Irish Indigenous Companies

Abstract

There was a substantial improvement in the growth of indigenous manufacturing in the late 1980s and early 1990s, but its performance became unsatisfactory later. In indigenous services, there was also a significant improvement and the trends mostly remained better for longer. New industrial policy measures were partly responsible for the improvement. There was success at first in promoting the growth of SMEs, particularly in the high-tech and medium–high-tech sectors, and in enhancing R&D and innovation. However, the growth momentum tended to weaken later as many successful companies were taken over by foreign MNCs. The indigenous contribution to net foreign earnings was much greater than it appeared to be when seen in terms of exports, and this was especially true for indigenous services, but the indigenous contribution to growth over the whole course of the boom was still significantly less than that of foreign-owned companies. However, the indigenous contribution was particularly important in the late 1980s and again in the 2000s. The improvement in indigenous industry was probably the key change that started the boom in the late 1980s.

This chapter focuses on the role of Irish-owned or indigenous companies during the boom and Chapter 6 deals with the role of foreign-owned multinational companies. Of course, some sectors of the economy were almost entirely indigenous, including agriculture, non-market services and construction, and consequently the indigenous/foreign distinction is of little relevance for those sectors. Therefore, this chapter and the next concentrate on manufacturing and market services. These are the sectors where foreign-owned MNCs did have a substantial and distinctive role alongside indigenous companies, and they are also the sectors that were particularly influential in driving economic growth since they were the source of most exports and net foreign earnings, as seen in Chapter 4. In this chapter, Sect. 5.1 deals with indigenous firms in manufacturing and Sect. 5.2 goes on to look at indigenous firms in market services.

5.1 Growth and Development of Indigenous Manufacturing

It was seen in Chapter 2 that, before the boom began in the late 1980s, much of Irish indigenous manufacturing had been experiencing persistent difficulties in coping with foreign competition in the domestic market after the introduction of free trade in the 1960s and 1970s. This existing weakness was compounded by a sharp deterioration in domestic demand conditions in the 1980s, resulting in a slump in indigenous industry. Against that background, there was a considerable improvement in the growth and development of indigenous manufacturing beginning in the late 1980s, although its performance was somewhat uneven with some strong points and some weak points. The growth of indigenous manufacturing looked quite good in some respects compared to other countries, although its performance was overshadowed for about half of the boom by much stronger growth in foreign-owned industry in Ireland.

5.1.1 Employment Trends

Table 5.1 shows employment trends in Irish indigenous and foreign-owned manufacturing in 1980–2006. Total manufacturing employment declined considerably during the recession of the 1980s with a particularly sharp decline in the indigenous industry which was reduced from 61.8% of the total in 1980 to 57.4% in 1988. Then from 1988 to 2000, there was strong growth in manufacturing employment. This growth was faster in foreign-owned industry than in indigenous industry so the indigenous share of the total declined further. Nevertheless, the employment growth seen in the indigenous industry was substantial, amounting to 1.9% p.a. in 1988–2000, which was in marked contrast to earlier in the 1980s and indeed was quite unprecedented in twentieth-century Ireland under free trade conditions. The decline that followed in 2000–2006 affected both foreign-owned and indigenous manufacturing employment, although the decline was a little less severe among the indigenous companies which slightly increased their share of the total over that period.¹

Table 5.1 Employment in Irish indigenous and foreign-owned manufacturing, 1980–2006

Before the boom, the employment trend in indigenous manufacturing had been exceptionally weak compared to international experience among developed countries but after 1988 it became exceptionally strong compared to other countries. Thus in 1980–1988 employment declined by 3.2% p.a. in indigenous manufacturing and by 2.2% p.a. in total manufacturing in Ireland while it declined by 1.6% p.a. in the EU-15 and by 0.7% p.a. in the USA. It can be seen in Fig. 5.1 that the growth of indigenous manufacturing employment which followed in 1988–2000 contrasted with the still declining trend in the EU and USA. Employment in Irish indigenous industry was clearly growing more slowly than employment in foreign-owned industry in Ireland, but it was growing quite rapidly by most other comparisons.

Fig. 5.1

(Source: Census of Industrial Production for Irish data, with a few minor adjustments as explained in Appendix. EU-15 data from EUKLEMS database [euklems.net]. USA from OECD’s STAN database. Creative Commons Attribution BY 4.0)

Manufacturing employment index (1988 = 100), Ireland total, Irish indigenous, EU-15 and USA, 1988–2007

After 2000, there was a substantial decline in indigenous manufacturing employment. However, Fig. 5.1 shows that this trend was quite similar to international experience. By 2006, employment in indigenous industry was still 8% above its 1988 level whereas industrial employment was 19 or 20% below the 1988 level in the EU and the USA.

As regards sectoral trends within Irish indigenous manufacturing, it is necessary to consider 1988–1990 separately from the later years because of a change in the industry classification system after 1990 (as discussed in Appendix). In 1988–1990 total indigenous manufacturing employment grew by 1.3% p.a., compared to 1.0% p.a. for the EU and −0.7% p.a. for the USA. About three-quarters of the sectors within indigenous industry had growing employment in that period but it was noticeable that the growth rates were well above average in the “high-technology” sectors—pharmaceuticals (5.6% p.a.), office & data processing machinery (35.3% p.a.), electrical engineering (6.0% p.a.) and instrument engineering (4.1% p.a.). However, employment in most of those sectors, with the exception of electrical engineering, was still very small in absolute terms being numbered in hundreds rather than thousands (O’Malley 1998).

Table 5.2 shows subsequent trends in indigenous manufacturing employment by sector in 1991–2000 and 2000–2006. In 1991–2000, when total indigenous manufacturing employment growth was quite rapid at 2.1% p.a., there was growth in all sectors apart from two. Nearly all the sectors that are classified by Eurostat as “high-technology” sectors grew exceptionally fast—communication equipment & technical instruments (14.2% p.a.), pharmaceutical products (8.3% p.a.) and office machinery & computers (7.6% p.a.). In addition, nearly all the sectors that are classified as “medium–high technology” grew at faster than average rates—electrical machinery & apparatus (9.8% p.a.), machinery & equipment (4.5% p.a.) and other chemicals (4.4% p.a.). In fact, the only sector that might possibly be seen as going against the general trend of above-average growth for the higher technology sectors was transport equipment.² Consequently, the composition of indigenous manufacturing employment was shifting significantly towards higher technology sectors (Jacobson and O’Malley 2018).

