In Part 1 of this paper, data were presented which contradicted the widely-held view that Ireland experienced a sharp loss of export competitiveness vis-à-vis its main trading partners over the period 2000-2007.  Ireland actually increased its share of total world exports (albeit marginally) in this period and did better in this respect than seven of Ireland’s twelve main trading partners.  In terms of share of global services exports, Ireland’s powerful performance easily surpassed all twelve main trading partners and, while Ireland’s performance with respect to merchandise exports was poor, it was still superior to that of Ireland’s two leading trading partners, the UK and USA, as well as Japan.

Part 2 of this paper interrogates two further myths relating to Ireland’s export competitiveness i.e. that rising unit labour costs negatively impacted on Ireland’s export competitiveness in the period after 2000 (thereby bringing about Ireland’s postulated loss of competitiveness in this period) and that labour costs are a key determinant of Ireland’s export competitiveness.  However, initially some explanations for Ireland’s loss of market share in global merchandise exports in the 2000s are advanced  – explanations which have nothing to do with trends in labour costs within Ireland.

Explaining Ireland’s loss of market share in merchandise exports 2000-2007

Some special factors contributed to Ireland’s loss of share of global merchandise exports between 2000-2007.  One of these was the rapid rise in the prices of minerals and fuels in this period which boosted the global market share of countries with high relative endowments in these natural resources at the expense of countries, like Ireland, with low natural resource endowments.  Between 2000-2007, the share of global merchandise exports taken by Fuels & Mining Products rose from 13.3% to 19.0%.  By contrast, the share of Ireland’s merchandise exports accounted for by Fuels & Mining Products was just 1.3% in 2000 and 2.3% in 2007.  Thus, even if there had been no change in the technical efficiency of the Irish export sector during the period, Ireland would still have lost market share due to its relative underdowment in a subsector with a substantial and rapidly rising global market share.

A second contributory factor to Ireland’s loss of global market share in merchandise exports was the knock-on impact of the collapse in 2000-2001 of the so-called “dot.com bubble” on the Office and Telecommunications Equipment (OTE) sector, upon which Ireland has had a heavy export dependence.  This was reflected in a fall in the share of global merchandise exports taken by the OTE sector from 15.0% in 2000 to 10.9% in 2007.  This sector alone accounted for one third of Ireland’s merchandise exports in 2000, according to WTO data.  In this case Ireland was penalised for an overdependence on a subsector which experienced a sharp reduction in global demand for reasons over which the Irish economy had no control.

A third factor which contributed significantly to Ireland’s loss of global market share in merchandise exports in the period 2000-2007 was the emergence, in this period, of China as a major competitor, again in the OTE sector. In 2000 China accounted for 4.5% of global OTE exports, but by 2007 this had jumped to 22.9%.  Within the OTE sector, China’s rise was particularly pronounced in the Data Processing & Office Equipment (DPOE) subsector (global share up from 5.0% to 30.3% between 2000-2007) which accounted for over two thirds of Ireland’s OTE exports in 2000.

The key factor here was China’s very low cost base in what is a highly-competitive industry, a factor which affected most western economies as much as – if not more than – Ireland.  Table 4 shows the global share of DPOE exports taken by Ireland and its main trading partners in 2000 & 2007.  This shows that the loss of global market share by Ireland’s main trading partners (excluding China) combined was almost as large as Ireland’s, with seven of the eleven comparator countries experiencing more sharp declines than Ireland while there is evidence to suggest that, in the case of two of the three countries which show a rise in market share (Germany and the Netherlands), this was at least partially due to re-exports from these countries of products originally imported from China.  Thus, Ireland’s loss of market share in this sector cannot be attributed to rising costs (including labour costs) relative to Ireland’s main trading partners, but to the emergence of a new competitor with a cost base profoudnly below those prevailing in the western economies in general.

The role of labour costs in Ireland’s export competitiveness

Our argument, therefore, is that, to a significant extent, Ireland’s loss of global market share in merchandise exports in the period 2000-2007 was due to external developments over which Ireland had no control but which had a disproportionate impact on Ireland’s export performance.  By contrast, the evidence suggests that changes in relative labour costs – the factor identified by most economic commentators as being responsible for Ireland’s purported loss of export competitiveness – had little if anything to do with Ireland’s export performance over the last decade.  While most public pronouncements by economic commentators do not state the sources they are using for their assertions that Ireland has been losing export competitiveness and that this is attributable to rising labour costs, it is assumed that the key sources concerned are publications of the ESRI and the Central Bank which routinely repeat these assertions.

