Growth of real wages in the EU15 since the start of 2014

Eurostat published Q3 2017 nominal hourly labour costs data earlier in the week.  Adjusting for inflation, using the Harmonised Index of Consumer Prices, we see the following picture emerge for the wages and salaries component of real labour costs since the start of 2014:

Using this approach it can be seen that Ireland has had the second-highest increase in real hourly wages since the start of 2014.  This might look good in relative terms but in absolute terms the increases remain modest.

Here is what emerges if we look at annual growth rates (smoothed over four quarters).

There are maybe two ways to interpret this.  The first, is that Ireland has the highest growth in real hourly wages in the EU15.  This growth has accelerated slightly in recent times but not massively so.  The second is that the growth of real wages has slowed in almost all EU15 countries with 11 of the 15 having a slower growth rate of real wages compared to a year ago.

The reason for this has not been a slowdown in nominal wage growth, if anything this has slightly picked up with 11 countries showing a higher nominal wage growth in Q3 2017 compared to a year earlier.  The arithmetic mean for the EU15 has gone from 1.1 per cent to 1.5 per cent.  Ireland too has seen an increase in nominal wage growth but not unusually so.  The smoothed average growth has gone from 1.2 per cent in Q3 2016 to 2.1 per cent in Q3 2017.

The reason for the slowdown in real wage growth across the EU15 has been the uptick in inflation.  The annual inflation rates from the HICP applied to the above nominal wages changes went from a mean of 0.3 per cent in Q3 2016 to a mean of 1.4 per cent in Q3 2016 with the UK, particularly showing a rapid rise in inflation. 

It’s all to with the second derivatives.  Inflation is rising faster than the increases in the growth of nominal wages hence real wage growth is slowing down – for most countries.

Ireland is one of the exceptions.  The increase in nominal wage growth over the past year in Ireland (0.9pp) is slightly above the simple average increase seen across the EU15 (0.4pp).  Inflation, though, has been non-existent in Ireland for an extended period now and remained close to zero throughout the period shown above.  Over the past year Ireland has had the smallest increase in inflation in EU15 and now has the lowest rate of inflation across the group – and this is including the large contribution made by the rapid increases in private rents. 

So put together our modest, though increasing, growth rates in nominal wages with our low rates of inflation and we get the result that Ireland has the fastest growth of real hourly wages in the EU15. For now anyway.

Why the breathless reactions to that Facebook announcement may be excessive

The announcement by Facebook that it was “moving to a local selling model” has been met with some breathless reaction with some thinking it will lead to sleepless nights.  There are a couple of reasons why such a reaction to this announcement may be excessive or why having sales recorded in the location of the customer isn’t the tax-changing panacea that some have made it out to be.  Anyway, here’s the key part of the Facebook announcement:

Today we are announcing that Facebook has decided to move to a local selling structure in countries where we have an office to support sales to local advertisers. In simple terms, this means that advertising revenue supported by our local teams will no longer be recorded by our international headquarters in Dublin, but will instead be recorded by our local company in that country.

There are a couple of reasons for a more reflective reaction.  First, Google announced that it was moving to such a model back in January 2016 while Facebook itself has been booking UK sales with its UK subsidiary since April 2016. Neither have caused the sky to fall in.

When HMRC concluded its six-year audit of Google in the UK in January 2016 Matt Brittin, Head of Google in Europe, gave an interview to the BBC:

“Today we announced that we are going to be paying more tax in the UK.”

"The rules are changing internationally and the UK government is taking the lead in applying those rules so we'll be changing what we are doing here. We want to ensure that we pay the right amount of tax."

The firm has now agreed to change its accounting system so that a higher proportion of sales activity is registered in Britain rather than Ireland.

"We are paying £130m in respect of previous years when the rules were to pay in respect of profits you make in a country and then going forward we will also be paying in respect of sales to UK customers," Mr Brittin said.

Asked whether the back-payments showed Google's critics were right that the company had avoiding paying tax in the past, Mr Brittin replied: "No."

He continued: "We were applying the rules as they were and that was then and now we are going to be applying the new rules, which means we will be paying more tax.

"I think there was concern that international companies were paying only in respect of profits that they make and those were the rules and the pressure was to see us pay in respect of the sales we make to UK customers - and the same for other companies.

"So, we are making a change because we want to continue to comply with the rules and the rules are changing."

I think the last word of the second last paragraph should be “countries” or be extended to “Google companies in other countries”.  So there is no precedent with the Facebook announcement.  Google got there two years before them.

And Facebook have been booking sales in the UK for over a year now as it states in the recently-filed 2016 accounts for Facebook UK Ltd.  In those accounts we are told:

The principal activity of the company in the year under review was that of providing sales support, marketing services and engineering support to the Facebook group and to act as a reseller of advertising services to larger UK customers.  The company’s function expanded to include the advertising reseller business in respect of large UK customers on 1 April 2016. 

Revenue for the year amounted to £842,429,955 (2015: £210,762,610), which is an increase of £631,667,345 on the value of services provided in 2015.  This increase was attributed to the commencement of advertising reseller services from 1 April 2016. 

The impact of this can be seen in the income statement for Facebook UK Ltd.



The first line shows the £632 million increase in turnover but this is immediately followed by a £450 million increase in the cost of sales.  After a £75 million increase in administrative expenses (mainly staff costs), all told, in 2016 Facebook UK Ltd. ended up with a profit of £58.4 million and a resultant tax bill of £2.6 million.  [Insert meaningless calculations of tax as a proportion of revenue around here.]*

The jump in the cost of sales figure is linked to the company becoming a reseller in April 2016.  In 2015, the company didn’t have any third-party sales.  All of its revenue was derived from providing marketing and support services to Facebook Ireland Ltd.  In 2016, it recorded the revenues from sales to the UK customers it dealt with directly but like with any company that doesn’t make the produce or service it sells it had to buy them from a supplier. 

There have been plenty of calls that revenues should be recorded where the customer is located.  And it could be that Facebook is reacting to the possibility that the implementation and/or interpretation of the BEPS proposals could mean that Facebook Ireland Ltd would be deemed to have a “permanent establishment” through the local company and that any sales done through that PE would be recorded in that country anyway.  By acting now to record the sales in the local companies Facebook is controlling the changes – albeit incurring significant cost to do so.

