10/07/2011: Irish Tax Rates in International Perspective

Some interesting international comparisons for tax revenues across the EU27, plus Israel, Norway and Switzerland (no Iceland, sadly), courtesy of the OECD dataset - last updated April 27, 2011. I added Ireland's tax ratios relative to GNP based on CSO data for all the years 1999-2009.

Let's run some comparisons:

• In 1999, total tax revenues in Ireland were 33.2% of GDP and 38.9 GNP which compares to 37.% of GDP for the simple average of 30 countries in the sample and 37.2 median. There was a slight (0.3) skew in the data. With a standard deviation of 7.0 that year, Irish tax/GNP ratio was well within the average, which is confirmed by the rank attained by Ireland as 12th highest tax economy in the group.
• In 2003, total tax revenues in Ireland were 30.3% of GDP, which of course would be consistent with FF/PDs 'low tax' policies the Left is keen of accusing them of. Alas same year total tax revenue in Ireland stood at 35.9% GNP which compares to 36.7% of GDP for the simple average of 30 countries in the sample and 36 median. So as Irish tax revenue as a share of economy declined, so did the sample average. The new skew was 0.2 lower than in 1999. Hence, with a standard deviation of 6.5 that year, Irish tax/GNP ratio was again well within the average - actually even closer to the average - which is confirmed by the rank attained by Ireland as 16th highest tax economy in the group.
• Now, note that within both of the above years, in terms of GDP comparative, Irish taxes were ranked 22nd and 26th highest in the sample.
• Zoom on to 2007 when Irish tax revenues accounted for 32.0% of GDP and 38.8% of GNP against the sample average of 38% of GDP and a standard deviation of 5.8. There was zero skewness that year. Once again, there was no statistical difference between Irish tax rates and the average. Ireland ranked 25th highest tax economy in comparison against GDP and 14th in comparison to GNP.
• 2009 is the latest year we have comparatives for and in that year, Irish Government tax revenue accounted for 29.6% of GDP and 35.9% of GNP, which (GNP figure) again was statistically indistinguishable from the mean which was 36.7% (with standard deviation of 6.1 and skew of 0.2).

So now, let's map the above data:
Notice the following features of the above chart:

• Irish tax returns as a function GDP are more volatile than in terms of GNP - in fact historical standard deviation for Irish tax revenues in terms of GDP is 1.406 against that for GNP of 1.210. The median standard deviation for the sample of 30 countries is 0.736.
• Irish tax returns as a function of GDP are always statistically significantly different from the average, but our tax returns as a function of GNP are never once outside the average. In other words, folks, our tax burdens are average. Not low, not high - average.
• Only within the period of 2001-2003 did our tax returns as measured in relation to GNP fall statistically significantly below those for Euro area (EA17).

Let's put our tax revenues against some comparable countries. I divided the following two charts into Small Open Economies that are members of the Euro area and those that are not:
Interestingly, for the Euro are countries, Sweden, Belgium, Austria and Finland have tax burdens in excess of the average (note they are above the 1/2 STDEV band relating to the mean. Notice that all of the countries in that group, with exception of debt-ridden Belgium, are experiencing declines in their tax burden since 1999. Apparently, to the chagrin of our friends in the Trade Unions, Tasc and Irish Times - the ones so keen on shouting about the FF/PD coalition tax policies - the Nordics too were run by right-wing free-marketeers.

Next, notice the countries within the trace band around the mean - these are the Netherlands, Lux, Slovenia, Ireland (GNP), Portugal and Czech. Greece has dropped below the average range around 2004. It's an interesting neighborhood we are in, which includes highly aggressive tax competitor such as the Netherlands.

Lastly, we have a truly aggressively competitive Slovakia.

So again, there is no evidence in sight that Ireland is or was a low tax haven.

Now, for non-Euro countries:
Speaks for itself, but let me cover one little point. Switzerland has ranked within lowest 5 tax economies in 10 out of 11 years between 1999 and 2009. The country with functional public services and great public infrastructure has managed its affairs on the average tax revenues of just 29.3% of its GDP against the average of 31.7% of GDP and 37.5% of GNP for Ireland. So, really, folks, cut this crap about 'low taxes have ruined Irish economy/society'. The Swiss do it on less than us, better than us and achieve great social cohesion, civility and cultural development while using three languages where we can't master two. It's not in how much you spend, it's how you spend it.

10/07/2011: Irish Trade Stats: some interesting points

Here are some interesting end-of-year numbers for 2010 in terms of our external trade. Note - these are from OECD stats via ST Louis Federal Reserve database, so slightly off compared to CSO data. All are reported in Euro, unless otherwise specified.

