributions to the committee hearing.  But note though that these are the words of one of the parliament’s translators rather than the original dialogue (which was in French).  Some precision may have been lost in translation but the central points are clear. 

Anyway here is some of what Facebook spokesperson, Delphine Reyre, said to the committee [preceded by timestamps in the video recording].

[15:56:34]  There was a question on the Cayman Islands. Our model is exactly the same as Google's.  We do have an operation there in the Cayman Islands, and as my colleague said, as our companies develop around the world, he talks about tax creativity but in fact it's tax conservatism to use these methods.  These are the methods that are used in order to be able to work well at an international level as I tried to explain in my introductory remarks. And that is exactly why we have our intellectual property offices in the Cayman Islands.  I also wanted to tell you that the snapshot you are taking of our business model at the moment is just that: it is only a snapshot. It is going to change; it is constantly changing for all the reasons that I set out before.  For all the work that the OECD is doing, for the BEPS group is doing.  And of course this is constantly evolving because of the decisions make in Ireland on business headquarters, tax headquarters.

[15:59:00] Let me just point one thing out.  The fact that out intellectual property office is in the Cayman Islands doesn't have an impact on most European countries but it does have an effect on how we pay taxation in Ireland, that's as far as non-US countries are concerned. I hoped this sheds light on the situation.

[16:01:52] The model we have plumped for is a usual model, a conservative model.  So when we moved abroad, if you like, we were only a four-year old and we just took up the general model that everyone was using. That was to put your intellectual property in a neutral place that allows you to cover the whole world.  And as Nicholas from Google said the difficulty is that the United States taxes intellectual property all over the world so companies therefore are encouraged to therefore find a neutral location.

[16:29:45] As for the issue of the Cayman Islands you are talking of tax dumping once again we don't agree with that description.  As my neighbour has described it is a question of taxes that are deferred for the United States.  Secondly, which don't effect at all all the countries in Europe apart from Ireland.  And as for Ireland there we are in total compliance with Irish law as it stands today.  We are in perfect compliance with Irish law in the future as well once changes have been made.  As regards the countries where we are set up: The Netherlands, there we have an office; in Switzerland, we don't have an office. And in Luxembourg we have a legal entity. And that's all we have by way of headquarters.

[17:03:16] When it comes to operations in the Netherlands, we have a commercial office. That's all.  So, there is no 'dutch sandwich' or whatever there. As regards Luxembourg there is a legal entity and you asked what these legal entity was doing.  I understand that the legal entity has allowed us to get an inter-company loan and support for an inter-company loan.  I don't have the details.

[17:04:55] I think that your question is whether or not there is a 'dutch sandwich'.  There is not. I can reassure you there is not.  It does not go through The Netherlands.

The changing nature of outbound royalties from Ireland

The subsidiaries of US MNCs operating in Ireland have annual revenues of around €400 billion.  These revenues are generated from sales of products and services that are the result of research and development that is primarily undertaken in the United States. 

These include pharmaceutical medicines and computer chips which are manufactured at Irish facilities and online advertising on platforms such as Google and Facebook which have their international headquarters in Ireland.

Paying for technology

It is non-controversial that the Irish operations of these US companies have to pay for the technology they use.  There are essentially two ways in which this can be achieved:

  1. An outright purchase with the Irish company acquiring the rights or license to use the technology;
  2. Recurrent payments with the Irish company getting access to the technology via royalty payments.

Historically, the latter was the main method used.  And the scale of the outbound royalty payments attracts significant attention.  For example, in its latest Country Specific Recommendations (CSRs) addressed to Ireland the European Commission notes:

[T]he high level of royalty and dividend payments as a percentage of GDP suggests that Ireland’s tax rules are used by companies that engage in aggressive tax planning, and the effectiveness of the national measures will have to be assessed.

How big are these outbound payments? Pretty big.

Royalty Imports 2012-2020

In 2019, outbound royalties from Ireland were €84 billion and on their own are enough to offset around 20 per cent of the revenues flowing into Ireland.  This looks like being a high watermark.  Figures for the first three quarters of the year suggest the total for 2020 will be less than €70 billion which is still a very significant sum.

The fall in 2020 is not the result of a reduction in sales leading to lower royalties.  Exports from Ireland in the key categories are set to reach record levels this year. 