The recording of sales in the local companies is fine but it must be remembered that the corporate income tax is paid on profit not revenue.  Revenue only translates into profit to the extent that a company has risks, functions or assets that add value.  And if a company doesn’t even make the product it sells it must buy it.  In fact, because of this it may be that the aggregate revenue-profit-tax outcome that arises through all these local companies will not be significantly different than what is achieved through the central company in Ireland now.  There may be differences in where tax is paid but it will be interesting to see if it results in more tax being paid overall. 

Anyway, back to the need for companies who don’t make what they sell to buy it.  This can be from a related or third-party supplier but regardless of the relationship between them the price should be the same, i.e. as based on the arm’s length principle.  So if Facebook UK is selling advertising on a platform that is owned by someone else it must pay the owner of the platform (or its licensee) for the right to sell that advertising or buy the advertising space before selling it on.  The latter appears to be what Facebook is doing in the UK and this is the cost of sales figure that offsets most of the revenue increase that resulted from recording the sales locally in the first place.

Who is Facebook UK buying the advertising space off?  We don’t know who this is but a good guess is probably the company that sold the advertising space to these customers the previous year.  And we all should know who that was.  So while the third-party sales may be recorded with the local company there will still be intra-company sales recorded in Ireland.  The final paragraph of the Facebook statement is not without its significance:

Our headquarters in Menlo Park, California, will continue to be our US headquarters and our offices in Dublin will continue to be the site of our international headquarters.

The revenues will still end up in Dublin it’s just that some third-party sales will be initially recorded with local companies.  Third-party sales to customers that do not have a local company as well as sales in those countries that do from customers who do not interact with the local companies will continue to be recorded in Ireland. 

As the only customers in question are those who interacted with their local company rather than Dublin anyway there is unlikely to significant change to the activities carried out by the staff in Dublin.  It is not as if there are staff in Dublin who will lose key customers or accounts because of this change.  They were been serviced by the local companies as it stands.  One of the important, and valid, criticisms of the old structure was that close on the only thing that happened in Dublin relating to these customers was the conclusion of the sale – the virtual signing of the contract.  The new structure better reflects Facebook’s selling activities.

It is also not clear how this change the amount of Corporation Tax that Facebook will pay in Ireland.  It is likely that Facebook Ireland Ltd is currently remunerated on some form of “cost-plus” basis with the main cost included in the calculation being the staff costs.  If there is no significant change in the activities of Facebook in Ireland, its cost here could remain similar and so too could its Corporation Tax liability here.

There may, however, be changes to the risks, functions and assets that Facebook has in Dublin but they are not detailed in the short statement issued by the company.  It could be that headlines like “Facebook will stop using Ireland as a global hub for tax and revenue” are wide of the mark.

As we have seen this is because it is likely that the advertising services sold in the local markets will still originate with the Irish company and, crucially, we don’t know what Facebook is going to do with its licenses and intellectual property.

As we looked at before (here and here) the transfer of these licenses out of the US is subject to an investigation by the IRS.  At issue there is the price paid for those licenses but the existence of them to grant the rights to sell advertising on the Facebook platform outside the US is likely to continue.  We don’t know where those licenses are currently held but some indications point to the Cayman Islands.

Unless Facebook are going to move significant DEMPE functions (developing, enhancing, maintaining, protecting or exploiting intangible assets) then payment to a cash-box in the Caymans may not be allowed as a deduction for tax purposes when the full implementation of the transfer pricing changes in the BEPS proposals comes into effect.

Facebook will likely seek to move its IP to a location where these DEMPE functions are located.  An obvious choice would be the US where the main innovation and R&D of the company takes place.  But even the proposed changes currently hurtling through Congress are unlikely to make that sufficiently attractive for companies who have managed to get part of their IP outside of the US.

Another alternative is Ireland.  We have already seen significant onshoring of IP and it is possible we will see more.  It would seem natural to co-locate the license to your international sales with your international headquarters.  Hence, the possible significance of that final paragraph in the company’s statement.

The thing is we just don’t know.  Maybe the advertising sold by the local companies won’t originate in Ireland.  Maybe the IP won’t be onshored here.  But until we know that let’s keep a lid on the breathless reactions.

There are risks to Ireland’s FDI model but I’m not sure ICT companies moving to a local selling model for some customers is one to knock us over.  There are tens of billions worth of goods and services sold from Ireland and most is via intra-company sales rather than sales to third-party customers.  We had €14.7 billion of exports to Belgium last year of which €13.5 billion were pharmaceuticals.  I don’t think they were all taken by Belgians. 

There could be benefits for Ireland.  As the sales are being recorded in these countries they will get first dibs at taxing the resulting profits.  If they think they’re not collecting the “right” tax from these sales (whatever that may be) the first bout of finger-pointing should be domestic. 

Of course, if the products or services originate in Ireland and/or the companies have their IP in Ireland then we will be next in line.   And maybe like India is attempting with Google countries may try to describe the payments to Ireland as royalties rather than sales revenues possibly bringing withholding taxes into play (for some non-EU countries at least). Those sleepless nights may arrive yet.

* It’s 0.3% by the way.

Getting somewhere with the Current Account

As part of its response to the 2008 crisis the EU has set up the Macroeconomic Imbalance Procedure which “aims to identify, prevent and address the emergence of potentially harmful macroeconomic imbalances”.  The identification is done via a scoreboard of indicators with various thresholds. 

Our concern here isn’t on the choice of indicators, the thresholds chosen or the one-sided nature of the adjustments but on the work done by the CSO to make some of the indicators logical from an Irish perspective.  We know that lots of macro aggregates in Ireland are distorted and unless this is identified a scorecard with arbitrary thresholds could lead misleading conclusions and adjustment requirements.

The CSO first published its version of the Macroeconomic Scorecard for 2013.  As we considered at the time this provided a very useful breakdown of private sector debt with debts of the NFC sector broken down by Irish-owned and foreign-owned parents.  There were lots of wild claims that debt in Ireland was four, five or more times national income.  The CSO’s work via the macroeconomic scoreboard was important in helping to get the actual position set out.

The work has continued with scorecards produced for 2014 and 2015 with the latest one for 2016 published last week.  The graphics have got flashier and the insights into things like the current account and net international investment position have got better.