Table 5.2 Employment growth by sector in Irish indigenous manufacturing, 1991–2006

In 2000–2006 total indigenous manufacturing employment declined. Table 5.2 indicates that there was a reversal of the earlier sectoral trends since the decline tended to be particularly severe among the high-technology and medium–high-technology sectors, with only pharmaceutical products among those sectors continuing to have a stronger than average record. However, these trends in 2000–2006 need to be interpreted with some care. The problem here is that, for the Census of Industrial Production (CIP) data used in Table 5.2, companies’ nationality of ownership is defined according to their nationality in each year. Consequently, if some companies are Irish-owned at the start of a period and are then taken over by new foreign owners during that period, their employment is included in the indigenous category in the initial year but not in the final year. This can result in a weak trend in indigenous employment in that period although there may have been no real competitive or commercial weakness resulting in declining businesses.

In fact, there is evidence that this type of effect was very influential in the indigenous high-technology and medium–high-technology sectors in 2000–2006. In the Forfás Annual Employment Survey, companies’ nationality of ownership was defined according to their latest nationality when responding to the survey, and then presentations of data on past trends in the survey reports applied each company’s latest nationality to all previous years so that changes of nationality did not affect the trends over time. With nationality defined in this way, the Forfás Annual Employment Survey 2008 (Appendix Table 5) indicated that total indigenous manufacturing employment declined by 1.2% p.a. in 2000–2006, which was less than the decline by 2.4% p.a. seen in the CIP data in Table 5.2. Furthermore, the Forfás survey data indicate that employment in the indigenous high-technology and medium–high-technology sectors declined by just 0.5% p.a. in 2000–2006, which was much less than the decline by 5.8% p.a. in the CIP data.³ In the low-technology and medium–low-technology sectors the rate of decline was similar in both data sets, at 1.6% p.a. in the CIP and 1.3% p.a. in the Forfás survey.

Thus, the combination of the two sets of data shows that there were significant net transfers of ownership from Irish to foreign during 2000–2006, with nearly all of these transfers being among the high-technology and medium–high-technology sectors. In the absence of such net transfers of ownership, the overall trend in indigenous manufacturing employment would have looked considerably stronger and the high-technology and medium–high-technology sectors would probably have continued to increase their share of the total. Also, the slight increase in the indigenous share of total manufacturing employment in 2000–2006, which was seen in Table 5.1, would have been greater in the absence of net transfers from Irish to foreign ownership.

Incidentally, the trends in indigenous manufacturing after the boom was over are of some relevance here because they show what happened when net foreign takeovers ceased for a while. In 2008–2014, the comparison between the CIP and the Forfás survey shows no significant net transfers from Irish to foreign ownership, perhaps because that period was dominated by the “great recession” in Ireland and elsewhere so that many multinational companies had to focus more on surviving in very difficult conditions rather than expanding by means of acquisitions. In the absence of net foreign takeovers, the trend of particularly rapid growth among the higher technology sectors resumed in Irish indigenous manufacturing. CIP data show that employment grew by 1.6% p.a. in the indigenous high-technology and medium–high-technology sectors in 2008–2014 despite the major international recession, while it declined by 2.9% p.a. in the low-technology and medium–low-technology sectors.⁴

To see how Irish indigenous employment compared to the EU, Table 5.3 shows indigenous industry’s percentage share of EU-15 employment in each manufacturing sector in 1991, 2000 and 2006. Indigenous industry’s share of total EU-15 manufacturing employment increased substantially from 0.32% in 1991 to 0.44% in 2000 and then declined a little to 0.42% in 2006. The net transfers of ownership from Irish to foreign in 2000–2006 were probably sufficient to account for the decline in the share of EU employment in that period.

Table 5.3 Irish indigenous industry’s share of EU-15 manufacturing employment, by sector, 1991–2006 (%)

At the sectoral level, Irish indigenous industry increased its share of EU employment in 1991–2000 in all manufacturing sectors apart from textiles & textile products and transport equipment. Initially in 1991, indigenous industry was relatively under-represented in the higher technology sectors compared to the EU. For example, indigenous industry had a 0.32% share of total manufacturing employment in the EU-15, but its share of EU-15 employment was a good deal lower in sectors such as pharmaceutical products (0.24%), machinery & equipment (0.14%), electrical machinery & apparatus (0.15%), and communication equipment & technical instruments (0.08%). That situation changed considerably during 1991–2000. Since Irish indigenous employment was growing particularly fast in most of the high-technology and medium–high-technology sectors, the Irish indigenous share of EU employment rose very rapidly in many of those sectors. In all of those sectors combined, indigenous industry more than doubled its share of EU employment from 0.14% in 1991 to 0.33% by 2000. The Irish indigenous share of EU employment in the low-technology and medium–low-technology sectors also increased in the same period but the rate of increase was slower, from 0.39% in 1991 to 0.49% in 2000.

As regards 2000–2006, if we focus on the CIP data that are used for Table 5.3, Irish indigenous industry’s share of EU manufacturing employment declined from 0.44% to 0.42%. The decline was particularly severe among the high-technology and medium–high-technology sectors. In those sectors combined indigenous industry’s share of EU employment dropped from 0.33% to 0.25%. However, if we use the Forfás Annual Employment Survey data, as discussed above, indigenous industry’s share of total EU-15 manufacturing employment held steady at 0.39% in both 2000 and 2006, while its share of EU employment in the high-technology and medium–high-technology industries increased slightly from 0.25% in 2000 to 0.26% in 2006.

Thus, the combination of the two data sets shows again that there were quite significant net transfers of ownership from Irish to foreign during 2000–2006, with these transfers being heavily concentrated among the high-technology and medium–high-technology sectors. In the absence of such net transfers of ownership, there would have been little or no change in the Irish indigenous share of total EU manufacturing employment while the indigenous share of the EU high-technology and medium–high-technology sectors could have increased a little.

5.1.2 Output Trends

Trends in the output of Irish indigenous manufacturing were generally consistent with its employment trends. Thus, there was a marked improvement in output growth in indigenous manufacturing beginning in the late 1980s.

Unfortunately, the data on this are not ideal because there is no output data series for indigenous industry in constant prices, which would show trends in the volume of Irish indigenous industrial production. Instead, published data on the volume of production in “traditional” and “modern” manufacturing have often been seen as proxy measures for data on Irish indigenous and foreign-owned industry respectively, because the “traditional” sectors were predominantly Irish-owned while the “modern” sectors were largely foreign-owned.⁵ Before the boom, the volume of production in traditional manufacturing was growing very slowly at just 1.0% p.a. in 1982–1987 but its growth then accelerated to 3.7% p.a. in 1987–1995 (O’Malley 1998).

However, traditional manufacturing is by no means perfect as a representative of indigenous manufacturing since about one-third of the output of foreign-owned manufacturing firms was in traditional manufacturing in the early 1990s, while almost 10% of the output of Irish indigenous manufacturing firms was in modern manufacturing.