In the first three quarters of 2019, computer services exports in Ireland’s Balance of Payments were €84 billion; in the same period of 2020 they were €92 billion.  Exports of pharmaceuticals in the External Trade data were €43 billion in the first ten months of 2019; in the same period of 2020 they were €51 billion.

Of course, it might have been expected that outbound royalty payments would have being falling before now. This is because of the onshoring of intangible assets that has been ongoing for the past number of years.  This has seen some companies switch from #2 above (recurrent royalties) to #1 (outright purchases).

Divergent patterns of royalty payments

It is likely that the chart above with total outbound royalties clouds some divergent patterns.  This can be seen if we look at the contribution of the pharmaceutical and ICT sectors to the total in recent years.

Royalty Imports ICT Pharma 2016-2020

This shows that, in 2016, the amount of outbound royalties for the ICT and pharmaceutical sectors were similar.  After that though the series diverge, with outbound royalties in the ICT sector rising and those for pharmaceuticals falling.  Unfortunately, data suppression by the CSO means that the quarterly series for the ICT sector does not extend beyond Q2 2019 but the general point is clear.  Indeed, the annual figures which are available show that royalty payments from the ICT sector went from €40.7 billion in 2019 to €50.7 billion in 2020 (the first half of 2020 was €22 billion).

Destination matters

Of course, royalties leaving Ireland is only a part of the story.  Where they go to is also relevant.  By far the largest destination of these royalties are what the CSO term ‘offshore financial centres’.  In their notes the CSO set out that:

This category overlaps with the regions referred to above and covers Andorra, Antigua and Barbuda, Anguilla, Netherlands Antilles, Barbados, Bahrain, Bermuda, Bahamas, Belize, Cook Islands, Curacao, Dominica, Grenada, Guernsey, Gibraltar, Hong Kong, Isle of Man, Jersey, Jamaica, St. Kitts and Nevis, St Maarten, Turks and Caicos Islands, Cayman Islands, Lebanon, Saint Lucia, Liechtenstein, Liberia, Marshall Islands, Montserrat, Maldives, Nauru, Niue, Panama, Philippines, Singapore, Saint Vincent and the Grenadines, British Virgin Islands, US Virgin Islands, Vanuatu, Samoa.

Plenty of zero-tax or no-tax jurisdictions in there.  By the end of 2019 around half of the outbound royalties (around €10 billion a quarter) were going to offshore financial centres.

Royalty Imports to OFCs 2010-2020

It can be seen that the amounts were relatively modest up to the end of 2013 after which they increased rapidly.  The reason for this is that up to then Ireland levied a withholding tax on outbound royalty payments to non-treaty countries.  If royalty payments were made by an Irish resident person or entity to these jurisdictions some of the payment would have to be withheld and paid to the Revenue Commissioners to cover potential tax liabilities.

There was a fairly easy workaround for companies to avoid Ireland’s withholding tax on outbound royalties to non-treaty countries.  Under the EU’s Patents and Royalties directive EU member states cannot levy a withholding tax on royalty payments made to a resident of another member state. 

And as is well known The Netherlands does not levy a withholding tax on outbound payments from there to locations such as Bermuda.  This means the chart above does not fully represent the amount of outbound royalties from Ireland that are directed to ‘offshore financial centres’.

In Q4 2019, just over €8 billion of royalties went from Ireland to The Netherlands – and it’s probably safe to assume that via this “dutch sandwich” they were further transferred on to jurisdictions such as Bermuda.

All the companies doing this are US companies.  The US tax system has allowed (and incentivised) US companies to locate licenses for the use of their technologies outside the US in low-tax jurisdictions such as Caribbean Islands. 

So while the Irish subsidiaries were correctly paying for the use of technology generated elsewhere these payments were going to locations where the companies had no substance rather than where the technology was actually developed, i.e. the United States.  The primary tax payments affected by these strategies are US tax payments.

There are many reasons why these strategies will no longer be effective.  Actions 8 to 10 of the OECD’s BEPS I Project means that such payments to jurisdictions which lack the substance to produce the service or technology being paid for may no longer be tax deductible.  The Tax Cuts and Jobs Act abolished the principle of deferral in the US tax code which was a primary motivation for the creation of these structures. 

Changes to Ireland’s residence rules for Corporation Tax meant it was no longer as feasible for US companies to avail of the deferral provisions in the US tax code such as the “same country exemptions” – the basis for the “double irish” being that two companies were registered in the same country.