The work on the Balance of Payments current account is particularly useful given the importance of this measure.  A distorted current account offers few insights.  Here is Ireland’s headline current account from the Balance of Payments.

Over the past few years the breakdowns provided by the CSO focused on the impact of redomiciled PLCs and the changing treatment of aircraft for leasing.  However, the extraordinary recent jumps in the balance are due to the impact of intangible assets and it was clear that this would have to be addressed.

This started with the publication of the modified national income measures back in June which included a modified current account, CA*.  As the CSO explained:

CA* is the current account balance (CA) adjusted for the depreciation of capital assets sometimes held outside Ireland owned by Irish resident foreign-owned firms, e.g. Intellectual Property (IP) and Aircraft Leasing, alongside the repatriated global income of companies that moved their headquarters to Ireland (e.g. redomiciled firms or corporate inversions).

The size of these adjustments was shown in this table (click to enlarge).

So if we subtract the depreciation of foreign-owned IP and aircraft for leasing as well as the net income of redomiciled PLCs the outcome is:

The 2016 figure shows that all is not well with this approach and as we discussed here the issue was likely related to the acquisition of aircraft and intangible assets.  These were being bought by Irish-resident entities but being funded by intra-company loans so any deficits arising from these acquisitions are of little concern to the rest of us. 

The CSO have included an updated approach to the modified current account in their 2016 macroeconomic scoreboard that takes this into account.  As they say in a revised version of their note on the modified current account:

Since the original publication the CSO has made a further change to CA* to exclude the cost of investment in aircraft related to Leasing and the cost of R&D related IP from the current account balance. Some firms borrow money abroad to finance their investment by purchasing IP from their parent company. In the long term this debt is repaid from the profit on the IP or the aircraft being leased. It means that this borrowing is not a liability of residents of Ireland and the purchase of this IP needs to be excluded when deriving CA*.

A couple of extra columns have been added to the table showing the adjustments (again click to enlarge).

As before the adjustments for depreciation and redomiciled income are subtracted while now  adjustments for imports are added back in to give the updated version of the modified current account.  The headline and modified current account balances are:

This is much better and there is no doubt that this modified current account gives a much more informed view of the underlying position of the economy relative to the headline current account.  This is further evidence of the work being undertaken by the CSO to provide meaningful indicators of the underlying conditions of the Irish economy.

As discussed here there may still be some concerns that the figures for recent years are a little high.  The modified balance is a surplus of €13 billion for 2016.  This may be related to the treatment of expenditure on R&D services as intermediate consumption for Balance of Payments purposes and gross fixed capital formation in the National Accounts. 

If this is an issue it may be remedied in due course and it does not require any changes to the adjustments now proposed to get the modified current account, CA*, as any revisions to the headline balance will automatically apply to the modified balance.  It may have taken a while but there’s no doubt that we’re now getting somewhere with the modified current account and as with other indicators of the Irish economy this is showing that we’re in pretty good shape.

The Aggregate Corporation Tax Computation for Companies with Net Income of More than €10 million

The last post looked at the corporation tax computation for companies with no net trading income.  Here we look at the other end of the scale and consider the aggregate outcome for companies with net trading income of more than €10 million.  Companies in this category make up about 500 of the 145,000 or so companies who filed tax returns for accounting periods ending in 2015.  It should also be noted that the composition of companies in the subgroup may differ across years.

Again, we will show a comparison between the figures for all companies and those in the chosen subgroup.    

Unsurprisingly these companies with the largest net trading income are the source of most on the net trading income in the economy.  Around 88 per cent of net trading income in 2015 (€72.4 billion out of €82.7 billion) arose in companies with a net trading income of more than €10 million.  

Perhaps surprisingly these companies only generated 56 per cent of the starting point: gross trading profits.  These reflects large use of capital allowances and previous losses by other companies especially, as we saw, companies with no net trading income.  The companies here has 16 per cent of the available capital allowances and, for 2015 at least, has less than two per cent of the available prior losses carried forward.

Of other income, these companies reported less than ten per cent of the total amount of rental income and foreign income but did have more than 40 per cent of the capital gains included in the Corporation Tax returns.

Unfortunately we are not given a breakdown of trade charges and group relief by range of net income but it is clear that the vast majority of these arise in these companies.  Before deductions and charges these companies had a Total Income of €74.2 billion in 2015 which after those items translated in a Taxable Income of €45.9 billion. 

Trade charges primarily refers to certain royalty payments.  We would usually expect such expenditure to be included as a deduction when Gross Trading Profits are being derived but Irish legislation set out that certain payments may not be deductible but rather should be deducted as a ‘charge on income’.   Hence we get Trade Charges between Total Income and Taxable Income.

Anyway, our companies with net trading income of more than €10 million had €45.9 billion of Taxable Income in 2015 and 0n this €5.1 billion of Corporation Tax was due giving tax due as a proportion of taxable income of 11.2 per cent in 2015.

The main reason for this being below the heading 12.5 per cent is the use of the R&D credit.  We don’t get a full breakdown of this but we can see that about 85 per cent of the R&D credit used against tax in the current year goes to these companies.  A split of the €359 million for the payment of the excess R&D credit would be nice but it is not provided.

To conclude, perhaps surprising is the amount of Double Tax Relief granted to these companies.  In 2015, they reported Foreign Income of €343 million but were granted €155 million of Double Tax Relief – €155 million is 31 per cent of €497 million (being the sum of the (after-tax) foreign income and the double tax relief).

The Aggregate Corporation Tax Computation for Companies with No Net Income

Among the Corporation Tax Statistics provided by the Revenue Commissioners is a breakdown of a number of the items that make up the aggregate corporation tax computation by range of net income.  Here we will look at the recent outturns for companies reporting “nil or negative” net trading income.  With no net trading income it might suggest that there is little going on from a tax perspective but that is far from the case.

The table below provides the aggregate computations for “All Companies” and for “Companies with no Net Income”.  Some items are not provided in breakdown used but more than enough is provided to see what is going on.  It should also be noted that this breakdown is done by individual entity so while a company or group may have an entity with no net trading income this does not mean that the group as a whole does not have net trading or other taxable income with tax liabilities associated with that.  Anyway here they are: 

In the panel on the right we can see the line of duck eggs in the row for Net Trading Income but there is a lot going on both above and below that.