As an alternative, O’Malley (1998) presented estimates of the output of indigenous manufacturing measured in constant prices in 1985–1995. These estimates were derived by constructing an “indigenous” manufacturing price index, based on the official price indices for each individual manufacturing sector, combined together in accordance with their weighting in Irish indigenous industry. According to these estimates the volume of production in indigenous manufacturing increased by just 0.6% p.a. before the boom in 1985–1987 and then grew at a rate of 4.0% p.a. in 1987–1995. Thus, the acceleration in indigenous output growth according to this estimate was somewhat greater than the acceleration in the growth of “traditional” industry.

Whether one goes by this estimate of 4.0% p.a. growth in 1987–1995, or the figure of 3.7% p.a. for traditional industry, this growth was a good deal slower than the 9.9% p.a. growth rate for all of industry in Ireland in the same period. Nevertheless, it was significantly faster than the growth rate of industry in the OECD, at 2.0% p.a., or the EU, at 1.7% p.a., whereas the growth of Irish indigenous or traditional industry had been weak compared to those countries in the years before the boom (O’Malley 1998).

Table 5.4 uses output data valued at current prices to show Irish indigenous industry’s percentage share of the value of EU-15 gross output in each manufacturing sector in 1991, 2000 and 2006. For the most part, the trends in indigenous industry’s share of EU manufacturing output were quite similar to the trends seen above in its share of EU manufacturing employment. Irish indigenous industry’s share of total EU manufacturing gross output increased from 0.34% in 1991 to 0.4% in 2000 and then declined to 0.37% in 2006.

Table 5.4 Irish indigenous industry’s share of EU-15 manufacturing gross output, by sector, 1991–2006 (%)

At the sectoral level, indigenous industry increased its share of EU output in 1991–2000 in all manufacturing sectors apart from textiles & textile products and transport equipment. Initially in 1991, Irish indigenous industry was relatively under-represented in the higher technology sectors compared to the EU. Thus, indigenous industry had a 0.34% share of total manufacturing output in the EU-15 in 1991 but its share of EU-15 output was much lower in chemicals (0.15%), machinery & equipment (0.09%) and electrical & optical equipment (0.09%). That situation changed considerably during 1991–2000. Irish indigenous output was growing particularly fast in the high-technology and medium–high-technology sectors and consequently the indigenous share of EU output rose very rapidly in those sectors. The indigenous share of EU output increased from 0.15% to 0.19% in chemicals, from 0.09 to 0.17% in machinery & equipment, and from 0.09 to 0.28% in electrical & optical equipment.

As regards 2000–2006, Table 5.4 indicates that Irish indigenous industry’s share of EU manufacturing output declined from 0.4% to 0.37%, including declines in all the high-technology and medium–high-technology sectors. However, given the general similarity of the employment trends and output trends, it can be assumed that, as in the case of employment, these declines in share of output probably reflected net transfers from Irish to foreign ownership rather than real weaknesses involving relatively declining businesses.

To conclude on output trends in Irish indigenous industry, these trends were generally consistent with the employment trends in a number of important respects. There was an acceleration in growth starting in the late 1980s, following significant weakness before then. The growth in Irish indigenous industry from the late 1980s onwards was distinctly slower than in foreign-owned industry in Ireland. Nevertheless, the growth in indigenous industry was fast growth by international standards, after being slow by international standards before the boom. There was a marked shift in the sectoral composition of Irish indigenous industry away from the lower technology sectors and towards the higher technology sectors, and this shift was relatively strong by international standards as shown by rapidly rising shares of EU employment and output in the higher technology sectors.

From about 2000 onwards the growth of indigenous industry weakened a good deal and its shares of EU employment and output declined, particularly in the higher technology sectors. However, much of this weakness was caused by net transfers of ownership of companies from Irish to foreign ownership. In the absence of such net transfers of ownership, the Irish indigenous share of EU manufacturing would probably have held up quite well while the indigenous share of the EU high-technology and medium–high-technology sectors could have increased a little.

5.1.3 Exports and Net Foreign Earnings

Regular official data on the exports of Irish-owned manufacturing first became available in the CIP in 1986. However, some earlier survey data on new foreign-owned grant-aided industry make it possible to estimate that exports of industries other than new foreign-owned grant-aided industry amounted to about 26% of their gross output in 1973 and about 27% in 1976 (O’Malley 1989, Table 6.5; 1998). These industries (other than new foreign-owned grant-aided industry) consisted very largely of Irish indigenous firms together with quite a small minority of older foreign-owned firms. Another estimate by Foley (1987) indicated that Irish-owned indigenous industry exported about 31% of its output in 1984.

Against that background, the first official CIP data on indigenous exports showed that indigenous manufacturing exported 26.6% of its gross output in 1986, which was about the same as in 1973 and 1976 but apparently somewhat lower than in 1984. At any rate, it seems reasonably clear that there can have been little or no increase in the export-orientation of indigenous industry over the period 1973–1986. In contrast to that previous experience, exports as a percentage of the output of indigenous manufacturing began to increase immediately after 1986 rising from 26.6% in 1986 to 33.4% by 1990.

In 1986–1990, the value of indigenous manufacturing exports (in current Irish pounds) increased by 12.2% p.a., which was slightly higher than the growth rate of 11.9% p.a. for exports from foreign-owned industry in Ireland. Indigenous manufacturing exports grew relatively fast in that period compared with other countries. Measured in terms of current US dollars, the growth rates of manufacturing exports were 18.3% p.a. for Irish indigenous industry, 15.2% p.a. for the EU-15 and 14.2% p.a. for the OECD.

In the years 1986–1990, the CIP did not show export data by both nationality and sector at the same time. However, it did include data on output by both nationality and sector, so that it is possible to identify the sectors in which the bulk of the output was produced by indigenous firms. It is clear that in nearly all of those sectors, there was an increase in the percentage of output being exported (O’Malley 1998), which suggests that there was probably an increase in export-orientation across most of the sectors in indigenous industry.⁶

In 1991, the CIP introduced the new NACE Rev.1 system for classifying industries so that there was a discontinuity in the data series between 1990 and 1991. In the period 1991–2000, the growth of indigenous manufacturing exports was less impressive than in 1986–1990 in several respects. The value of indigenous manufacturing exports grew by just 5.0% p.a. in 1991–2000 measured in current Irish pounds, and this was much slower than the growth rate of 20.7% p.a. for exports from foreign-owned manufacturing firms in Ireland. Furthermore, there was no increase in the export orientation of indigenous manufacturing in 1991–2000 since 34.8% of its output was exported in 1991 declining a little to 33.2% in 2000. In addition, indigenous industry’s share of the EU’s manufacturing exports declined substantially from 0.407% in 1991 to 0.309% in 2000 (see Table 5.5).

Table 5.5 Irish indigenous industry’s share of EU-13 manufacturing exports, by sector, 1991–2000 (%)

This export performance by indigenous industry was undoubtedly poorer than in 1986–1990, although the absence of any increase in export-orientation was not necessarily a clear sign of weakness given the context of the 1990s. Since there was exceptionally fast economic growth in Ireland and hence exceptionally fast growth of domestic demand, it was possible for the domestic sales of many indigenous firms to grow unusually rapidly. Consequently, it was not necessarily a clear indication of export weakness if their exports grew no faster than their domestic sales.