Table 4: Share of global exports of Data Processing & Office Equipment (%)

2000 2007 2007 as % 2000
Belgium 1.19 0.93 77.9
France 2.73 1.57 57.4
Germany 4.63 5.95 128.6
Italy 0.88 0.43 49.0
Japan 9.46 4.54 48.0
Netherlands 7.41 8.83 119.2
Spain 0.50 0.25 49.3
Sweden 0.20 0.48 234.0
Switzerland 0.37 0.19 52.4
United Kingdom 5.92 2.45 41.3
United States 15.48 8.64 55.8
11 Trading Partners 48.78 34.26 70.2
Ireland 4.70 3.15 67.0
China 5.01 30.28 604.5

—-

Source: WTO

An interesting fact about most of these pronouncements – and indeed many academic papers on the topic – is the absence of any factual data to demonstrate that Ireland has actually been losing competitiveness.  Most commentators appear happy to assume that if Irish labour costs are rising faster than our competitors, then Ireland must be losing competitiveness.

One commentator who has been stating his sources on labour cost trends and who has been presenting data on trade performance to demonstrate what he considers to be the consequences of these trends is Garret Fitzgerald.  In his Irish Times columns, Fitzgerald has shown that unit labour costs (costs relative to productivity) in the Irish economy rose relative to the USA and Western Europe during the 2000s while growth of industrial exports virtually ceased.  From this he draws the conclusion that the latter phenomenon was a direct consequence of the former.

Apart from the fact that, as shown above, factors other than labour cost trends had a significant negative impact on Ireland’s merchandise export performance in this period, there are serious flaws in Fitzgerald’s analysis.  The labour cost indicator which he uses is the OECD’s Total Economy Unit Labour Cost (ULC), which is based on labour costs throughout the economy and not just in exporting firms (direct employment in export activities accounted for only 12% of Ireland’s total employed workforce in 2000).  Apart from the dubious merit of calculating a national average from such a disparate array of economic sectors (many of which are effectively sheltered from international competition), the assumption (implicit in Fitzgerald’s argument) that unit labour cost trends in the export sector reflect those in the economy at large does not hold water.  Fitzgerald makes no reference to the OECD data on Manufacturing ULCs, which provide a more accurate view of labour cost trends in the Irish export sector (as some 80% of Irish manufacturing output is exported).  These data show that not only have Manufacturing ULCs shown an overall downward trend in the period 2000-2007, but they have been falling relative to most of Ireland’s main trading partners.

Figure 3: Trends in Total Economy & Manufacturing Unit Labour Costs (ULCs) relative to the OECD average, 2000-2007 [2000 = 100]

Source: OECD

Figure 3 shows how Ireland fared relative to the OECD average with respect to both Total Economy and Manufacturing ULCs between 2000-2007.  This shows Total Economy ULCs falling slightly between 2000-2003 and then rising consistently between 2002-2007.  Manufacturing ULCs fell sharply between 2000-2002, then rose again to the OECD average in 2005 & 2006 before falling again in 2007.  The overall trend lines show a diverging trend over time, with Total Economy ULCs rising and Manufacturing ULCs falling over the period.

Table 5 shows the change between 2000 and 2007 for both labour cost measures for Ireland and ten of Ireland’s twelve main trading partners (no data available for China and no Manufacturing ULC data available for Switzerland).  This shows that while Ireland’s Total Economy ULC rose more quickly than any of the other countries, Ireland’s Manufacturing ULC actually fell over the period whereas they rose in six of the ten comparator countries, while of the four which experienced a fall in Manufacturing ULCs, only Japan’s and Sweden’s rate of fall exceeded Ireland’s.