At the top we can see that companies with no Net Trading Income had Gross Trading Profits of just over €40 billion, a rise of €26.5 billion.  Net income is derived by subtracting capital allowances and previous trading losses carried forward.  The dataset gives the amount of these that are available rather than used but the effect of them is to reduce the starting figure for gross trading profits to a net figure of zero. 

Most of the losses likely relate to financial institutions and a large portion of them will likely never be used as some of the companies are in liquidation.  The use of previous losses is only possible if there are current gross trading profits against which to offset them.

The changes for capital allowances are more interesting.  In 2015, the amount of capital allowances available for all companies increased by €27.8 billion.  We can see that of that €25.6 billion was related to companies with no Net Trading Income. 

Thus we have an increase of around €26 billion in the gross trading profits offset by an increase of around €26 billion in capital allowances resulting in a net trading income of nil. 

Moving down the computation we can see that while these companies might have no net trading income they do have significant amount of other income and in particular foreign income.  In 2015, the Total Income of companies with no net trading income was over €10 billion with almost €8 billion of this being Foreign Income.  The remainder is made up of capital gains (regrossed to reflect the difference between the rate of Corporation Tax and Capital Gains Tax) and net rental income.  Since 2011, around 70 per cent of the income in these categories has arisen in companies with no net trading income.  In most years,  these companies had limited deductions to use against this and in 2015 the €10 billion of other income translated into a Taxable Income of €9.3 billion.

We don’t get a breakdown of gross tax due by range of net trading income but applying the 12.5 per cent rate would give a rough approximation of around €1.1 billion.  We can see from the bottom line that the amount of Tax Due was €177 million and that almost all of that reduction is the result of €730 million of Double Taxation Relief granted to these companies in 2015.  The €177 million of tax due likely arises due to rental, capital gains and other income earned by companies in this group.

The foreign income included in the table relates to external activities of Irish-resident companies.  This is in the Corporation Tax computation as Ireland has a worldwide regime and remitted branch or related company dividends are included in the Taxable Income base.  To avoid double taxation a relief is granted in the form of a foreign tax credit.  This credit is one of the main reasons the tax due as a proportion of taxable income is below 12.5 per cent (as shown it was 9.6 per cent in 2015)

The data by range of net income shows that almost 80 per cent of the €948 million of Double Taxation Relief granted in 2015 was to companies with no net trading income.  In overall terms there are ten of thousands of companies in this category but Double Taxation Relief was granted to just 413.  In 2015, there was also €213 million granted under the Additional Foreign Tax Credit.  This is included under “Other Tax Relief” in the table above but is only available for all companies as a breakdown by range of net income is not provided.

It is likely we are dealing with a small group of Irish-resident companies.  These companies may have no or limited domestic trading activities though other companies in the group may have significant operations, and tax liabilities, here.  These companies have large amounts of foreign income included in their Taxable Income in Ireland but the application of the various double tax reliefs will significantly offset their gross tax due in Ireland.  The end point of tax due as a proportion of taxable income was 1.9 per cent for for the group of companies with no net domestic trading income in 2015.

As a result of these large foreign profits and small tax amounts due in Ireland it is likely that these companies make up a significant proportion of the 13 companies from the Top 100 ranked by Taxable Income identified by the Comptroller and Auditor General as having effective tax rates of 1 per cent or less.  We looked at this at the time here. 

While the Irish tax due on these foreign profits may be nil it will be the case that the companies will have paid amounts of foreign tax on these profits (that it why they are getting the foreign tax credit) so that the effective tax rates of the companies (as opposed to just their effective rate of Irish tax) will much higher. 

If we want to remedy some of the near-zero effective tax rates identified by the C&AG one solution would be to move to a territorial which would take large amounts of foreign income out of the Irish tax base and negate the need to provide a foreign tax credit.  This would reduce companies Taxable Income to that generated by their Irish activities (or any income attributed to their Irish activities by any CFC rules Ireland might introduce) and the effective tax rate would be much closer to 12.5 per cent.  The C&AG report that 79 of the Top 100 companies had effective tax rates above 10 per cent with 59 having effective tax rates greater than 12.5 per cent.  Moving to a territorial system would increase those numbers.

The second main reason for the low effective tax rates identified by the C&AG is the R&D tax credit.  We do get a breakdown of the R&D credit used against tax in the current year by range of net income.  As shown above just €2 million of the €349 million of the credit used in this manner in 2015 was for companies with no net trading income and this was spread across 94 companies.  The figures show that €302 million went to 99 companies with to net trading incomes greater than €10 million.

When looking at low effective tax rates it would probably be much more informative to look at the distribution of the €359 million of the R&D tax credit that was paid to companies as the credit they were entitled to exceeded their tax liability. Unfortunately a breakdown of this component of the R&D tax credit is not provided in the Revenue statistics. 

The very fact that this is a payment of an excess credit means that these companies will have a negative effective tax rate.  We do not know how many such companies  are in the Top 100 as set out by the C&AG but again note, that like the foreign tax credit, it is a logical explanation, and deliberate policy choice, behind the low effective tax rates identified by the C&AG. 

So, if we want to further remedy the low effective tax rates the solution in this instance is to abolish the R&D tax credit.  Do that and even more of the Top 100 will have effective tax rates close to 12.5 per cent. 

The Revenue Commissioners have actually considered this and in recent evidence to the Public Accounts Committee, the Chair of the Revenue Commissioners said:

…if the effective tax rate of each of the 13 companies is calculated before taking account of double tax relief and the R&D tax credit, each would have an effective tax rate in excess of 12%.

So this really is only something to get excited about unless we think companies are paying tax elsewhere or incurring R&D expenditure to avoid Irish taxes.  If anything the C&AG report confirms that the Taxable Income of companies is taxed at close to the 12.5 per cent rate.  It is in the determination of that Taxable Income where most of the action is.

Very Low Work Intensity by Household Type

Ireland’s rate of people living in households with very low work intensity has received increased attention over the past few years.  It’s not hard to see why.

It could be that this is an issue of composition.  That is, maybe our population structure is such that we have a greater proportion of households with a higher tendency in general towards the end of the work-intensity scale.  Looking at very low work intensity by household type should throw some light on this.