Google changes tack

One of the most high profile companies with the a “double irish with a dutch sandwich” structure was Google.  But this time last year Google came out and said:

“A date of termination of the Company’s licensing activities has not yet been conīŦrmed by senior leadership, however management expects that this termination will take place as of 31 December 2019 or during 2020.

“Consequently, the Company’s turnover and associated expense base generated from licensing activities will discontinue as of this date” the filing with the Dutch Chamber of Commerce added.

“We’re now simplifying our corporate structure and will license our IP (intellectual property) from the US, not Bermuda,” a spokesman said in a statement.

Can we see the impact of this in Ireland’s Balance of Payments data? It seems we can.

Eurostat don’t have a good quarterly series for royalty payments from Ireland to the Netherlands (many of the values are redacted) but there is a compete series for payments to the euro area and the available numbers suggest that payments to The Netherlands make up 90 per cent of royalty payments from Ireland to the euro area. 

For example, we know that there were €8.1 billion of royalties from Ireland to The Netherlands in Q4 2019; the total for the euro area was €8.9 billion. Anyway, let’s get to it.

Royalty Imports to Euro Area 2008-2020

Pretty clear that there was a significant change at the start of 2020.  This change doesn’t make a huge amount of difference to Google’s operations in Ireland – in the short-term at any rate.  The Irish company must still pay for the right to use Google’s technology but these license payments or royalties are no longer going to Bermuda via The Netherlands. 

Can we see where they are going?  Again it seems we can.  Here are outbound royalty payments from Ireland to the United States.

Royalty Imports to United States 2008-2020

Note that the vertical scale of both charts are the same.  It is not a huge leap to conclude that the royalty payments that went to The Netherlands in Q4 2019 went to the US in Q1 2020. 

So now rather than being parked offshore in Bermuda this income will be taxed under the provisions of the TCJA (FDII, GILTI etc.) though will likely be at a rate that is less than the 21 per cent headline rate of the federal US corporate income tax.

And so does Facebook

And it’s not just Google making these changes.  As we looked at previously Facebook’s outbound royalties from Ireland were paid to the Cayman Islands (with no stopover in The Netherlands).  Facebook have now said that:

“Facebook International Holdings Unlimited is being wound up as part of a change that best aligns with our operating structure. In preparation for the unlimited company winding up, FIH’s assets were distributed to its US parent company.

Intellectual property licenses related to our international operations have been repatriated back to the US. This change, which has been effective since July this year, best aligns corporate structure with where we expect to have most of our activities and people. We believe it is consistent with recent and upcoming tax law changes that policymakers are advocating for around the world.”

The above chart of Eurostat data of royalty payments from Ireland to the US goes to Q2 2020.  The Facebook announcement of a change in its structure from Q3 2020 means that there will be more to add to this when data for subsequent quarters are published.

Manufacturing concern

The European Commission are concerned with the level of outbound royalty payments from Ireland.  Will they still be concerned as a greater and greater share of these payments, particularly for the ICT sector, flow to the United States?  And what about in manufacturing where it is likely that the payments will continued to fall?  Will that reduction satisfy the concerns of the Commission?

When they published the Q1 2020 national accounts the CSO said:

However, royalty imports for pharmaceutical products and preparations decreased by almost €1 billion in Quarter 1, 2020 when compared with Quarter 1, 2019 and the accumulation of the IPP relocations to Ireland in recent years may now be leading to reduced quarterly royalty payments abroad.

It seems in manufacturing, that US companies in Ireland have responded to the changes in tax rules by onshoring their licenses to Ireland with their subsidiaries here undertaking an acquisition of the license.  The capital cost incurred will be deductible for tax with the eligible amounts set out under Ireland’s rules for capital allowances.

The BEPS project’s overarching objective was to try and ensure that company profits were aligned with substance.  US MNCs had little or no substance on Caribbean Islands and maintaining the licenses they had there was no longer feasible.. 

But for US MNCs that have operations in Ireland is there a reason why ICT companies such as Google and Facebook have decided to relocate these licenses to the US while pharmaceutical companies and other manufacturing companies have onshored the licenses to Ireland?

It may come down to that BEPS objective – or, more particularly when it comes to intangible assets, DEMPE. This stands for the development, enhancement, maintenance, protection and exploitation of the intangible asset. 

