The Pfizer impact on Irish statistics

At the end of December 2014 the retained earnings on Pfizer’s balance sheet was $72.2 billion.  Irish GNP is going to look pretty impressive whenever this inversion is completed as much of this will probably be counted as Irish income.  And on an on-going basis Pfizer’s retained profits will count as an income inflow in the Balance of Payments and be added to Irish GNP.

For the last few years Pfizer has paid total dividends on it stock of around $6.5 billion.  Net income for the past three years has been $14.6 billion, $22.0 billion and $9.1 billion.  Thus the retained earnings have been $8.1 billion, $16.5 billion and $2.6 billion.  These are now included in the $72.2 billion currently on the balance sheet but it can be seen that the annual amounts can be very significant and - volatile. 

Interpreting Irish GNP figures may become even more meaningless.  What won’t be meaningless is the additional EU contributions that we may to make on the basis our of our reported National Income. Is it time for an Irish rebate in the mould of the British rebate?

A less significant impact of the Pfizer inversion will be on the calculation of effective Corporation Tax rates in Ireland.  This is because dividends from Pfizer’s subsidiaries will be added as ‘Foreign Income’ in the determination of taxable income in Ireland.  However, Pfizer will get credits under ‘Double Taxation Relief’ for corporate income taxes paid elsewhere so the amount of net Corporation Tax due will not increase to the extent that Taxable Income will increase.

Sound arithmetic; shocking logic

The Irish Times yesterday featured a piece which said:

Pfizer reported international revenue last year of $28.5 billion and a group profit margin of 18.4 per cent. On that basis, its international profits were about $5.3 billion, translating to a tax bill of €620 million at the Irish corporation tax rate at the current euro/dollar exchange rate if it was all taxed here.

Yes, $28.5 billion multiplied by .184 is $5.24 billion and that multiplied by .125 is $0.65 billion which at the current €/$ exchange rate of 1.06 corresponds to €618 million.  Flawless arithmetic.

The problems with this are many.  Let’s go through the numbers using Pfizer’s most recent 10k SEC filing.

The $28.5 billion revenue figure is Pfizer’s non-US biopharmaceutical revenue.  Pfizer’s total non-US revenue was $30.5 billion so it would seem this is the appropriate revenue to use.  But why would the distinction between US and non-US revenue be important for determining Irish Corporation Tax.  So what if $30.5 billion of Pfizer’s $49.6 billion is earned outside the US.  What makes the $19.1 billion of US revenue earned by Pfizer irrelevant when Pfizer becomes an Irish company?

Next up is the profit margin used of 18.4%.  Firstly, this is the group profit margin so there is no way of knowing whether this applies to the $28.5 billion of revenue used in the calculation.  Next is the problem that it is the post-tax profit margin.  The objective seems to be to calculate some form of taxable income so it should be the margin before existing taxes are subtracted.  Pfizer’s 10K form shows that this was 24.7% for the group in 2014.

So maybe we should be multiplying €49.6 billion by .247 which gives us $12.24 billion.  The next step in the arithmetic was to multiply the taxable income figure by the Irish 12.5% tax rate to get the tax due.  Using the $12.24 billion we have a figure $1.53 billion.  Wouldn’t $1.5 billion of extra tax have made for an even better headline?

Of course, it is a nonsense calculation.  The easiest problem to see is that Pfizer already has operations in Ireland so is already paying our 12.5% Corporation Tax on a share of its profits. Just because they move the headquarters here doesn’t mean we can tax them twice on their Irish profits!

However, the most serious error is the assumption the piece makes the Ireland’s 12.5% Corporation Tax could be due on all of Pfizer’s profits (even if it makes the division between US and non-US profits which is irrelevant for Irish purposes).  The piece says:

In Ireland, it will pay our 12.5 per cent tax rate on any international income routed through the new Dublin operation.

It is true that Ireland levies the 12.5% rate on dividends received by Irish-resident companies from their trading subsidiaries in EU and treaty countries [It is 25% for non-treaty countries].  So when Pfizer becomes Irish-headquartered the dividends the parent receives from subsidiaries will be taxed at the 12.5% rate.

However, the crucial element is that the tax is applied at a net effective rate of 12.5%.  Companies get a tax credit for any foreign tax they have already paid on the profits and only have to pay additional tax to the extent that the tax already paid is below 12.5%.  The hints at this but seems to get the logic in reverse:

The precise benefit will depend on how much international revenue is channelled through the merged Irish business. Pfizer may choose to pay tax locally in countries with even lower tax rates.