So is it an issue of composition? No.  There is no household type for which Ireland are better than fourth last.  We have very low work intensity across the board.

Computation and Concentration of UK Corporation Tax

Ireland’s Corporation Tax statistics get a fair bit of attention.  Similar statistics are produced for the UK by HRMC.  First up the aggregate Corporation Tax computation.

The common origins of the two systems mean that many of the descriptions are similar and it follows a form similar to the Irish version (the most recent update of which can be seen here).

For the UK we can see that the starting point of Gross Trading Profits and after subtracting capital allowances and other deductions and adding Other Income & Gains we get to Total Chargeable Profits.  The applicable rates are set against this to give the Total Tax Charge and after allowing for reliefs and set-offs we get to the bottom line: Corporation Tax Payable.

The sequence of numbers shown above don’t necessary add to the outcomes shown as in some cases (such as capital allowances) the amounts are those available in the year rather than those actually offset against profits in the year as losses brought forward are not included in Net Trading Profits.

The majority of Chargeable Profits are taxed at the Main Rate.  This has been reduced over the past decade (it was 30 per cent in 2008) and for the period shown in the table above there were reductions every year as it fell from 26 per cent to 20 per cent.  This has the effect of reducing the “tax payable as a % of chargeable profits” figure in the final row. 

This figure is fairly close to the Main Rate for all years and it can be seen that the largest item reducing the Total Tax Charge is Double Tax Relief which relates to non-UK tax already paid on non-UK profits included in the Chargeable Profits of companies.  This is also a key reason why this measure of below the 12.5 per cent headline rate for Ireland though it has been joined by the R&D tax in recent years as explored here. 

A second issue of relevance to Ireland is the concentration of payments.  Statisticians from the Revenue Commissioners have done excellent work on this recently and we now know that 10 companies account for around 40 per cent of Irish Corporation Tax payments.  For the UK the key points on concentration made in the HMRC report are:

Key points:

1. The distribution of companies’ tax liabilities is highly skewed. In 2015- 16 about 7,000 companies (under 1 per cent) had liabilities of £500,000 or more, between them contributing around 54 per cent of total Corporation Tax payable.

2. Companies with liabilities of less than £10,000 comprised about 65 per cent of the total number of companies liable for corporation tax in 2015- 16, but owed only around 7 per cent of the total Corporation Tax payable.

3. In 2015-16, around 50 companies had more than £50 million each in Corporation Tax liabilities (totalling £5.4 billion or 12 per cent of the total Corporation Tax payable). The figures for 2014-15 were around 60 companies paying £6.9 billion or 16 per cent of the total Corporation Tax payable.

4. There was an increase of around 140 thousand in the number of companies with any liability between 2014-15 and 2015-16. This increase was largely concentrated in companies with a Corporation Tax liability of under £50,000.

So the receipts are concentrated but not anything near to the same extent as they are in Ireland.

Finally, the HMRC report has lots of detail on the use of capital allowances – but there is no mention or breakdown provided for capital allowances for capital expenditure on intangible assets. Pity.

Ireland’s capital stock continues to expand

The CSO have provided the 2016 update of Ireland’s fixed capital stock.  This has been affected in recent years by aircraft for leasing and intangible assets.  This continued in recent year.

Here is Ireland’s gross capital stock in 2015 prices.

This measure gives an estimate of the amount of fixed capital held at the end of each year.  Gross depreciation is not taken into account and the amounts reflect the value of the fixed assets (in 2015 prices) as if they were new.  Changes are the result of additions to the capital stock and obsolescence of existing fixed assets.

There have been a number of data suppressions in recent years which means that we only have a combined category for aircraft and R&D intangible assets.  There were increases in all categories of fixed assets though the largest contribution to the overall increase came from aircraft and IP.

Unsurprisingly it is estimated that there was a very small increase in the stock of dwellings, rising by just one per cent in 2016.  Other buildings and structures rose at an appreciably faster pace at 4.5 per cent.

Next is a net measure of the capital stock which accounts for depreciation.  This better reflects the value in use of fixed assets rather than their replacement cost.  Again we will consider this in constant prices to look at changes in the quantity of assets from year to year.

We know there were large depreciation charges for aircraft and IP over the past few years (it is was it messing up the GNP and GNI figures) but we can see that even accounting for depreciation the stock of these assets increased by almost five per cent in 2016.  They are still arriving faster than the existing stock is depreciating.  Again on dwellings it is not surprising to not that our stock of dwellings barely changed in 2016, with a rise of just 0.5 per cent.

And finally, here are the change in the depreciation on these assets (again in constant prices)

Although the increase in depreciation in 2016 was much lower than the huge-level shift that occurred in 2016 we can see that around three-quarters of the 2016 increase is due to aircraft and intellectual property.  This, of course, matches the first table where we saw that an equivalent amount of the increase in the gross capital stock was related to these assets which have few direct links to activities in Ireland.  The aircraft are flying all over the world and many of the products that are derived from the IP are made in other countries with the R&D behind it all also having being undertaken elsewhere.

The figures for “total ex. aircraft and IP” at the bottom of each of the tables give a better indication of the underlying changes in Ireland’s capital stock.

Non-Profit Institutions Serving Households

As well as giving a split of the non-financial corporate sector the 2016 institutional sector accounts also give a split for the household sector.  Up to now households (S.14) and non-profit institutions serving households (S.15) were combined.  We now have separate current and capital accounts for each.  Non-profit institutions serving households (NPISHs) are relatively small in the overall scheme of things (they are around two per cent of the combined sector) but we do get to see how they are funded and what, in rough terms, they spend the money on.

It should be noted that this doesn’t cover all charitable or voluntary organisations many of which are counted in the government sector given the nature of their links to the public sector.  As the CSO note:

Many charities, such as hospitals and social care providers, which receive most of their funding from government and which provide services under contract with government, are treated as part of the government sector (S.13) in national accounts, and not S.15.

Anyway here is the aggregate current account of those entities included in this sector.

On the income side we can see that these institutions had around €3.3 billion of income in 2016.  Most of this came in the form of transfers with around €780 million raised from their own activities.  The CSO note:

The institutions in the NPISH sector get most of their income as transfers from households (for example donations, church collections, union dues or members' subscriptions) and also receive transfers from government.  They also earn profits on their market activities, such as charity shops and match tickets.  Relatively little of their income is from investments.