It seems that US manufacturing companies have decided they have sufficient DEMPE activities in Ireland (or that they could move to Ireland) to justify co-locating the license to use their intellectual property with their production facilities in Ireland.  On the other hand it seems that US ICT companies have decided they do not have, and will not have, sufficient DEMPE activities in Ireland to justify co-locating their licenses here.

Should this be a concern for Ireland? Perhaps.  By onshoring their licenses to Ireland the manufacturing companies are further embedding their presence here.  It is another spoke in wheel solidifying their presence here.  ICT companies such as Google and Facebook have not moved their IP to Ireland.  As a result their international headquarters are a bit more footloose.

Tuesday, November 10, 2020

The Contribution of the Domestic Non-Financial Corporate Sector to Recent Growth and Some Taxing Questions

In an earlier post we went through some of the remarkable improvements for the household sector in recent years.  Of course, an outstanding question is: where did the growth come from?  We can get some insight into this from the same sector accounts. 

Current Account

Here we will look at the Irish-owned non-financial corporate sector which as we will see has been a major contributor to recent growth.  First, the current account. Click to enlarge.

Domestic NFC ISA Current Account 2013-2019

Some of the changes here are very significant.  Between 2013 and 2019, output of domestic NFCs increased by 55 per cent, wage paid by 66 per cent and gross operating surplus by 83 per cent.  Compared to where they were in 2013, last year NFCs produced €53 billion more output, paid employees €18 billion more and generated €12.5 billion more profit. 

These have been a significant factor underlying the economy’s recent growth.  A useful sectoral breakdown is also provided.  Here are the domestic sources of compensation of employees (with domestic financial corporates (NACE K) also included))

Domestic Corporate COE Paid 2013-2019

The largest relative increase was the 130 per cent rise in employee compensation from domestic ICT companies (NACE J). The largest absolute increase was the €2.6 billion increase in employee compensation from professional, scientific and technical activities companies (NACE M), with mining and manufacturing (B,C), wholesale and retail (G), and administrative and support services (N) as well as ICT also recording increased greater than €2 billion.

Some of the other notable changes in the current account of the domestic NFC sector are the near 1,000 per cent increase the retained earnings of foreign subsidiaries and related entities of Irish-owned NFCs, the 500 per cent increases in dividends received and the 13 per cent increase in Corporation Tax paid. 

Two per cent tax rates?

It is surprising to see such a small increase in Corporation Tax paid by domestic NFCs over a period when all performance indictors improved significant.  The figures are this item are a significant revision on what the CSO reported this time last year. For example, it was previously reported that domestic NFCs paid €1.6 billion of Corporation Tax in 2018.  This is now estimated to be just €0.5 billion. 

It is not clear why there has been such a significant downward revision.  Nor is it clear that the figure is correct.  If we take the figures for depreciation from the capital account we can get the Net Operating Surplus of the sector.  This is probably the closest proxy of the tax base for Corporation Tax in the sector accounts and is akin to EBIT (the FISIM adjustment excepted).  Anyway, using Net Operating Surplus we can get some effective tax rates.

Domestic NFC Tax Paid 2013-2019

Those certainly are some eye-catching numbers down at the bottom.  Perhaps there is some support for this in the Revenue Commissioners research paper on recent Corporation Tax receipts which shows that non-multinationals (i.e., domestic companies) have the lowest effective tax rates in their data.

However, that conclusion was based on an effective tax rate in 2018 or 7.4 per cent which is almost three times larger than what the national accounts are showing.  Differences in definitions and coverage aside that is a large gap to be explained. No explanation is offered here.

Capital Account

We dipped into the capital account with the depreciation figures used above and here it is in full.

Domestic NFC ISA Capital Account 2013-2019

Again, we are looking at some large increases over the period.  In line with the rise in profits (and seeming absence of a rise in taxes) the annual increase in net worth rose significantly over the period.

Of this, ever more was used for gross capital formation, with the investment spending of domestic NFCs rising from €7 billion in 2013 to €19 billion in 2019.  There was no year where this fully exhausted the gross saving available and for each of the past seven years the domestic NFC sector has been a net lender. 

That is, after working through the current and capital accounts there was a residual surplus that was transferred to the financial balance sheet.  And it could be that some this is going on the financial balance of foreign subsidiaries or related companies given the increase in the retained earnings on FDI shown in the current account.

Financial Balance Sheet 

We don’t have a financial transactions account so move straight to the financial balance sheet.  As we said last year this has been exploding.