Table 5: Change in Total Economy & Manufacturing Unit Labour Costs – 2000-2007

Total Economy Manufacturing
2000 2007 2000 2007
Belgium 100.0 113.6 100.0 103.7
France 100.0 114.2 100.0 100.8
Germany 100.0 97.8 100.0 91.0
Italy 100.0 120.8 100.0 120.4
Japan 100.0 85.8 100.0 77.8
Netherlands 100.0 114.9 100.0 102.3
Spain 100.0 124.0 100.0 119.4
Sweden 100.0 110.3 100.0 87.1
United Kingdom 100.0 119.5 100.0 107.1
United States 100.0 115.5 100.0 92.7
Ireland 100.0 125.4 100.0 90.5

Source: OECD

Comparative data on Unit Labour Cost trends in export services are not available, but the relevant trends for Ireland can be estimated from Forfás surveys of the economic impact of state-aided economic activities.  These show that Unit Labour Costs in Export Services (pay costs per worker divided by value added per worker) fell by 26.6% between 2000-2007 in the foreign-owned export services sector (which accounted for 94% of services exports in 2000).

Fitzgerald’s assertion, to the effect that Irish manufacturing exports were negatively affected in the period after 2000 by rising labour costs is therefore based on a false premise, since relative labour costs in the export manufacturing sector actually fell in this period.  It has been pointed out by some commentators that real unit labour costs in manufacturing are understated due to the effect of transfer pricing (whereby TNCs operating in Ireland overstate the value of their output in order to avail of Ireland’s low corporation tax rate) in overstating value added.  However, this refers to the absolute value of unit labour costs – there is no obvious reason why this should impact on the rate of change of unit labour costs over time, which is our concern here.

A more obvious flaw in Fitzgerald’s analysis refers to his treatment of services exports.  While acknowledging their strong growth in the 2000s, he fails to explain why the purported rise in Irish labour costs should not have impacted on these in the same way as he suggests they impacted on manufacturing.  To put this in perspective, according to Forfás survey data, in 2007 labour costs accounted for 9.9% of total costs in the foreign services export sector compared with 11.1% of total costs in the foreign manufacturing export sector (i.e. there was no major difference in the relative cost of labour in both sectors).

This last statistic brings us to what is possibly the biggest flaw of all in the argument of those who assert that labour costs are the key ingredient in export competitiveness i.e. labour costs in fact only account for a relatively small proportion of costs in Irish export activities.  To explore this further, Table 6 presents some relevant data from the Census of Industrial Production (CIP) relating to Ireland’s main export manufacturing sectors for the period 2000-2006 (with the exception of the Food & Beverages sector in which natural resource endowment has a substantial impact on competitiveness).  Due to changes in coverage, the CIP data for 2007 are not directly comparable to previous years and have been excluded from this table.

Garret Fitzgerald in his Irish Times articles has placed much emphasis on the job losses sustained in the 2000s due, in his view, to rising labour costs.  One might expect, therefore, that sectors in which labour accounted for a high proportion of costs to have the weakest employment performance over the period, either due to loss of market share or through substitution of capital for labour.  In fact, the data in Table 6 show no relationship whatever between proportionate labour costs and rate of employment change (Pearson’s R = -0.03).  Thus, for example, the Medical, Precision & Optical Instruments sector, in which labour costs accounted for 25% of total costs in 2000, experienced 41% employment growth over the period, whereas the Electrical Machinery & Apparatus sector, in which labour costs were almost equally important (23%) experienced a decline in employment of 50%.  A distinctive feature of the table is the major differences between sectors in employment performance, which demonstrates the futility of making industry-wide statements about employment trends and their causes.

A further distinctive feature of the table is the relatively low proportion of costs accounted for by labour costs in all sectors, ranging from just 4.1% for Office Machinery & Computers to 28.7% for Machinery & Equipment, with an overall average for the nine sectors of 10.8%.  It may be noted that the very low labour cost share in the Office Machinery & Computers sector did not spare that sector from serious job losses in the 2000s, indicating that China’s meteoric rise in this highly competitive sector is attributable to low costs in general, and not merely low labour costs.