The large loss in share of EU exports was a more telling indication of export weakness. The weakness here related to some sectors more than others, and it was partly structural in nature, in the sense of being a result of a very unfavourable sectoral composition. Thus, at the start of the period indigenous industrial exports were relatively highly concentrated in sectors that had relatively slow growth for all countries’ exports during the period. As a result, 56% of the total loss in share of EU exports occurred because of the unfavourable sectoral composition of indigenous industrial exports, while the remaining 44% of the loss in export market share occurred because individual indigenous sectors had declining shares of EU exports of their type of products.⁷

Much of the loss in export market share related to the food, drink & tobacco sector, which was a relatively large component of Irish indigenous industry. This sector suffered from a slow-growth environment since the growth of all countries’ exports of food, drink & tobacco products was exceptionally slow, while the Irish indigenous sector also had a substantial loss of export market share as shown in Table 5.5. If we leave that sector aside, Irish indigenous industry’s share of EU exports in other manufacturing sectors was quite mixed. In non-food manufacturing as a whole, the export share increased marginally, from 0.17% in 1991 to 0.171% in 2000, as shown in the last row of Table 5.5. It was probably a considerable improvement over previous experience before the boom for the exports of this broad group of indigenous industries to keep up with the growth of EU exports, particularly since the 1990s was a time when EU exports were growing rapidly.⁸

It is also of interest to note that, as in the case of employment and output, there was relatively strong growth of exports from indigenous firms in the high-technology and medium–high-technology sectors, after starting from a position of being under-represented in those sectors. In Table 5.5 those higher technology sectors are represented by electrical & optical equipment, machinery & equipment, and chemicals & chemical products. Although chemicals & chemical products had some loss in EU export market share in 1991–2000, the other sectors mentioned had large increases in their export market share. Taking all these higher technology sectors as a group, indigenous industry had a relatively low share of EU exports of their products at just 0.116% in 1991 but its share of EU exports then increased to 0.175% by 2000. This group of sectors also increased a good deal in importance in terms of its share of total indigenous manufacturing exports. Its share grew from 11.3% of the total in 1991 to 25.4% by 2000.

After 2000 the growth of indigenous manufacturing exports slowed down, from 5.0% p.a. in 1991–2000 to 3.4% p.a. in 2000–2007, measured in current values. At the same time, the growth of exports from foreign-owned manufacturing firms in Ireland slowed down far more dramatically from 20.7% p.a. in 1991–2000 to 3.6% p.a. in 2000–2007. Thus, exports from the indigenous and foreign-owned groups grew at almost the same rate in 2000–2007. There was only a small increase in the export-orientation of indigenous manufacturing as the proportion of its output being exported rose from 33.2% in 2000 to 34.9% in 2007. This export performance by indigenous manufacturing looked relatively weak by international standards since its share of EU-27 manufacturing exports declined from 0.28% in 2000 to 0.239% in 2007.⁹

It was noted above that in the 1990s the growth of indigenous manufacturing exports was held back by an unfavourable sectoral composition, but that was not the case to any significant extent in 2000–2007 (O’Malley 2013). However, their growth was restrained by the substantial net transfers of ownership of companies from Irish to foreign ownership in the period after 2000. As was pointed out above when discussing trends in employment, these transfers of ownership were very largely concentrated among the high-technology and medium–high-technology sectors. These sectors of indigenous manufacturing were a good deal more highly export-oriented than most other sectors, exporting 48% of their output in 2000 compared with just 30% for the rest of indigenous manufacturing. Consequently, transfers of ownership which primarily affected these higher technology sectors probably had a substantial negative influence on indigenous export trends.

Although it is not possible to present precise figures on this, it seems safe to make two qualitative statements here. First, indigenous manufacturing exports would have grown faster than exports from foreign-owned manufacturing in Ireland in the period after 2000 if there had been no net transfers of ownership. Second, most of the loss in indigenous industry’s share of EU exports in that period can be attributed to transfers of ownership rather than to companies actually losing market share.

As was discussed in Chapter 4, it is useful to look deeper than the figures on exports to examine trends in “net foreign earnings”—meaning the value of exports minus the value of imported inputs contained in the exports minus profit outflows that arise from the production of the exports. Among Irish indigenous companies, net foreign earnings have generally amounted to a substantially higher proportion of the value of exports than they have in the case of foreign-owned companies. This is so partly because (a) indigenous companies have tended to be more concentrated in the sectors where net foreign earnings are a relatively high proportion of exports, and partly because (b) within each individual sector, indigenous companies have generally had higher net foreign earnings as a proportion of exports than their foreign-owned counterparts.

To illustrate the first point (a), at the broad macro-sectoral level in 2005, we estimate that net foreign earnings as a percentage of exports were highest in agriculture at 82%, lower in market services at 62% and lowest in manufacturing at 30%.¹⁰ In this context, indigenous producers accounted for virtually 100% of agricultural exports, 19% of services exports and just 8% of manufacturing exports. At a more disaggregated sectoral level within manufacturing there was a similar pattern. Some sectors had relatively high net foreign earnings as a percentage of exports, such as non-metallic mineral products at 78% and food, drink & tobacco at 58%, while others had particularly low net foreign earnings as a percentage of exports, such as metal & engineering products at 26% and chemical products at 20%. In this context indigenous producers accounted for 55% of exports in non-metallic mineral products and 29% of exports in food, drink & tobacco, compared to just over 4% of exports in metal & engineering products and just over 1% in chemical products.

Table 5.6 illustrates point (b) above—that, within each individual sector, indigenous companies generally had higher net foreign earnings as a proportion of exports than their foreign-owned counterparts. The estimated figures were higher for indigenous companies than for foreign-owned companies in every sector listed in the table. This was partly because there was generally an outflow of profits from the foreign-owned companies in each sector, but not from the indigenous companies. In addition, foreign-owned companies in each sector tended to import a higher proportion of their purchased inputs, whereas indigenous companies in the same sector were more likely to source a higher proportion of their inputs from suppliers in Ireland.

Table 5.6 Estimated net foreign earnings as a percentage of exports, Irish indigenous and foreign-owned by sector, 2005

The bottom row of Table 5.6 shows the aggregate results of these influences in 2005. Net foreign earnings amounted to a much higher proportion of the value of exports in indigenous companies than in foreign-owned companies. Specifically, net foreign earnings amounted to an estimated 87% of the value of exports of indigenous companies in all manufacturing and market services, compared to just 34% for all foreign-owned companies.

Table 5.6 also shows that net foreign earnings were a higher proportion of the value of exports in services than in manufacturing, with figures of 62% for total market services and 30% for total manufacturing. Consequently, net foreign earnings were a particularly high proportion of the value of exports in indigenous market services, at 94%, while the figure for foreign-owned manufacturing was particularly low at 26%.