Domestic NFC ISA Financial Balance Sheet 2012-2019

The financial assets of the Irish-owned NFC sector have gone from €167 billion in 2012 to €513 billion last year.  On the other side of the balance sheet, financial liabilities have gone from €175 billion to €633 billion.   Even with the huge increase in the gross figures, the net financial worth of the sector has been largely unchanged, going from –€108 billion in 2012 to –€120 billion last year.

Such is the scale of the increases across the available categories it’s actually pretty hard to determine what is going on.  But for both financial assets and liabilities there has been extraordinary growth in the equity category.  Are there some form of circular transactions taking place?

While the balance-sheet amounts are massive we didn’t really see significant distortions in the current account.  Yes, there has been a surge in the retained earnings of foreign related entities.  This was €3 billion in 2019 and it is possible that some of this should be included in operating surplus.

It would be nice to see the interest figures before the FISIM adjustment but they are only published by Eurostat who do not use the bespoke sectoral breakdown employed by the CSO.


All-in-all it looks like there are two things going on in the domestic NFC sector.  The first is a large, and very real, contribution to the growth of the Irish economy in recent years with significant increases in output, wages and profits.  This has driven a lot of the improvements that occurred pre-COVID.

The second is some kind of financial engineering on the balance sheets of (some) domestic NFCs.  It is possible this involves some sort of circular international transactions with huge increases on both sides of the balance sheet but limited impact on net worth.  To the extent, that there are flows which might, for example, reduce the tax base in Ireland, is unclear.  And, indeed, the tax paid on the tax base of this sector that remains seems unusually low with an estimated effective rate of not much more than two per cent.

So a mix of the good and the unexplained.

Friday, November 6, 2020

The convergence of urban-rural incomes in the SILC

In recent years, the Survey of Income and Living Conditions has shown remarkable improvements for Irish households with significant rises in income and significant falls in inequality.  One feature within the findings has been the narrowing of the urban-rural divide.

To start, just look at real income.

SILC CSO Median Real Equivalised Disposable Income Urban-Rural 2004-2018

In 2013, median real equivalised disposable income in rural areas was just 85 per cent of that in urban areas.  For the past couple of years it has been 95 per cent or higher.

For the purposes of the SILC, households are classed as living in an urban or rural area based on population density (as measured by the 2016 Census).  Areas with a population density of less than 1,000 per square kilometre are classified as rural.  In 2019, 30.9 per cent of the weighted sample were in rural areas.

With incomes by area we can also look at at-risk-of-poverty rates by area.  This is the percentage of people living in households with an equivalised income below 60 per cent of the national median.  For 2019, the threshold used by the CSO was an income after taxes and transfers of €33,400 for a household of 2 adults and 2 children.

In previous years, as the rural category had a lower income, it had a higher at-risk-of-poverty rate. This reversed in 2019.

SILC CSO At-Risk-Of-Poverty Rate Urban-Rural 2004-2018

We don’t have inequality figures by status but this could be taken to imply that there is less inequality in rural areas at the bottom of the income distribution.

The at-risk-of-poverty rate is a relative income measure.  To help with understanding differences in living conditions we can look at deprivation rates.  This is the percentage of individuals living in households who experience two or more items of deprivation from a list of 11.  According to the CSO for the population as a whole:

In 2019, the most common item of deprivation experienced was the inability to afford to replace any worn out furniture (18.1%), followed by being unable to afford to have family or friends for a drink or meal once a month (13.6%) and being unable to afford a morning, afternoon or evening out in the last fortnight (11.7%).

Here are the rates for urban and rural areas.

SILC CSO Deprivation Rate Urban-Rural 2004-2018

These have always been closer than the income levels or at-risk-of-poverty rates might imply.  This is likely because those in rural areas face a lower price set compared to those in urban areas, most notably for housing.

Over the past few years, rural area have had a lower deprivation rate than urban areas and though the fall that had been evident in both since 2013 stopped in 2019, the gap between them increased.  There was no significant change in the deprivation rate for rural areas in 2019 while it rose in urban areas.

Finally, we can look at the consistent poverty rate which is a measure that combines those at-risk-of-poverty and also experiencing deprivation.

SILC CSO Consistent Poverty Rate Urban-Rural 2004-2018

Given what this shows it wouldn’t be out of place to talk about urban-rural divergence in the SILC rather than convergence.  In 2019, the consistent poverty rate for rural areas continued to fall and was less than half what it was in urban areas (3.1 per cent versus 6.5 per cent). 