On the current expenditure side almost all goes on compensation of employees and final consumption expenditure which between them were €3.2 billion in 2016:

Their income is spent on final consumption expenditure and compensation of employees in a roughly 3:1 ratio. There are over 55,000 people in direct paid employment in the NPISH sector.

Expenditure on staff was €1.14 billion in 2016 which gives an average of €20,700 per person in direct employment.  We also get a capital account for the sector but there isn’t a whole lot going on there.

For the past few years the gross capital formation of the sector has been less than the gross saving from the current account and investment grants received shown here.  This has resulted in the sector being a net lender averaging around €45 million for the past three years.  The figures for net capital formation are positive which means that gross capital formation was lower than the consumption of fixed capital (depreciation).

New breakdown of Non-Financial Corporates in the Sector Accounts

An earlier post went into the gory details of the household sector.  Here we do something similar for the non-financial corporate sector and assess what can be learned from a new breakdown of this sector provided by the CSO.

There is lots going on in the current account of the non-financial corporate (NFC) sector but it is hard to tell what the underlying patterns are.  Here is the NFC current account since for the past five years from the 2016 institutional sector accounts.

The big changes happened in 2015 when gross value added jumped by more than 50 per cent to reach €180 billion with an increase of a near similar scale showing for gross operating surplus.  We know this was the result of activities of foreign-owned MNCs and it probably wouldn’t be much of any issue if the pollution was limited to the estimates of GDP but we can see that after profit and interest distributions that gross national income in 2015 still jumped by almost €30 billion. 

So more than half of the increase in gross profits of the NFC sector in 2015 was attributed to Irish residents.  There was a bit of fumbling around when the figures first came out but now we have a fairly good handle on what happened.

Now the CSO are giving a further useful breakdown that allows us to see what happened by trying to isolate some of the distortions.  Figures have been provided for two sub-components of the NFC sector:

• large foreign-owned NFCs
• other NFCs

As the CSO say in the background notes:

Non-financial corporations are sub-divided into Large foreign-owned MNEs (S.11a) and the Other (S.11b) in these accounts. Large foreign-owned MNEs are those companies surveyed by the CSO's Large Cases Unit. This division is not prescribed in ESA2010 but is an additional level of detail provided because of the nature of the Irish economy. This sub-division is a step towards a full separation of domestic and foreign-owned corporations, and allows a more informed perspective on the purely domestic economy.

In the release they further say:

These 50 largest foreign MNEs (out of approximately 114,000 enterprises in S.11) are presented as a proxy for all the MNEs in this release. A more comprehensive account of foreign-owned enterprises is currently under development.

So what do we see if we split the current account into these 50 foreign-owned MNCs and the rest? Lots.  Here are their current accounts for the last three years.

A wider table that also includes the overall totals is available here.  Breakdowns aren’t provided for all of the constituent parts of the current account but most of the important ones are included.  The panel on the right hand side is a huge step in giving us a view of the underlying trends in Ireland’s business sector.  Over the past three years we see that gross value added has been increasing (+7.3% in 2016), compensation of employees is growing (+6.3%) and gross operating surplus is rising (+8.3%). 

There may be a little bit of an issue with redomiciled PLCs or some other issue in the distribution of income account as gross national income recorded an increase of 17.9% in 2016 but all in all the new breakdown is very useful.  Corporation Tax payments from these companies rose a further 13.2% in 2016 to reach €3.55 billion (and up €1.3 billion on 2014).  Thus, the right panel presents a story of a business sector growing strongly. 

And that means that most, but not all, of the problems are corralled in the large, foreign-owned NFC subsector.  This is a small group of companies but ones which have a disproportionate, and distortionary, effect on Ireland’s national accounts.

The €50 billion jumps in gross value added and gross operating surplus that occurred in 2015 are obvious.  As stated above the problems really kick in after the distribution of income.

As these are foreign-owned companies we would expect their direct contribution to Irish gross national income to be minimal.  Their contribution would have been paid out to other sectors: buying goods and services in their intermediate consumption from domestic suppliers and wages paid to the household sector.  In fact by the time we get to gross national income all we would expect to be left is whatever is needed to cover their Irish Corporation Tax bill.  We would expect any remaining profits to be either distributed or attributed to the foreign shareholders.

But that is not the case.  The gross national income of the large, foreign-owned NFCs far exceeds their Corporation Tax payments and at the bottom we would expect their gross disposable income to be close to nil.  We can see that it was €3.1 billion in 2014, jumped hugely in 2015 and rose again in 2016 to stand at €31.2 billion.  This is counted as our income. It is not.

As an quick aside here are the Corporation Tax payments attributed to the subcategories of NFC and also to Financial Corporates over the past four years.

Compared to 2014, Corporation Tax payments for 2016 are shown to be €2.9 billion or 57.6 per cent higher.  All the categories shown paid more but the small relative growth was for these large, foreign-owned NFCs which paid 53 per cent more Corporation Tax in 2016 compared to 2014.  For other NFCs the increase was 59.7 per cent and it was 60.4 per cent for financial corporations.  It should be noted though that these tax amounts are inclusive of the R&D tax credit (which as a payable tax credit related to capital spending (as research and development is now treated) is counted as an investment grant received).

Anyway, back to this huge level shift in GNI from foreign-owned NFCs in 2015.  The reason is because nearly €25 billion of profits of foreign-owned companies weren’t counted as a factor outflow.  There are two possible reasons:

• The first is different treatment of certain items in the national accounts (where gross operating profit is estimated) and in the balance of payments (where factor outflows are derived).  We previously considered some issues around the treatment of spending on R&D service imports.
• The second, and most significant, is that profit outflows are based on net operating surplus and there is now a huge amount of gross operating surplus that is consumed by depreciation.

We can see some things that point to the second issue in the capital account.

Unfortunately, both the capital accounts of both sub-groups are bit of a mess.  For the large, foreign-owned group the acquisition and depreciation of intangibles is throwing the numbers awry while for other NFCs it is likely that the acquisition and depreciation of aircraft are muddying the waters (not forgetting that gross savings is inflated by the net income of redomiciled PLCs).