Table 6: Labour costs and employment change, main export sectors – 2000-2006

Sector Export % 2000 Labour % Input Costs 2000 Empl Change 2000-2006 (%)
Reproduction of Computer Media 97.2 21.1 -27.5
Organic Chemicals 99.8 8.2 +23.3
Pharmaceuticals 96.0 16.4 +35.3
Perfumes & Toilet Preparations 98.5 27.8 -28.5
Machinery & Equipment 71.6 28.7 -22.7
Office Machinery & Computers 86.0 4.1 -36.4
Electrical Machinery & Apparatus 83.1 23.2 -50.3
Communications Equipment & Electronic Components 93.3 13.2 -43.4
Medical, Precision & Optical Instruments 95.4 25.0 +41.2
Total 92.7 10.8 -14.8

Source: Census of Industrial Production (CIP)

Economic commentators appear to be unaware of the low share of labour costs in the total cost structure of Ireland’s exporting firms, frequently pointing to the high share of GNP accounted for by labour remuneration as indicative of the importance of labour costs to competitiveness.  Thus, in a recent article dealing specificially with the measurement of labour cost competitiveness by Central Bank economist Derry O’Brien in that institution’s prestigious Quarterly Bulletin, the following passage appears on page 105: “The largest costs incurred by firms are typically labour costs”.  This is presented as a self-evident truth with no supporting data, yet as is obvious from the data presented above, labour costs account for only about ten per cent of total costs in the export activities which were the very focus of the paper in question.  This means that the five per cent reduction in wages which the ESRI’s John Fitzgerald sees as being necessary to restore Ireland’s competitiveness (see quotation above) would have the effect of reducing the overall costs of the average exporting firm by less than one half of one per cent – hardly likely to have a major impact on competitiveness.

Conclusion

This part of the paper has shown that, while economy-wide unit labour costs in Ireland did rise more rapidly than in Ireland’s main trading partners in the period 2000-2007, unit labour costs in both manufacturing and export services moved in the opposite direction.  While this was associated with a rapid rise in services exports (and might have been in the case of manufacturing exports were it not for the intrusion of several extraneous factors which had a particularly negative impact on Ireland’s merchandise exports), this should not be construed as establishing a causal link between trends in labour costs and trends in export performance.  An examination of the cost structure of Ireland’s main manufacturing export sectors revealed that not only was there was no link between the proportion of total costs accounted for by labour costs and employment change in these sectors, but that labour costs in general account for a low proportion of costs in the sectors in question (as is also the case with export services).

The third and final part of this paper reviews a number of alternative approaches to assessing national export competitiveness which are much more complex and sophisticated than the simplistic – and erroneous – reduction of competitiveness to international differences in unit labour costs.  These alternative approaches generally portray Ireland’s competitiveness position in a much more favourable light than has become the norm in public discourse of the topic in Ireland.

Reference

O’Brien, Derry (2010) Measuring Ireland’s Price and Labour Cost Competitiveness Central Bank Quarterly Bulletin, 2010 No 1, 99-113.

TO BE CONTINUED…

Proinnsias Breathnach

One of the more disquieting aspects of the current economic crisis is the almost complete absence of any debate on the question of Ireland’s ability to resume its economic dynamism via export-led growth whenever the global economy recovers. There is almost complete unanimity among Irish economists in support of the following propositions:
• Export competitiveness is a function of Ireland’s level of unit labour costs (labour costs relative to productivity) vis-à-vis those countries with which Ireland competes in international markets (usually seen as being equated with Ireland’s ‘main trading partners’).
• Ireland’s labour costs became increasingly uncompetitive in the decade since 2000 due to wage levels rising too fast relative to productivity growth.
• This led to a massive loss of export competitiveness over the last decade.
• As a result, economic growth in Ireland during that decade was based almost entirely on the construction-led consumer boom whose collapse has resulted in the current sorry economic mess.
• Ireland therefore needs to reduce wage costs in order to restore competitiveness so that Ireland will be in a position to exploit the next upturn in the global economy through resumed export growth.

What is remarkable about the unanimity surrounding these propositions is that they are almost entirely unsupported by the available evidence. In what follows, evidence is presented in support of two key arguments which contradict the received wisdom as outlined above:
• That the widespread view that Ireland experienced a major loss of competitiveness over the last decade vis-à-vis its main trading partners is, for the most part, erroneous.
• Following from this, that labour costs are a factor of, at best, minor importance in determining the competitiveness of export businesses operating in Ireland.