The “rural Ireland being left behind” narrative is a common refrain.  There is no evidence of it here.  

Wednesday, November 4, 2020

The position of the household sector as we entered the crisis

Last week, the CSO published the 2019 Financial and Non-Financial Institutional Sector Accounts.  Obviously, with the emergence of COVID-19 this is a world away from where we find ourselves now but it is instructive to see where we were with the onset of the crisis.  We previously looked at some of the COVID-generated chaos in the household sector with the Q2 2020 sector accounts here.

The non-financial accounts of last week’s publication give the incomes and expenditures of each sector and then how they translate through the positions for items like loans and deposits is shown in the financial accounts. 

Here we will look at the household sector.  First, an extended version of the current account for the household sector. Click to enlarge.

Household Sector Current Account 2015-2019 Full

OK, lots more detail there then we need to go through.  The headline results are that Gross Disposable Income rose 7.3 per cent in 2019 to reach €114.2 billion.  Household Final Consumption Expenditure rose 5.6 per cent and was €102.0 billion.  Income rising faster than consumption meant savings rose and the Savings Rate was 12.2 per cent in 2019.

The most significant change for the household sector in recent years has been the rise in employee compensation received.  In 2015, this was €77 billion.  By 2019 it exceeded €100 billion.  The was driven by an expansion of employment and an increase in earnings.

Household Sector COE Paid 2003-2019

By far, the largest source of the increase were non-financial corporations, and COE from this sector was €15 billion higher than it was at the previous local maximum in 2007 and 2008.  For more recent years, we have a breakdown of this between domestic and foreign NFCs, noting that foreign NFCs include traditional FDI as well as foreign-owned companies with a presence to serve the domestic market.

Household Sector COE Domestic and Foreign NFCs 2013-2019

In recent years, COE growth from domestic firms has outstripped that from foreign firms.  In 2015, were responsible for 35.5 per cent of the COE that arose in NFCs, €16.8 billion out of €47.4 billion.  By 2019, the foreign share was down to 32.6 per cent, €21.4 billion out of €65.7 billion.

Further up the table it can be seen that that the mixed income of the self-employed was much less buoyant than employee compensation. Self employed income, which includes agriculture, went from €11.6 billion in 2015 to €12.3 billion in 2019. 

There was a substantial increase in the Gross Operating Surplus of the household sector from €12.2 billion in 2015 to €18.9 billion in 2019.  The GOS of the household sector is derived from the provision of housing services.  Most of this is income from “imputed rents” said to be paid by owner-occupiers to themselves (this nets out further down the table via “imputed rent” expenditure in consumption). Household GOS also includes the surplus generated from actual rents for housing earned by households who own buy-to-let properties.

Household Sector GOS and MIxed Income 2003-2019

In the allocation of primary income account there isn’t a whole lot going on.  We will consider interest flows – though before the adjustment is made for Financial Intermediary Services Indirectly Measured (FISIM). No, let’s not go there. 

Here are the gross interest flows for the household sector from Eurostat.

Household Sector Interest 2003-2019

Interest received is down to less than €300 million while interest paid has been €4 billion or slightly under for the past six years or so.  As stated, most else in this part of the current account has been relatively stable.  There has been a bit of a bounce in dividends received by the household sector which exceeded €2 billion for the first time in 2019.

The next part of the table works through income taxes and social contributions paid, social benefits received and other transfers paid and received.  The net outcome here gives us the difference between Gross National and Gross Disposable Income. 

The gap between the blue and red lines below can be considered part of the contribution of the household sector (including employers’ social insurance contributions) to the non-cash-transfer cost of government.

Household Sector Income and Consumption 2004-2019

We can see that back in 2010 there was close to no difference between Gross National and Gross Disposable Income for the household sector.  This means the amount the government collected from income tax and social contributions from household income was almost fully matched by the amount the government paid in social benefits to households. 

Household Sector Net Tax Social Position with Government 1995-2019

Since 2010, the gap has opened significantly and exceeded €15 billion in 2019 as the amount of tax and social contributions collected (Income Tax, USC & PRSI) has risen but the amount of social contributions paid has been relatively stable (n.b. this only goes up to 2019!).

Going back to the previous chart, the gap between the red and green line gives the amount of disposable income that is not used for consumption: household savings.