First, consider the flows of trade and trade balance:
There is a clear regime shift in the data since 2009 with a rise in trade surplus. This confirms that Irish net external trade has entered a recovery stage post-crisis in 2009, not in late 2010-early 2011 as the IMF officials claimed recently. The second thing the chart highlights is the dramatic rise in trade balance in 2009-2010, even compared to the strong performance pre-2002. In fact, we reached beyond our trend (for 1997-2010 period) back in 2009.

This might suggest validity to the 'exports-led recovery' thesis, except for two issues:

1. Two years are hardly a trend, especially if coincident with extremely robust global trade recovery post-crisis, and
2. The trade balance is only relevant to Irish economy as a whole if we actually get to keep it here - in other words, if it accrues to companies with really sizeable investment and employment activities here. Note that in the chart above, the last two years have actually seen a negative relationship between growth in the economy and growth in the trade balance.
   

The latter issue is easy to see if we net out of the trade balance the remittances of profits and payments abroad, as done in the chart below:
Notice the decline in Net Factor Income from Abroad (NFIAF) in 2009-2010 period. This is linked directly (more closely than in the case of GDP and GNP changes) to our trade balance:
In other words, what gets produced here in terms of trade surplus gets remitted out of here. As we become more open to trade - as shown below - by any metric possible, we get more open to exporting profits and surpluses accumulated in the economy.
This is similar to an analogy of draining water out of a sinking boat with a coal bucket - when you scoop up water, the bucket is full, by the time you turn it overboard, the bucket is empty...

Some interesting correlations to that effect - all for data from 1997 through 2010, so small sample bias obviously is there:

• Trade balance correlations with GDP and GNP are 0.613 and 0.543, but with NFIFA it is -0.866
• NFIFA itself is correlated with GDP and GNP at -0.904 and -0.861.

So NFIFA has more sgnifcant links to GDP and GNP than our trade balance. In other words, the propensity of our MNCs to take out profits from Ireland has more effect on our GDP and GNP than the trade balance. The recovery, therefore, if it were to be driven by external trade, has less to do with our Exports and Imports, than with profits expatriation decisions by MNCs.

09/07/2011: Construction Activity : Ireland 1980-2010

Rummaging through the Federal Reserve database, I came across a fascinating set of numbers on the number of construction permits issued in Ireland. These are based on index with 100=2005 level of activity.

• By the end of 2010, new dwelling construction activity has fallen from the high of 102.4 attained in 2004 to the low of 15.3.
• Year on year, 2010 activity was down 56.8%. 2010 marks a decline of 80.7% on 5 years ago, 80.6% decline on 10 years ago, 56.4% decline on 15 years ago, 26.1% decline on 20 years ago, 8.5% rise on 25 years ago and 58% decline on 30 years ago.
• The only sustained decline period - other than current - was 1983-1996 period, when activity dropped from 35.2 in 1983 to the trough of 12.8 in 1988 - 4 years of decline and the cumulative drop of 63.6% (much more benign that the current drop of 85.1% to the end of 2010). The recovery in that contraction took over 13 years.
  

So we had a cycle of over 17 years and if one were to count 1981 as a peak with 1982-1983 as a temporary bounce, then the last cycle took 19 years to unwind. Good luck to anyone still hoping for a return to "normal" unless your normal is pre-boom average activity at 51-52 or roughly a half of the construction activity in 2004-2005.

Here's the chart:

08/07/2011: Effects of the spending stimulus on unemployment

An interesting study on the effectiveness of fiscal spending on unemployment was recently published in the CESifo working paper series. The full study can accessed here: Steinar Holden and Victoria Sparrman, "Do Government Purchases Affect Unemployment?" CESifo Working Paper No 3482, May 2011.

The study presents estimated effects of 1% increase in Government purchasing of goods and services on unemployment in 20 OECD countries for the period 1960-2007, controlling for a number of factors, including the size and the openness of the economy, the exchange rate regime and the economy position in the business cycle.

To summarize relevant results (found in Table 7) in the case of small open economies within the currency union, the effect of 1% increase in government purchases of goods and services translates into 0.37 decrease in unemployment rate. The effect can be as high as 0.47% decrease. Year after there is no net effect of jobs creation from the purchasing.

So what does this mean in the case of Ireland? Per latest QNA, Irish GDP in current market prices was €155,992 million in 2010. 1% of that spent on new purchases of goods and services amounts to €1,559,920,000. Q1 2011 unemployment, per QNHS, amounted to 295,700 and the unemployment rate stood at 14.1%. These are our inputs into the estimate.

Now, let's make an assumption concerning jobs created - suppose these pay €35,000 per annum in wages. Suppose that they pay €7,067 in income-related taxes (inclusive of USC etc), as consistent with single tax filer with no deductions. Suppose the social welfare benefits savings amount to €350 per week (note these are taken on purpose to be larger to account for other benefits that might be foregone) to the annual total of €18,200. Suppose that additional 30% is collected on income tax contributions due to higher consumption taxes contributions in employment - generating savings of additional €2,120 per annum.