Let us begin by examining the data relating to Ireland’s export performance over the last decade, drawn mainly from the OECD and World Trade Organisation (WTO) websites.
Between 1998-2003 there was a substantial acceleration in the rate of inflation in Ireland which has generally been attributed to a major expansion in public spending in this period. Between end-1997 and end-2003, Ireland’s nominal GDP, after allowing for real volume growth, grew by one third, three times the rate for the Eurozone and the USA, and over twice the rate for the UK. This was reflected in a sharp rise in Ireland’s economy-wide relative unit labour costs (output relative to pay) which rose at twice the rate for the Eurozone and 1.5 times the rate for the USA.

The fact that Ireland’s rate of export growth fell dramatically in the early 2000s (to as low as one half of one per cent in 2003) led to the easy – and widely-held – conclusion that Ireland experienced a massive fall in export competitivness vis-à-vis our main trading partners in this period and that this was attributable to the rise in relative labour costs which coincided with the sharp fall in the rate of export growth. However, this conclusion is simply not supported by the available data. Between end-2000 and end-2007, Ireland’s exports of goods and services grew by 48.3% in volume terms, ahead not just of the overall rate of growth for the OECD (39%) but for eight of Ireland’s ten leading trading partners (the exceptions being Germany and China).

The star of the show in this respect were services exports, which more than doubled (to 2.6%) their share of global services exports in this period. If one excludes the relatively low-tech travel and transport subsectors, Ireland’s share of global services exports reached 4.6% in 2007. This rapid growth embraced a range of sectors including software services, insurance and other financial services, merchanting and operational leasing. The picture was not so positive as regards merchandise exports, where Ireland’s share of global exports fell by over one quarter over the period. However, this overall decline masks significant intersectoral differences – while Ireland’s share of global exports of office and telecoms equipment fell by 40% over the period, our share of pharmaceuticals exports rose by almost 20%.

The sharp fall in Ireland’s exports of electronics hardware which commenced in 2001 can be attributed largely to two factors. The first of these was the dotcom collapse of 2000 which induced a substantial fall in the demand for electronics products – in 2001 global exports of office and telecom equipment were 13% lower in nominal value terms than a year previously, and did not regain the 2000 level again until 2004. The second factor in the fall-off of Irish electronics exports was the emergence of China as a major competitor in the sector (China’s share of global exports of office and telecoms equipment jumped from 4.5% to 23% between 2000-2007). Ireland was not alone in experiencing market share loss to China’s ultra-low production costs in routine electronics production – in fact, all of Ireland’s nine main trading partners (other than China) lost market share to China in the office & telecoms equipment sector except Germany and Netherlands, and there is evidence to suggest that the export growth recorded in the latter two cases consisted of re-exports of products which originated in China.

Because of the very high proportion of Irish merchandise exports accounted for by electronics products (40% in 2000), the sharp contraction in exports from this sector in the early 2000s had a substantial impact on Ireland’s overall export performance in this period. However, this contraction had nothing to do with rising labour costs vis-à-vis our main trading partners. This is borne out by the relatively strong export performance of our other main manufacturing export sector (pharmaceuticals) and the powerful export performance of a range of services sectors in this period. It is further borne out by the fact that, while Ireland’s unit labour costs rose by 80 per cent more than those of the USA between 2000-2007, Ireland’s exports grew by 60% more than those of the USA (in volume terms) in the same period.

Of course, while economy-wide trends in unit labour costs are widely cited as evidence of Ireland’s declining export competitiveness, they bear little relationship to labour cost trends in exporting industries (especially where, as in Ireland, most of these sectors have an ‘enclave’ nature with few links with the rest of the economy). In the Irish case, calculation of unit labour costs for manufacturing industry may be significantly distorted for some sectors by transfer price manipulation which exaggerates value added in Ireland. However, trends over time may have more reliability than fixed point calculations. Thus, whereas Ireland’s ‘total economy’ unit labour costs rose by over 10% relative to the OECD average between 2000-2007, unit labour costs in manufacturing (most of whose output is exported) actually fell by 9% relative to the OECD average in this period. However, this again is an overall average which does not allow for varying trends between manufacturing sectors, and when calculated for individual sectors is a poor predictor of export performance. Thus, unit labour costs (labour costs as a proportion of gross value added, calculcated as a three-year average) rose by 20% in the pharmaceuticals sector and fell by 9% in the office machinery and computers sector between 2000-2005. Yet it was the former which experienced significant growth in market share in the period while the latter sector lost out substantially.