Since net foreign earnings were generally a higher proportion of exports in indigenous companies than in foreign-owned companies, indigenous companies were generally responsible for a larger share of Ireland’s net foreign earnings than their share of Ireland’s exports. In that sense, indigenous companies were generally more important for the economy than their share of exports would suggest. Similarly, since net foreign earnings were a higher proportion of exports in services than in manufacturing, service exports were more important for the economy than their share of total exports would suggest.

Table 5.7 shows that in 2000 Irish indigenous manufacturing companies accounted for 7.0% of total exports, whereas they accounted for 14.5% of total net foreign earnings. In the same year, Irish indigenous producers in all sectors combined accounted for 12.0% of total exports compared to 27.5% of total net foreign earnings. A similar point applies to each sector in each year shown in Table 5.7—Irish producers always accounted for a larger share of net foreign earnings than their share of exports.

Table 5.7 Percentage distribution of exports and estimated net foreign earnings, Irish indigenous and foreign-owned, 1985, 2000 and 2007

As regards trends over time, it is clear in Table 5.7 that foreign-owned companies increased their share of both exports and net foreign earnings over the total period of the boom from the mid-1980s to 2007, with their share of exports rising from about 70% to 87% while their share of net foreign earnings rose from about 56% to 70%. Within this long period, however, there were differences between shorter periods, so that the trend in indigenous companies was relatively strong in the late 1980s and the 2000s while the 1990s was the time when growth was clearly faster among foreign-owned companies.

It was already noted above that the value of indigenous manufacturing exports increased at a slightly faster rate than the growth of exports from foreign-owned industry in Ireland in 1986–1990. Since indigenous manufacturing accounted for almost 30% of total net foreign earnings in 1985, and almost 40% of net foreign earnings from manufacturing, its rapid export growth at that time made an important contribution to transforming the 1980s recession into an economic boom.

Later, between 2000 and 2007, indigenous companies increased their share of total exports from 12.0% to 13.4% and they increased their share of net foreign earnings from 27.5% to 30.4%. This primarily reflected trends in manufacturing as the trends in both exports and net foreign earnings were a good deal weaker in foreign-owned manufacturing than in indigenous manufacturing at that time. In the services sector, the growth of exports and net foreign earnings was rapid in both indigenous and foreign-owned companies, so both groups participated in the rising importance of services in total exports and total net foreign earnings.

5.1.4 R&D and Innovation

It was noted above that during much of the boom the sectoral composition of Irish indigenous industry was changing, moving away from the lower technology sectors and towards the higher technology sectors. This shift was relatively strong by international standards in the 1990s, although it ceased in the last phase of the boom largely because of foreign takeovers of Irish companies in the higher technology sectors.

This shift towards the higher technology sectors is not necessarily conclusive evidence that there was a move into genuinely higher technology activities because the higher technology sectors would include some niches where companies were making relatively simple products with mature technologies. Therefore, it is relevant to ask whether there was any supporting evidence of technological upgrading in Irish indigenous industry—in the form of more research & development (R&D) and a high rate of innovation.

To take the question of R&D first, at the start of the boom a relatively small amount of R&D was being performed in Ireland compared to many other countries. In 1986, gross expenditure on research & development (GERD) amounted to just 1.0% of GNP in Ireland,¹¹ whereas the corresponding figures for most other OECD countries were between 1.4 and 2.7%.¹² The Irish figure rose considerably from 1.0 to 1.43% by 1998 and remained at that level in 2004, although that was still relatively low compared to 1.85% for the EU and 2.24% for the OECD.¹³

GERD data refer to all R&D performed in a country, including in higher education, government research organisations, etc., as well as in businesses, whereas business expenditure on research & development (BERD) is a focused measure of R&D in businesses. In most developed economies, BERD accounts for a substantial majority of GERD, but BERD was so limited in Ireland at the start of the boom that it accounted for only about half of GERD and 0.5% of GNP in 1986 (Eolas 1990). BERD then increased relatively fast so that it accounted for 65–70% of GERD in the late 1990s and 2000s. BERD as a percentage of GNP also rose quite fast from 0.5% in 1986 to between 0.95 and 1.05% in the late 1990s and early 2000s. The Irish figure then was quite close to, but still below, the EU level which ranged between 1.06 and 1.17%. The Irish figure was further below the OECD level which was generally close to 1.5%.¹⁴

To focus more specifically on Irish indigenous industry, indigenous companies accounted for less than half of BERD throughout the boom, but that partly reflected the fact that their output and sales were lower than those of foreign-owned firms in Ireland. If we look at BERD intensity, BERD as a percentage of gross output was 0.5% in indigenous manufacturing at the start of the boom in 1988, and that was just a little lower than the figure of 0.6% for foreign-owned manufacturing.¹⁵ BERD intensity in indigenous manufacturing increased substantially to 1.1% of gross output by 1997, while there was also a similar increase in foreign-owned manufacturing, to 1.2% of gross output. The figure for indigenous industry declined later to 0.75% by 2003, while BERD intensity in foreign-owned firms declined more sharply to 0.65% in 2003, almost the same level as in 1988.¹⁶

Even at the peak level in the late 1990s, BERD intensity in both indigenous and foreign-owned firms was low compared with the OECD. Thus, the 1997 BERD intensity figures of 1.1% in Irish indigenous manufacturing and 1.2% in foreign-owned manufacturing in Ireland were both a good deal lower than the OECD figure of 2.4%. However, the reasons for this were quite different in indigenous and foreign-owned industries.

In indigenous industry, overall BERD intensity was relatively low because of the sectoral composition of indigenous industry—being relatively highly concentrated in sectors that generally had low R&D intensity in most countries, and relatively less concentrated in the sectors which generally had the highest R&D intensity. When viewed sector by sector, Irish indigenous firms did not have systematically lower R&D intensity than OECD firms in the same sectors. Thus, Table 5.8 shows that in 1997 Irish indigenous companies had higher R&D intensity than the OECD in 8 out of 14 manufacturing sectors, including electrical & electronic equipment—a high-tech sector, and machinery & equipment—a medium–high-tech sector. Irish indigenous companies also had almost the same R&D intensity as the OECD in medical & technical instruments—another high-tech sector.¹⁷

Table 5.8 BERD intensity (BERD as % of gross output), by nationality of ownership and manufacturing sector, compared to the OECD, 1997

In contrast, foreign-owned industry in Ireland was particularly highly concentrated in the sectors that generally had high R&D intensity in most countries, but for the most part foreign-owned firms in Ireland tended to have systematically lower R&D intensity than OECD firms operating in the same sectors as themselves. Thus, Table 5.8 shows that in 1997 foreign companies in Ireland had lower R&D intensity than the OECD in 11 out of 14 manufacturing sectors. In most of the high-technology and medium–high-technology sectors, R&D intensity in the foreign companies in Ireland was much less than the level seen in the OECD. It was possible for them to prosper in this way, despite their lower R&D intensity, because they could benefit from the R&D performed by other branches of the same MNCs in other countries.