This seems to suggest the tax paid in Ireland would be lower if Pfizer pays lower taxes in other countries.  The opposite is the case.  The less tax Pfizer pays in other countries the more tax will be payable here.

What countries does Pfizer operate in? Most of Pfizer’s R&D takes place in the U.S. Pfizer has major manufacturing facilities in Belgium, China, Germany, Ireland, Italy, Japan, Puerto Rico, Singapore and the U.S and operates multiple distribution facilities around the world.  It is likely that the tax paid in many of these countries will exceed our 12.5% rate.

The exceptions are maybe Puerto Rico and Singapore where the tax paid by the these subsidiaries might result in additional tax being due in Ireland to bring the tax on these profits up to an effective 12.5%.  However, Pfizer will be able to avail of “dividend pooling” so that the tax credits on all the foreign dividends received can be combined before they are offset against the Corporation Tax due in Ireland.

Even after the inversion Pfizer will earn a substantial portion of its profits in the U.S.  This will now be counted as a dividend from a subsidiary for the Irish HQ.  It is likely that Pfizer will get sufficient tax credits from its US tax to offset any additional tax liability from Ireland it might face.  And then there will be the credits from corporate income tax paid in Belgium, Germany, Japan and other countries.  It is very difficult to see how the inversion alone will result in Pfizer paying any additional Corporation Tax in Ireland even though dividends from its subsidiaries around the world will become liable to our 12.5% Corporation Tax rate.

The appropriate calculation is foreign dividends received multiplied by 0.125 minus double taxation relief equals nil.  But where would be the headline in that?

Pfizer have said that they expect their effective tax rate to go to around 17 per cent once the inversion has washed through.  If a large part of the company’s profits were going to be taxed in Ireland wouldn’t this be closer to our 12.5% headline rate?

Now, it is possible that Pfizer will pay more Corporation Tax in Ireland.  We simply do not know the full detail of what its new tax structure will look like.  Pfizer likely has a major tax restructuring that it wishes to implement of which the inversion is just one part.  But to suggest that Pfizer will be paying our 12.5% tax on its net non-US profits is misleading in the extreme.

So why is Pfizer engineering this inversion? The reason is in this table from its 10k filing.

The reason is the deferred U.S. corporate income tax which was $725 million in 2014.  The table also shows that Pfizer pays lots of ‘international’ income taxes ($2,321 million in 2014) which doesn’t seem to leave much for our double-dip to bring them up to 12.5%.  The problem for Pfizer is that the US’s worldwide regime is applied at 35%.

The table also shows that a large part of Pfizer’s tax provision is not actually paid – it is deferred. The inversion will not really change the income tax Pfizer pays but it will change the corporate income tax it owes.  The cash tax paid of $2,100  million in 2014 (and $2,874 million for 2013) can be compared to the tax provisions in the above table.

Pfizer’s balance sheet shows that it has $25.0 billion of non-current deferred tax liabilities.  This is mainly the US corporate income tax due on its non-US profits that is has deferred using various provision in the US tax code.  This tax does not become payable until Pfizer repatriates the profits to the US.  Pfizer’s 10k statement also says that:

As of December 31, 2014, we have not made a U.S. tax provision on approximately $74.0 billion of unremitted earnings of our international subsidiaries. As these earnings are intended to be indefinitely reinvested overseas, the determination of a hypothetical unrecognized deferred tax liability as of December 31, 2014, is not practicable.

This is the issue the inversion is trying to address.  Pfizer wants greater flexibility to utilise its profits without running the risk of tripping a major US tax payment on profits which are earned outside the US.

Pfizer’s Inversion: Cui bono?

It looks like Pfizer has found a dance partner and that the inversion with Allergan is set to go through.  So who is set to gain from this?

The big advantage for Pfizer is that it is no longer subject to the US corporate income tax of 35 per cent on its worldwide profits.  The US is unusual in that it requires US companies to pay US corporate income tax on their global profits and not just those profits earned in the US.  As the US has the highest corporate income tax rate in the OECD this means that US companies will have to pay additional tax to the US government to bring the overall tax paid on their profits up to the US 35% rate. 