Overall, therefore, we contend that the extraordinarily high levels of export growth which Ireland enjoyed in the 1990s were unsustainable in the long run, and the fall-off observed in the 2000s to a considerable extent reflects an adjustment to more ‘normal’ levels of growth. Even in this period, Ireland’s export growth outstripped that of most of our main trading partners, and, in the one major sector where Ireland lost significant market share (electronics hardware), we argue this was due to specific circumstances which were largely beyond Ireland’s control.

We find no evidence that labour costs had any significant bearing on Ireland’s export performance in the 2000s (apart from the exceptional case of very low Chinese costs in electronics). Rising general labour costs relative to our main trading partners have not prevented very rapid growth in services exports or superior export performance on Ireland’s part compared with countries whose labour costs have fallen substantially relative to Ireland’s. Furthermore, export performance in the main export sectors in manufacturing has been inversely related to unit labour cost trends in these sectors.

The preoccupation of many economists with labour costs as the key to export competitiveness is difficult to fathom given the minor role played by such costs in the overall cost structure of Ireland’s main export sectors. In manufacturing as a whole, labour costs amounted for just 15.5% of total direct costs of Irish manufacturing plants in 2006 (the latest year for which the relevant data are available in the Census of Industrial Production). This figure varies intersectorally, and ironically was lowest of all (at just 3%) in the office machinery and computers sector, which has experienced substantial loss of market share, while it stands at 18% and 22%, respectively, in pharmaceuticals and medical devices, our two most successful manufacturing export sectors in recent times. According to Forfás data, labour costs in foreign-owned service export firms (which account for the great bulk of service exports) were, at 10% of total input costs, even lower than in export manufacturing.

It is noteworthy that labour costs are given a very minor role in the computation of the authoritative Global Competitiveness Index compiled annually by the World Economic Forum (WEF) under the direction of Harvard University competitiveness guru Michael Porter. This index is compiled from no less than 113 indicators which are given different weightings depending on a country’s level of development. For countries at Ireland’s level of development, the key to competitive success is seen by the WEF as the ability to produce new and different products employing cutting-edge production processes. In the system of weightings applied to this group of countries, labour cost factors account for just 1.7% of the total value of the competitiveness index. Based on these criteria, Ireland has actually been moving up the WEF’s competitiveness league table, from 30th position in 2002 to 22nd in 2009.

This positive view of Ireland’s competitiveness performance is shared by other influential international league tables. Thus, in its report for 2010, the World Bank’s annual Doing Business survey placed Ireland 7th out of 183 countries in terms of ease of doing business (labour costs are not among the 40 variables used in compiling the table). Similarly, IBM’s Global Location annual report for 2009 rated Ireland as the most successful country in the world (up from 10th place in 2008) for attracting foreign investment, measured in terms of jobs per 100,000 population. This undoubtedly is related to the profitability of foreign firms operating in Ireland. In 2007, the average rate of return on investment by US firms in Ireland, at 22.5%, was up 3.5 percentage points since 2000 and was the fourth highest in the world after China, India and Singapore.

It is also noticeable that spokespersons for foreign firms operating in Ireland rarely make reference to labour costs when reflecting on their operations here. This was brought home in the recent address by former Intel chief Craig Barrett to a meeting of the Royal Irish Academy. In this address, Barrett specifically rejected the notion that Ireland could strengthen its competitiveness on the basis of cutting costs, emphasising instead the need to expand investment in the development of high-tech activities and in enhancing educational provision in science and mathematics. To be competitive, we need to be smart and innovative, not cheap, according to Barrett..

There is a real danger that, encouraged by the economists’ chorus on the need to cut costs as the way to get us out of our current economic troubles, the government will actually undermine Ireland’s long-term competitive position through cutbacks in research and education (in fact, such cutbacks are already being imposed). We desperately need to substitute serious evidence-based analysis for the rote incantation of inherited mantras which currently passes for expert economic advice in the realm of competitiveness policy in this country.

Proinnsias Breathnach