When R&D intensity declined later in the 2000s in both Irish indigenous and foreign-owned industries, the reasons for this were again quite different in indigenous and foreign industries. In foreign-owned industry, the main reason was because of the decline of the electronics industry, which was already mentioned above. In 1997 the electrical & electronic equipment sector accounted for 49% of all R&D in foreign-owned firms, so trends in that sector had a great influence on total R&D in foreign firms. Between 1997 and 2003, the sector’s share of the total output of foreign-owned industry declined significantly and its R&D intensity dropped from 1.7% of gross output to 0.7%. It seems that cutting R&D activity was an aspect of a broader reduction of electronics production.

In Irish indigenous industry, the decline in R&D intensity, from 1.1% of gross output in 1997 to 0.75% in 2003, was less severe than in foreign-owned industry. When viewed sector by sector, there was no general pattern of decline in R&D intensity across most of the R&D intensive indigenous sectors, and there was also no single one of them that experienced a very sharp decline in R&D intensity comparable that seen in foreign-owned electronics. Rather, the main cause of the overall decline in indigenous R&D intensity was the generally declining share of the most R&D intensive sectors in total indigenous manufacturing output. As was discussed above in connection with Tables 5.2, 5.3 and 5.4, the high-technology and medium–high-technology sectors tended to constitute a declining portion of indigenous manufacturing in the early 2000s, largely because of foreign takeovers of indigenous companies in these sectors. Since these sectors were far more R&D intensive than total indigenous manufacturing, their declining share of total indigenous manufacturing output had the effect of reducing the total R&D intensity of indigenous manufacturing.¹⁸

In order to be of real benefit for companies and industries, business expenditure on R&D has to lead to innovation, which in turn can be expected to result in commercial or economic benefits such as increases in productivity, sales, profitability and employment. In the Irish context, there is good evidence that R&D did in fact tend to have a significant positive influence on innovation. Hewitt-Dundas and Roper (2008) found that in 2003–2005 there were “strong positive R&D effects on both product and process innovation as well as innovation success”. Thus, they found that having in-house R&D significantly increased the probability that a plant would engage in product innovation and process innovation, while it also increased the share of plants’ sales accounted for by new and improved products.

Using a different and independent set of data for 2004–2006, Doran et al. (2012/2013) again found that having in-house R&D significantly increased the probability that plants would engage in product innovation, and they found that this effect was almost twice as strong for Irish indigenous businesses as for foreign-owned businesses. They also found that having in-house R&D significantly increased the likelihood that Irish indigenous plants would engage in process innovation, although this effect was not significant in foreign-owned companies. Thus, the influence of in-house R&D activity on innovation was generally more important for Irish indigenous plants than for foreign-owned plants, presumably because a good deal of the innovation that occurred in the foreign-owned plants was coming from R&D performed in other branches of the same MNC located in other countries.

Levels of innovation activity in Ireland were quite high compared to other European countries in the 1990s. Data from the Community Innovation Survey 1 (CIS1), collected in 1993, indicate that Ireland ranked highest among a group of ten European countries in terms of the percentage of establishments that had introduced a technologically changed product or process during the previous three years. When ranked in terms of the percentage of innovating plants’ sales derived from innovative products, Ireland was in the middle of the group of ten countries. In a later survey carried out in 1997, CIS2, Ireland ranked highest among a group of fifteen European countries in terms of the percentage of firms introducing innovative products and also ranked highest in terms of the percentage of firms introducing innovative processes. In that 1997 survey Ireland ranked somewhat higher than previously, at fourth out of fifteen countries, in terms of innovative products as a percentage of sales (O’Malley et al. 2008, Table 5.1).

These findings refer to all manufacturing plants in Ireland, including Irish indigenous and foreign-owned plants. A different series of surveys, the Irish Innovation Panel surveys, found that Irish indigenous plants were generally less likely than foreign-owned plants to be product or process innovators (Hewitt-Dundas and Roper 2008, Table 2). However, since the proportion of indigenous plants who were innovators was just 3 or 4 percentage points below the overall national figures for Ireland in the late 1990s (Hewitt-Dundas and Roper 2008, Tables 1 and 2), Irish indigenous industry still ranked quite highly by European standards.

As regards trends in innovation over time, the pattern in Irish indigenous industry looks broadly consistent with the trend in R&D intensity in the 1990s and early 2000s, although the innovation figures did not change by large amounts. For example, the proportion of indigenous plants that were product innovators increased from 57.8% in 1991–1993 to 61.8% in 1997–1999 and then declined to 53.6% in 2000–2002. Also, the proportion of indigenous plants that were process innovators increased from 51.3% in 1994–1996 to 61.7% in 1997–1999 and then declined to 51.3% in 2000–2002 (Hewitt-Dundas and Roper 2008, Table 2).

In order to be of real economic significance, R&D and innovation must have beneficial consequences for output, productivity, employment, etc. There is good evidence that such consequences did occur in Ireland in the 1990s and early 2000s.

Roper and Love (2004) used the proportion of a plant’s sales that was derived from innovative products as an indicator of a plant’s innovation success, and they found that there was a strong positive link between product innovation success and business growth. The effects of product innovation on productivity were negative in the short-term due to temporary disruption but then the longer term effects on productivity were positive. The benefit of process innovation for productivity was more immediate and enduring.

Kearns and Ruane (1998) found that technological activity in plants was an important determinant of their probability of survival. Looking at Irish indigenous manufacturing plants that existed in 1986, they found that technologically active plants had a higher probability of surviving until 1996 than comparable plants that were less technologically active. This was true for several different variables that were used to represent “technological activity”, including the scale of R&D, R&D intensity and sales of innovative products developed within the plant. Kearns and Ruane (1999) found that foreign-owned plants that undertook R&D in Ireland were more likely to survive in Ireland for longer than those that did not undertake R&D. The R&D performers also had lower rates of job loss and their jobs lasted for longer than among those that did not undertake R&D.

As was outlined in Chapter 2, the state’s industrial policy began in the mid-1980s to put a somewhat greater emphasis on the aim of developing Irish indigenous industry. Part of that effort was the introduction of enhanced measures to support technological development in indigenous companies. Consequently, the share of the industrial policy budget going to science and technology measures increased from 11% in 1985 to 21% in 1992.

Given that background, it is worth noting here that Roper and Love (2004) found evidence that grant support for product innovation was a statistically significant and positive influence on innovation success. They found that there was a significant policy effect working through specific grants for R&D and innovation and through stimulating investment. Similarly, Hewitt-Dundas and Roper (2008, 2010) found that state support was important in increasing the probability that plants would be engaging in product and process innovation as well as having success with their innovations.