Of course, as lots of companies can engineer a deferral of this tax payment until the profit is formally repatriated to the US, many US companies choose to keep profits "offshore" rather than repatriate them and pay the additional tax due.  This hinders what the company can do with the profits such as transferring to the parent to invest in R&D or to distribute to shareholders. 

Ireland is also in the bracket of unusual countries in that we too have a worldwide regime - but in this instance at one of the lowest corporate income tax rates in the OECD.  This means that Irish companies have to pay very little additional tax on their non-Irish profits as they will already have paid sufficient tax in the countries they earn the profit to bring them up to (and over) the 12.5% rate.  By becoming headquartered in Ireland this is the position Pfizer hopes to move to. They will continue pay tax at the appropriate rates in the countries they operate it, but now won't have to pay additional tax when they aggregate the profit in the company’s HQ unlike they would have to do if the HQ remained in the US.

This won't change hugely the amount of tax that Pfizer pays – though it will change the amount of tax that it owes.  Up to now Pfizer has owed US corporate income tax on its non-US profits but can defer the actual payment of most of it using various provisions in the US tax code.  But it is by doing this that Pfizer faces restrictions on what it can do with its profits.  The inversion will have no impact on most of the countries Pfizer operates in (UK, Germany, France etc.) and the company will continue to pay corporate income tax on the profits it earns in those countries at the rates those countries levy.  The move will also have little impact on the amount of Corporation Tax the company will pay in Ireland.

By becoming Irish-headquartered all of Pfizer's global profits will become subject to Ireland's Corporation Tax and not just the profits it earns from  its existing activities here.  However, as the tax paid in the countries where these profits are earned is likely to exceed our 12.5% rate the amount of additional tax that will be due in Ireland will be nil.

There are almost no benefits to Ireland from these inversions.  There are likely to be further distortions to our national accounts and if the switch to becoming Irish-headquartered causes Irish GNP to be higher we could potentially face a higher bill for our contribution to the EU.  So if anything it could cost us.

The benefits are for the company that escapes the clutches of the US 35% corporate income tax rate on its profits earned outside the US though Pfizer will continue to be subject to the US 35% rate on the profit it earns in the US. 

Of course, once the structure is in place this will impact on future decisions of the company so the amount to be taxable in various jurisdictions may be different to what it would be had the inversion not taken place.

On the unequal distribution of our increased income

TASC have released a report which states the following:

According to figures released by Revenue, the total income of those liable for income tax was €77 billion in 2011. By their calculations, this is expected to rise to €98 billion in 2016. This €21 billion in new income represents an increase of more than a quarter in just five years. This includes 355,000 new tax cases (either couple or individuals), a rise of 17% from just over two million to 2.4 million (see appendix of this paper for all calculations).

The most startling thing about these figures is how unequal the gains have been. Of the €21 billion in extra cash, about €12 billion, more than half, has gone to the top 10% of earners. Two thirds of it - €14 billion - has gone to people who earn more than €70,000, despite the fact that they make up less than 15% of Revenue’s income tax cases.

On the other hand, less than a third (€6.5 billion) has gone to the middle 60% of earners, with only 6% of all the increase going to the bottom 50%.

This represents a fundamental shift in how the market distributes income. From 2011 to 2016 the share of income going to middle income earners fell from 52% to 46% which is a loss of more than 10% of the share of all income. At the same time, the Top 1% went from having 9% of all income to having 11% of all income – a gain of more than 20%.

Given that there is an expected €21 billion increase in Gross Income from 2011 to 2016 one wonders whether the definition of Gross Income has changed?  Will Gross Income really have increased by 27 per cent over this five year period? 

If true it is a remarkable performance.  However, if “more than half, has gone to the top 10% of earners” then that would knock a lot of the lustre off it.  But the report itself points to a couple of points which indicate that the distribution of the increase may not be as skewed as the headline-grabbing points suggest.

The first is that there are 355,000 new tax cases.  By definition these have to have received additional income.  However, we don’t know where they have entered the distribution so it’s hard to draw out how much of the additional income has gone to new earners and how much has gone to existing earners.  If new earners enter towards the top of the distribution is this a bad thing?

And this points to the second issue with the analysis.  Who is in the group that comprise the Top 10 per cent?  Actually the report notes that we don’t have figures for the Top 10 per cent. Footnote 19 says:

Because of the nature of the Revenue data, the groups are approximate. In 2011 the bottom 50% was 49% of all tax cases, the Middle 60% was 58%, the Top 10% was 9.8% and the Top 1% was 0.91%. In 2016 the bottom 50% was 46.3% of tax cases, the middle 60% was 56.4%, the Top 10% was 11.8% and the Top 1% was 1.2%. This will account for some variation in the data.