Household Sector Savings Rate Revised 1995-2019

In 2019, the household savings rate is estimated to have been just over 12 per cent.  There have only been two years since 1995 when it was higher: the crisis years of 2009 and 2010. 

The income of the household sector had been growing rapidly in advance of the COVID-19 crisis but the gains were not being used for current consumption.  What was the household sector doing with the income it wasn’t spending?  The possibilities are that the money was used for capital spending or put on the financial balance sheet.  To assess that we can look at the capital account.

Household Sector Capital Account 2015-2019

The bottom line here is the household sector has not been using the additional savings from the current account for capital spending.  There are a couple of other minor capital flows to account for but the main take from the capital accounts is [S – I], savings minus investment.

In 2015, households had €8.0 billion of gross saving and did €4.4 billion of capital spending leaving €3.6 billion for net lending,  By 2019, household gross saving was nearly €14 billion and while household capital spending had increased it was still less than €7 billion.  This meant in increase in net lending to €7.4 billion.

Near the bottom of the above presentation of the capital account we can see net capital formation: the difference between gross capital formation and consumption of fixed capital (depreciation).  This shows that there have been very modest net additions to the household sector capital stock. 

Household Sector Gross and Net Capital Formation 2003-2019

Indeed, as recent as 2015, household gross investment was not sufficient to cover depreciation.  In 2019, the net increase in the household capital stock was just €1.1 billion which isn’t much more than the amount investment grants received by the household sector for the likes of retro-fit projects.  It’s probably not a surprise that household capital formation remains muted given the modest increases in the supply of new housing, which is the main capital formation the household sector engages in.

But let’s look at what the household sector did with the €7.4 billion that was left after all current and capital spending has been accounted for.  So we switch from the non-financial to the financial accounts and start with the Financial Transactions Account.

Household Sector Financial Transaction Accoun 2015-2019

The bottom line here, net financial transactions, is slightly larger than the net lending we saw at the end of the capital account, but not significantly so.  What are the household sector doing with the resources they have left after their spending has changed in recent years. 

Household Sector Contribution to Net Wealth

Back in 2012, the increase in net household wealth due to financial transactions was mainly happening via loan repayments with only a modest increase due to adding to deposits.  Over the following years this reversed and in 2019 most of the contribution to net wealth due to financial deposits was from increases to deposits.

Of course, this may not simply be due to changes in what the household sector is doing with its net lending surplus.  For example, if one household takes out a mortgage to buy a house from another household who put that money on deposit that will reduce the contribution of loans to net wealth and increase the contribution of deposits in net wealth but in overall terms the impact of this on the net wealth of the household sector is zero.

Anyway, let’s look at the impact on the financial balance sheet of the household sector.

Household Sector Financial Balance Sheet 2015-2019

Here we can see that the change from 2015 to 2019 in the values on the balance sheet roughly corresponds to the sum of transactions over the same period as shown in the previous table.  The only significant exception to this is insurance and pensions reserves which, as well as increasing due to household contributions, have also increase due to revaluation effects, i.e. rising asset values.

We can see that the above reflects the rising deposits and falling loans from the financial transactions account.  The stock of loan liabilities of households fell from €146.9 billion at the end of 2015 to €130.9 billion at the end of 2019.  On the asset side of the balance sheet, the amount of current and deposits rose from €129.0 billion to €147.2 billion.

Here is a chart with a longer series highlighting that the reversal of the relative size of loans and deposits takes us back to a position last seen in 2002.

Household Sector Loans and Deposits 2001-2019 CSO

Using these stocks of loans and deposits and the amount of interest paid we can calculated implied interest rates.  The average interest on loans was around three per cent.  The average interest rate on deposits is essentially zero.  After narrowing post 2008, the interest differential or margin between loans and deposits is back to where it was in the early 2000s (around three percentage points).

Household Sector Implied Interest Rates 2001-2019

We can get a better relative position if we put the stocks of loans and deposits as a share of income rather than just looking at the nominal levels.  This shows that the 2019 loan/deposit position of the household sector, in aggregate terms at least, was pretty much back to where it was in 2001/02.

Household Sector Loans and Deposits to GDI 2001-2019 CSO

The current account showed us that the income of the household sector has recovered all the losses suffered in the aftermath of the 2008 crash.  And now we can see that the worst of the excesses of the credit bubble have been purged from the household sector financial balance sheet. The COVID crisis is wrecking havoc.  The consequences of that would be far worse if the pre-crisis position of household sector was not as strong as it was.