For the group of large foreign-owned NFCs we can see the large changes that occurred in the depreciation charge.  Consumption of fixed capital for these 50 companies was €5.5 billion in 2014 and this surged to €29.4 billion in 2015 with a further increase to €32.8 billion in 2016.

In the National Income and Expenditure Accounts the CSO provided details of a modified measure of national income, GNI* and one of the adjustments made is for the depreciation on certain foreign-owned intellectual property assets.  This depreciation  went from €0.8 billion in 2014, to €25.0 billion in 2015 to €27.8 billion in 2016.  This is what has driven the changes in the consumption of fixed capital for the large, foreign-owned NFCs shown in the table above.

Although these companies have large amounts of gross savings their expenditure on gross capital formation is volatile and can exceed the level of gross savings.  The financial transactions account for the subcategories might throw some additional light in the thing but although great strides forward have been made we haven’t got that far – yet. For the large, foreign-owned NFCs we can surmise that some of these are running large surpluses to repay loans they assumed in the process of acquiring large amounts of intangible assets. 

At the same time other companies are borrowing to acquire intangibles so it is hard to tell what is happening.  So, in 2015 there was net lending available to repay debt (net lending of €10.7 billion) while in 2016 additional borrowing for intangibles swamped the repayments that some companies were making (resulting in net borrowing of €16.5 billion).

While we don’t have the financial transactions account we do, though, have the financial balance sheets of the two sub-groups.

Plenty of big numbers there.  Unfortunately the loans liability category is suppressed.  However, we do have total financial liabilities.  We can see that for the group of 50 large, foreign-owned NFCs this jumped from €198 billion at the end of 2014 to €516 billion at the end of 2015.  A year later and it was €519 billion.  This is vaguely supportive of the idea of some large loans being repaid while other, relatively smaller, loans are being taken out as part of the onshoring of intangibles.

The balance sheet of the Other NFCs category tells us nothing about the domestic business sector.  The numbers are huge.  By the end of 2016 these companies has €1.2 trillion of financial assets and €1.5 trillion of financial liabilities.  There’s still a bit of stripping out to do here.

None of this is straightforward but this latest release is another step along the way.  GNI* is a promising measure that will likely improve in subsequent rounds.  The current account of the balance of payments is still a bit of a mystery but maybe we know where we’d like to end up.  For the sector accounts we’d definitely like a foreign/domestic split for the NFCs.  The split provided here gives some reasonable growth measures for output in the current account but there’s still room for improvement on the income, capital and balance sheet side of things.

Some snapshots of the aggregate improvements in the household sector accounts

The last post went a bit heavy on the detail in the household sector accounts.  Here we try and pull out a few snapshots from the accounts that give visual pointers to the aggregate improvements over the past few years.  All the data here is nominal and is for the household sector combined with non-profit institutions serving household but the impact of these on the aggregates is relatively minor and they have almost no impact on the trends.

We’ll start with income and consumption and we have series for these that are showing steady growth for the past few years (with the series also having been extended back to 1995).

Which combined give the following gross savings rate:

Since 1995 this has averaged 8 per cent so the 2016 level is about a percentage point below that.

The increase in income has largely been driven by a rise in the compensation of employees received by the household sector which, in nominal aggregate terms, is back to the pre-crisis peak.

Though tax and social contributions are now higher than they were pre-crisis:

And, of course, rents are increasing.  The aggregate amount of actual rentals paid for housing exceeded €4 billion for the first time in 2016.

It is from a very low base (compared to pre-crisis levels at any rate) but household sector capital formation is beginning to pick up:

After a number of years where household spending (consumption plus capital formation) was less than total household income, the household sector has returned to being a net borrower in the past few years though no where near the levels that were seen pre-crisis.

This pattern is reflected in household financial transactions with household transactions increasing both assets and liabilities up to 2008 and reducing them since.

By and large net financial transactions from the financial transactions account and net (borrowing)/lending from the capital account track each other:

Here is the impact of these transactions on household deposits and loan liabilities:

After peaking at €203 billion in 2008, household loan liabilities have been steadily declining since then and had reduced to €143 billion by the end of 2016.  Household deposits have been relatively stable for the past decade but what is noticeable is that the level of household deposits almost matches the level of household loans for the fist time since 2002 – though these aggregate data tell is nothing about the distribution of these assets and liabilities.

Add in the impact of other financial assets and liabilities and we get the overall balance sheet position:

The divergence of financial assets (rising) and financial liabilities (falling) since 2008 is clear. This has meant that the household net financial asset position has been increasing and stood at €210 billion at the end of 2016.  This compares to €130 billion at the end of 2006.

Finally, here is a measure of debt to income:

This ratio of total financial liabilities to total disposable income has fallen from 220 per cent in 2011 to 160 per cent in 2016 and is now back near levels last seen in 2004.  The progress is understandably slow but we’re getting there.

Digging into the detail of the household sector accounts

The CSO have published the Institutional Sector Accounts, Non-Financial and Financial, for 2016.  These are very useful datasets and the sectors included have been further broken down.  For the household sector we now have separate account for households and non-profit institutions serving households while the non-financial corporate sector has been usefully split out in large, foreign-owned NFCs and other NFCs.  We will come back to these.  Here we will focus on developments in the household sector. 

First, here is the current account in all its gory glory.  Most of the figures come from the CSO release but some of the breakdowns are only available from Eurostat so will be updated for 2016 in due course.  Some of these breakdowns have been crudely adjusted to fit the revised total published by the CSO but the broad changes are correct.  Only figures with published outturns for 2016 are given an annual change in the final column. Click to enlarge and also get rid of the fuzziness it seems.

An more palatable abridged version with only updated figures is reproduced here.  Gross disposable income is estimated to have grown by 3.9 per cent in 2016 to reach €92.6 billion with consumption expenditure rising a touch more at 4.1 per cent.  Both of these growth rates are largely unchanged from the preliminary estimates.  This means that the gross savings rate for 2016 was little changed from what it was in 2015 with the final row of the table showing figures of around 7 per cent for both years.

However, while the growth rates of income and consumption may be little changed, the level of gross savings of the household sector has been significantly revised down from the preliminary estimates.  The Q4 ISAs put it at €11,714 million and we can see above that it is now put at €6,386 million.