So total savings per person moved off welfare into employment are roughly speaking €27,287. In other words, we assume that for each €35,000 job created, the Government get back almost €28,000 through various tax returns and savings.

Now on to the estimated impact of 1% increase in Government purchases of goods and services:

1. Case 1: maximum effect of 0.47% reduction in unemployment rate will result in 9,857 jobs created with the total cost of €158,260 per job created. Net of Government returns and savings, this means net cost per each job created of €130,873. Total impact is to generate a loss of 0.84% of GDP due to 'stimulus'. If we are to assume that all of the jobs created remain for ever after the 'stimulus' (a very tall assumption, but let's be generous), while the Government finances the stimulus at a constant interest rate of 6%, it will take almost 7 years for the economy to recover the costs of the 'stimulus' (if the rate of borrowing is zero - e.g. by using NPRF or some other 'free' funding, the period to recovery shrinks to 5.8 years).
2. Case 2: most likely effect of 0.37% reduction in unemployment rate will result in 7,760 jobs created with the total cost of €201,033 per job created. Net of Government returns and savings, this means net cost per each job created of €173,646. Total impact is to generate a loss of 0.88% of GDP due to 'stimulus'. If we are to assume that all of the jobs created remain for ever after the 'stimulus', while the Government finances the stimulus at a constant interest rate of 6%, it will take over 9 years for the economy to recover the costs of the 'stimulus' (if the rate of borrowing is zero, the period to recovery shrinks to 7.3 years).

07/07/2011: What's in the interest rates hikes

Working away on the data for PIIGS, I was interested in a question, what if the ECB were to go to the equilibrium repo rate consistent with the current inflation & growth environment?

In a recent post (here) I did analysis of the ECB historical rates in relation to eurocoin leading indicator of growth. This chart is reproduced here with suggested ranges for the repo rates consistent with current and with higher inflation.
So if the equilibrium rates are in the neighborhood of 2.25-2.75 percent, what would 1% increase in interest rates from June 2011 rate of 1.25% do to the cost of fiscal debts financing across the PIIGS?

Using IMF projections for debt levels for PIIGS through 2016 and assuming that all interest payments are financed out of deficits / borrowing, the chart below shows the extent of the increase in the cost of interest charges on government debt by 2016:
This translates into an increase in the annual cost per capita (2016 forecast) of:

• €560.48 in Greece
• €834.84 in Ireland
• €546.74 in Italy
• €309.24 in Portugal
• €319.02 in Spain

Overall, the increases in interest costs for PIIGS will amount to ca €47.06 billion per annum or 1.23% of the PIIGS GDP and 0.44% of the Euro area GDP. Oh, and by the way, this does not take into account the additional costs of financing banks lending by the ECB.

So that should put into perspective my view of today's hike in the ECB rate, expressed earlier here. So happy wrecking ball swinging, Mr Tri(pe)chet & Co.

06/07/2011: Profit Margins in Irish Services and Manufacturing

Based on latest Manufacturing and Services PMIs, let's update my index of profit margins in Irish economy.

• Profit margins continued to decline in Services, with the rate of decline slowing down from -19.74 in May to -16.02 in June. Profit margins declines are still steeper than 12mo MA of -14.6, but are now more moderate than the Q2 2011 average of 20.1 and Q1 2011 average of 17.4. However, profit margins declines for Q1 2011 were more benign than in Q2 2011.
• Profit margins in Services are now on the declining trend for 24 months straight and have accelerated significantly since 2009 and 2010. 2010 H1 average was -6.9 and 2009 H1 average was -11.0.
  
• Profit margins volatility has risen steeply during the crisis. The standard deviation for profit margins in Services was 5.29 for the entire history of the series and 5.31 for the period from 2000 through today. However, volatility now reads 7.90 for the period from January 2008 through today - the period of the crisis.
  
• Profit margins in Manufacturing also continued to decline in June, with the rate of decline moderating even more than in Services from-21.19 in May to -16.22 in June. Profit margins rates of decline are now more moderate than 12mo MA of 20.6. Q1 2011 profit margins rate of decline was -20.8 with Q2 2011 declines steeper at -23.6.
• Profit margins in Manufacturing are now on the downward trend for 28 months in a row. 2010 H1 average decline was -19.3 and 2009 H1 average was -3.7, implying that 2011 deterioration in profit margins is steeper than in both previous years.
• Profit margins volatility also rose in manufacturing during the crisis. Historical standard deviation for profit margin indicator in manufacturing stands at 6.95, while since 2000 through today volatility is 6.98. However volatility since January 2008 is 8.92.