You can be sure it will “account for some variation in the data”!  Comparing the top 9.8 per cent in 2011 to the top 11.8 per cent in 2016 is misleading.  Of course this group has extra income – you just put an extra 20 per cent of tax cases into it!  Here are the details of tax cases with Gross Incomes over €75,000 in the two years.

The changes are pretty clear.  What is noteworthy is that for a group that apparently got more than half of the €21 billion of additional income is the what happened to the average income of the group.  Yes, averages have their problems and there are lots of moving parts here but the average increased by 2.5 per cent whereas the total income in the group increased by over 45 per cent.  This suggests that most of the increase in the income of this group is down to the increase in the number of people in the group rather than increased income for the people who were in the group to begin with.

If the comparison was a like-for-like comparison so that 9.7 per cent of tax cases were analysed in each year then there would be 234,440 in the 2016 group rather than the 283,674 that were actually in the group.  This means that 49,234 tax cases were added to the “Top 10%” because it actually went from the “Top 9.7%” to the “Top 11.8%”.

If these 49,234 extra tax cases put into the group had a Gross Income of €75,000 (the lowest possible) then this means there is a minimum of €3.7 billion of extra income in the “Top 10%” that is only there because of the increased size of this group.  That is 31 per cent of the additional income attributed to this group.

A headline that the “top earners main winners in recovery” is clearly what the analysis was designed to engineer.  But it is not true.  The main winners are the recovery are the 355,000 new tax cases since 2011.  From the data available is it impossible to say where these new tax cases have entered the distribution.   This would paint a much different picture than doing arithmetic acrobatics with the distribution.

For example, every €10,000 of Gross Income that these new tax cases have is €3.5 billion of Gross Income.  How much of the €21 billion of Gross Income is the result of these new tax cases?  If they’re at an average of €30,000 that is €10.5 billion.  That is half the increase in Gross Income from 2011 to 2016.  It is unequal but it is desirably unequal.

However, the distribution of income Gross Income among recipients in 2016 seems set to similar to what it was in 2011.  Here are the Lorenz curves for the two years:

Using the shares of gross income and tax cases the approximate Gini co-efficient for each year is:

• 2011: 0.475
• 2016: 0.485

There has been an increase in inequality between the two years However without testing for the significance of this difference or knowing what happened in the intervening years we cannot be too conclusive about the pattern.  It is also likely that the distribution of market income across the population is more equally distributed as there are 355,000 new tax cases in 2016 compared to 2011.

Resources available for public service: how does Ireland compare

An article in The Irish Times today has a surprising opening line:

The share of our income that goes on Government services such as welfare, health and education is very low by European standards.

Surprising because here is the conclusion I came to when looking at public spending on health, education and social welfare excluding old age social protection.

Using a hybrid measure of national income for Ireland (between GDP and GNP) Ireland is the fourth highest spender in the EU on health, education and welfare excluding old age social protection.  It is not what we spend that matters (we spend plenty) it is how we spend that matters.

We can do lots of acrobatics to see why Ireland is below average in the EU when it comes to public spending but one reason stands above all others: spending on old-age social protection.

In most countries this is the largest single area of government spending.  There are two reasons why Ireland is at the bottom in this category: firstly, as a proportion of the population, we have fewer older people than every other country in the EU, and, secondly our public pension benefits are flat-rated.

The mean expenditure on old-age social protection for the other 27 countries in the EU is 9.4 per cent of GDP.  That is more than double the Irish level.  If Ireland was at this level in 2013 it would have meant spending an extra €9.5 billion on old age social protection.  To reach the level of spending of the countries at the top (14 per cent of GDP ) would require spending almost €18 billion more.

If other countries have better public services than us (what evidence is there that they do?) it can’t be because they devote more of their national income to them than we do.  Other countries do have far more comprehensive systems of public pensions than we do but in those systems those who make the most contributions receive the most benefits.

Below are some tables that look at the breakdown of spending in health, education and social welfare.  No national income adjustment is made so all the data is by GDP. No commentary is provided. Click to enlarge.

For health and education we have a alternative breakdown by type of spending rather than area. Again click to enlarge.