To sum up on R&D and innovation, there was a substantial increase in R&D intensity in Irish indigenous manufacturing between the start of the boom and the late 1990s, which helps to confirm that there was a genuine shift into higher technology activities. By the late 1990s, R&D intensity in individual indigenous manufacturing sectors was generally comparable to the corresponding sectors in the OECD, including the higher technology sectors, while the relatively low overall R&D intensity of indigenous manufacturing was a result of its sectoral composition. When indigenous R&D intensity declined after the late 1990s, that was mainly an effect of foreign takeovers of indigenous companies in the higher technology sectors. The evidence on innovation levels and innovation trends is broadly consistent with this account of R&D levels and trends.

There is also evidence that confirms the expectations that R&D should lead to innovation while both R&D and innovation should lead to economic effects such as faster growth, higher productivity and better employment prospects. Finally, the state’s industrial policy provided important assistance which supported R&D and innovation.

5.1.5 Company Size

As was outlined in Sect. 2.2 in Chapter 2, the industrial policy changes that were introduced during the 1980s aimed to develop larger and stronger Irish indigenous firms by adopting a somewhat more selective approach. It was intended to focus state support and incentives more on building larger Irish companies that would be able to export more successfully by developing greater competitive advantages in areas such as scale of production, technological capabilities and export marketing.

This policy objective arose in the 1980s at a time when a major long-term decline had been occurring among the larger Irish indigenous companies in the more internationally traded activities. The larger companies were generally engaged in activities in which there were significant economies of scale—hence their relatively large size by Irish standards. But they were generally not large enough to match larger and longer established foreign competitors in a free trade environment. At the same time as many larger firms were declining, growth was occurring among small firms. The small firms were generally engaged in activities in which economies of scale were not important and most of them were limited to serving the Irish market.

In Table 5.9, the first column shows this process of fragmentation or deconcentration during 1973–1987. In a context of declining total indigenous manufacturing employment, the employment decline was very rapid in the relatively large establishments employing over 200 people, the decline was slower in the middle size establishments employing 50–199, and employment increased in the smallest establishments.

Table 5.9 Average annual percentage employment change in Irish indigenous manufacturing, by size class of establishments, 1973–2001

This pattern changed in the early years of the boom, 1987–1990, as seen in the second column of Table 5.9. Total indigenous manufacturing employment started to grow but the smallest size class contributed very little to this growth while instead the middle size class grew by most. The decline of the largest size class did not stop but it did slow down a good deal. This did not amount to a clear-cut reversal of the long-established trend of fragmentation, but it was a distinct change from that previous trend.

Subsequently, in 1991–2001, total indigenous manufacturing employment grew quite fast and all the size classes grew at much the same rate. Thus, there was a definite end to the previous process of fragmentation. While the largest size class did not grow any faster than the other size classes, it did grow quite fast just by growing at close to the average rate, and the number of indigenous establishments employing over 200 increased from 61 to 74 while the number employing over 500 increased from 12 to 16.

In the final phase of the boom, in 2001–2007, there was some re-emergence of the fragmentation trend, as seen in Table 5.10, although the contrast between decline in the largest size class and growth in the smallest size class was much less marked than in 1973–1987.¹⁹ It seems quite possible that the fragmentation trend in 2001–2007 was partly a reflection of takeovers of indigenous companies in the higher-tech sectors by foreign companies, which were discussed above, since the indigenous companies that attracted the attention of foreign buyers would probably have tended to be relatively successful and hence probably larger than average.

Table 5.10 Average annual percentage employment change in Irish indigenous manufacturing, by size class of enterprises, 2001–2007

A notable aspect of the development of Irish indigenous companies was the fact that, relative to the overall size of indigenous industry in Ireland, there was extensive development of Irish MNCs operating successfully in other countries. This feature was already visible in the 1980s before the boom began (O’Malley 1989, Chapter 6) and it increased a good deal over the next two decades. By 2010, Irish MNCs in all sectors employed 249,000 people in other countries, with 75,000 of these being in affiliates of Irish manufacturing companies.²⁰ Compared to 101,000 employed in indigenous manufacturing in Ireland in 2010, this seems like a large number.

While many indigenous manufacturing companies had foreign subsidiaries by then, the bulk of the employment was in the overseas branches of just the top ten or so. Forfas (2001) observed that most of the largest manufacturing companies concerned were in traditional industrial sectors such as construction materials, paper, packaging & printing, and agribusiness. They tended to be dominant home market players, often in non-traded sectors, and they used their existing strong positions in terms of expertise and finance in the domestic market in order to develop overseas. It is noticeable that most of their expansion overseas took the form of acquiring existing foreign companies in their own or related lines of business, rather than establishing new greenfield operations.

This type of growth produced some prominent companies that were clearly great successes in business terms, but it is unlikely that this growth had a commensurate impact on Ireland’s economic development. Such growth probably had some benefits for the home economy, such as more employment in higher value-added functions, and additional income coming from the flow of profits back to Ireland with consequent benefits for tax revenue and employment, etc. (Forfas 2007). However, such benefits could hardly compare to the effects of a similar amount of growth occurring in Ireland.

The extensive growth of Irish MNCs does at least show that Ireland was reasonably well endowed with the entrepreneurial ability and management skills required to build large and internationally successful companies. Therefore, if the development of indigenous manufacturing in Ireland remained unsatisfactory, this was not due to an absence of business ability and skills but was more an effect of the constraints confronting indigenous industrial companies in a late-industrialising country such as Ireland (as discussed in Sect. 2.2 in Chapter 2).

5.2 Growth and Development of Indigenous Market Services

5.2.1 Output, Exports and Net Foreign Earnings

The services sector accounted for a large part of total output and employment long before the boom, but traditionally it was not seen as a significant source of exports because most of the more traditional services had to be provided locally for local customers. There were always some exceptional services which could be exported such as transport services, tourism and some financial services but in the mid-1980s, just before the boom began, these exports were relatively small in Ireland compared to manufacturing exports.

In 1985, services accounted for 39% of total output but only 12% of total exports. Just 10% of services output was exported while 62% of manufacturing output was exported. This picture changed substantially over the next two decades as technological and organisational changes resulted in services becoming a growing part of international trade in Ireland and in the wider world. By 2005 services still accounted for 39% of output in Ireland, but the proportion of their output that was exported had increased greatly and they accounted for 33% of Ireland’s exports.²¹

During the period 1985–2005, the value of services exports increased by €40 billion, from €2 billion in 1985 to €42 billion in 2005. Over the same period, the value of total exports from Ireland increased by €114 billion, so services accounted for 35% of the total increase in exports.