We examined these revisions here and note that the relate to the line right at the very top of the household current account: gross value added.  The level of output of the household sector, and in particular the self-employed, has been significantly revised down.  This means the estimate of the gross household savings rate for 2016 has gone from 11.5 per cent in the preliminary figures to 6.9 per cent now.  The savings rates have been revised down for all years since 2010.  There is no “right” savings rate but for 2016 these revisions seems to be a shift from one that seems a little high to one that seems a little low.  This has implications for forecasts of consumption growth relative to forecasts of income growth.

Anyway, that minor quibble aside the household current account is continuing to show solid improvement. We might have a new level but the output of households and the self-employed grew, in nominal terms, by 5.2 per cent in 2016 (though the extent that this is due to increases in imputed rentals for homeowners remains to be seen).  Compensation of employees received by the household sector grew by 5.4 per cent though there were declines in property income received. All in all gross national income of the household sector grew 4.9 per cent.

Taxes paid on income and wealth grew by 3.2 per cent while social contribution paid to the government (mainly PRSI and pension contributions by public sector employees) grew by 5.8%.  The lower growth of taxes on income reflects the impact of policy measures (primarily on USC) with PRSI growing in line with the increase in compensation of employees.  Social benefits received remained flat with €22.8 billion paid out by the government sector. 

Combine all these changes and we get to the 3.9 per cent increase is gross disposable income.  In line with this consumption grew by 4.1 per cent.  Consumption items growing by more than 10 per cent were the purchases of vehicles (+16%) and actual rents for housing (+13%).

All told, the household sector was left with gross savings of €6.4 billion in 2016 after all current items are accounted for.  We next turn to the capital account.

In 2016 gross capital formation of the household sector was just under €8 billion which exceeds the gross savings that arose in the current account which means (after accounting for relatively minor amounts of capital taxes and transfers) that the household sector was a net borrower in 2016 of €1.4 billion.  This roughly means that household expenditure for current and capital purposes exceeded the income available to meet such expenditure.

We can see that this different significantly from 2012 when the household sector was a net lender to the tune of €4.8 billion.  The €6 billion change is roughly divided between a €3 billion reduction in gross savings from the current account and a €3 billion rise in gross capital formation in the capital account.  In order to cover its current and capital expenditure in 2016 the household sector was a net borrower.

We can try to get some insight into this shift from net lending to borrowing by looking at the financial transaction account.

For most of the items the outcome of transactions over the past five years is what we would largely expect.  As the second last row shows over the past five years net financial transactions have increased household financial net worth by almost €16 billion.

In net terms, households have added nearly €11 billion to their deposits while net loan transaction (drawdowns minus repayments) have reduced loan liabilities by €27 billion.  We have continued to contribute a net two to three billion a year to insurance and pension reserves.

The standout figure is the minus €36 billion for equity transactions, of which €34 billion relates to unlisted shares.  It is these transactions that are hanging the financial account together.  In the final row of the table we can see that there are some differences between net financial transactions in the financial transaction account and net(borrowing)/lending in the capital account.  However, over then years these sum to less than €10 billion and are within the realms that they could be explained by re-valuations, write-offs or sales of non-financial assets to other sectors.

But without the minus €36 billion of equity transactions the plus €11 billion of net deposit transactions and minus €26 billion of net loan transactions would be hard to explain.  But how much of an explanation is it? What exactly was the €34 billion of unlisted shares that we sold (assuming that the minus figure for financial assets refers to their sale)?

Anyway, we can see how these transactions have impacted household net financial worth by looking at the financial balance sheet.  Click to enlarge.

The last line shows that the net financial worth of the household sector has increased by €76 billion over the past five years, rising from €133.8 billion at the end of 2012 to €209.9 billion at the end of 2016.  The final column shows that this is €60 billion more can be explained by net financial transactions (which as we saw were plus €16 billion).

Looking at the liability side of the balance sheet shows a fairly close comparison between changes in the stock of financial liabilities and net liability transactions.  Loan liabilities over the past five years fell by just over €30 billion so the minus €27 billion of transactions doesn’t leave much to be explained by re-valuations or write-offs.

But there seems to be lots going on on the financial asset side of the balance sheet.  The €10 billion rise in deposits closely matches the plus €11 billion of transactions.  However, insurance and pension reserves rose by €35.6 billion on the back of plus €12.8 billion of transactions.  The remaining €22.9 billion is due to revaluations which we can presume is linked is rising share and debt asset prices.

And again we turn to direct holdings of equity.  Over the last five years the value of the stock of equity assets held by the household sector has been largely unchanged: €46.4 billion in 2012 versus €46.8 billion in 2016.  Of course, we have just seen that there has also been minus €36 billion of transactions with assets in this category with almost all of this relating to unlisted shares..

The detailed table shows that the value of listed held by households rose from €9.9 billion in 2012 to €14.1 billion in 2016.  On the other hand the value of unlisted shares fell from €36.5 billion to €32.8 billion but a drop of just under €4 billion pales in comparison to the minus €34 billion of transactions.  We are left with a related question.  First, we didn’t know what the household sector was selling, now we have to wonder how the household sector sold €34 billion of an asset it had €36.5 billion to begin with and still be left with nearly €33 billion at the end.  That’s a pretty strong revaluation effect!

But this is nothing new for this category.  Since 2003 the net financial transactions of the household sector in unlisted shares sum to almost €60 billion yet in spite of these negative transactions the value has always been between €33 billion and €42 billion.  The top of that range was recorded in 2005 while 2016 gave rise to the bottom of the range.  It seems we’re creating the value almost as quick as we’re realising the value.

When the revisions of to household income and the savings rate were first published in the Q1 2017 ISAs we said the coherence that appeared in the last annual accounts was no more – what large net lending explaining most of the deleveraging by the household sector.  The 2016 annual accounts once again present a coherent story and though the household sector may now be a net borrower the continued reduction in debt liabilities (and accumulation of deposits) is balanced by negative transactions in unlisted equity.  We obviously don’t know the distribution of these transactions but they do add up in aggregate.

Anyway, after all that digging what are we left with?  We have a household sector that is showing improvement across the current, capital and financial accounts.  It is a hard slog but we are slowly working through the excess that built up in the run-up to 2008.  We will look at these improvements visually in the next post.