Net foreign earnings were generally a higher percentage of the value of exports in services than in manufacturing, largely because imported material inputs were generally a smaller proportion of the value of exports in services than in manufacturing. Consequently, the services sector accounted for a greater share of total net foreign earnings than its share of exports. By 2005, services accounted for 50% of net foreign earnings according to our estimates,²² and services had accounted for 56% of the total increase in net foreign earnings since 1985.

Given the key role of exports, and especially net foreign earnings, in driving overall economic growth in Ireland, the services sector clearly played a very important part in the boom. Unfortunately, however, the data on services are generally not as comprehensive or as detailed as the available data on manufacturing, and there is a particular scarcity of data on services distinguishing between Irish indigenous and foreign-owned firms.

The CSO’s Annual Services Inquiry (ASI) did not include data on services by nationality of ownership until 2001. The data by nationality that it began to publish in 2001 included turnover, gross value-added and employment—but no data on exports—and then exports by nationality of ownership was first included in 2008. The data concerned referred to service enterprises with 20 or more persons engaged and did not include financial services and insurance.

The ASI data show that the services sector was mostly Irish-owned in 2001–2007. Although the Irish indigenous part of the sector grew quite fast during that period, its share of the sector’s employment and turnover was declining as the foreign-owned part of the sector was growing even more rapidly. Irish indigenous firms accounted for 79% of the sector’s employment in 2001 declining to 73% by 2007, while they accounted for 65% of the sector’s turnover in 2001 declining to 56% in 2007.²³

Irish indigenous firms had a much smaller share of services exports than their share of employment or turnover. When the ASI published data on exports by nationality of ownership in 2008, Irish indigenous firms accounted for 22.7% of non-financial services exports. Our own estimate for 2007 indicates that indigenous firms accounted for 18.5% of all services exports, including financial services.²⁴

Net foreign earnings were generally a higher percentage of the value of exports in Irish indigenous services than in foreign-owned services. Consequently, although indigenous companies accounted for about 18.5% of total services exports in 2007, their share of total net foreign earnings in services was a good deal higher at an estimated 28.4%.

As a result, in terms of net foreign earnings, indigenous services were making a quite important contribution to the overall economy in 2007. As was seen in Table 5.7, indigenous services accounted for about 17% of all net foreign earnings, compared to 12.1% for indigenous manufacturing, 1.3% for agriculture, and a total of 30.4% for all indigenous firms.

As regards trends over time, we have reasonably good estimates for 2000 as well as 2007. These estimates indicate that indigenous services exports and net foreign earnings grew faster than total exports and total net foreign earnings from all sectors during that period, increasing their share of total exports from 4.2% to 6.8% and increasing their share of total net foreign earnings from 11.1% to 17.0% (Table 5.7). Furthermore, indigenous services exports were growing faster than total commercial services exports from the EU-27 in 2000–2007, increasing their share of this total from 0.65% to 1.06%.²⁵ At the same time, however, exports and net foreign earnings from indigenous services were growing more slowly than exports and net foreign earnings from foreign-owned services in Ireland.

Taking a longer view over the whole period of the boom, it is known that services accounted for 12.5% of all exports in 1985, as was seen in Table 5.7, but we do not have any adequate estimate of how much of that came from indigenous or foreign-owned companies. In Table 5.7, a simple assumption is made that half came from indigenous companies and half came from foreign-owned companies, but this is just a technical assumption that is not based on evidence and consequently it could be wrong by a wide margin.

However, services exports were so limited in 1985 that, even if there is a high degree of inaccuracy in breaking down this small amount between indigenous and foreign companies, it is still possible to make some valid remarks on long-term trends. Thus, using the estimates underlying Table 5.7, we can say that the increase in the current value of exports in 1985–2007 was roughly four and a half or five times greater in foreign-owned services than in indigenous services. In a similar vein, we can conclude that the increase in the current value of net foreign earnings in 1985–2007 was about two and a half or three times greater in foreign-owned services than in indigenous services. These observations are valid regardless of whether indigenous companies accounted for as little as one-quarter or as much as three-quarters of services exports in 1985.

Although the long-term growth of exports from Irish indigenous services looks rather weak when compared with foreign-owned services in Ireland, it was actually quite strong when compared with wider international experience. The broader context here is that the growth of total services exports from Ireland was much faster than in most other countries. Loungani et al. (2017, Table 2) found that Ireland was the 32nd largest exporter of services in the world in 1980, rising to 30th in 1990, 22nd in 2000 and 14th in 2010. They also reported that Ireland’s services exports increased from $1.4 billion in 1980 to $89.1 billion in 2010 (valued in current US dollars), which means that they grew by 14.8% p.a. Much of this was due to the growth of exports from foreign-owned services in Ireland, but indigenous growth was also significant. If we take it that Irish indigenous companies accounted for no more than 90% of services exports in 1980 and at least 17% of the figure for 2010,²⁶ the value of their exports increased by at least 8.6% p.a. in 1980–2010. For comparison, the corresponding growth rates for the largest services exporters in the world were USA 8.6% p.a., UK 6.9% p.a., Germany 6.8% p.a., France 5.2% p.a. and Japan 6.6% p.a.

As regards the composition of services exports from Ireland, it was noted in Chapter 4 that at the start of the boom they consisted largely of transport services (41% of total services exports) and lodging & catering (23%), with smaller contributions coming from a wide variety of other services.²⁷ We have no data on this distinguishing between indigenous and foreign-owned companies, but it seems likely that the transport and lodging & catering services were mostly indigenous. In the case of transport services, there were a few prominent Irish companies involved in international transport in 1985—Aer Lingus, B&I line and Aer Rianta—which were large enough on their own to have accounted for a substantial majority of Ireland’s exports of transport services if we assume that about 35% or more of their sales were exports.²⁸

At the end of the boom, when we have data on services exports by nationality of ownership, the data show that exports of transport services were predominantly indigenous as Irish-owned companies accounted for 82% of all exports of transport services—including air, water and land transport as well as warehousing and support services for transport.²⁹

These transport services were still a substantial component of all indigenous services exports at the end of the boom, although their share of the total was probably lower than it had been in the 1980s. Table 5.11 shows the sectoral composition of Irish indigenous services exports in the period 2008–2012, based on data from the Annual Services Inquiry. The four transport categories in the table together accounted for 33% of the total.³⁰

Table 5.11 Irish indigenous services exports by sector, 2008–2012

Two other types of services made up most of the remainder—(1) software, with about 27% of the total, and (2) a range of professional, scientific and technical services with about 26% of the total. Software services are in the first item in Table 5.11, which includes the publishing of software (as well as publishing of books etc.), together with computer programming, consultancy and related activities. The range of professional, scientific and technical services in Table 5.11 includes scientific research & development, legal & accounting activities, head office activities and management consultancy, and architectural & engineering activities.

There are two significant omissions from Table 5.11, financial services & insurance as well as accommodation & food services. Financial services & insurance are left out because the data in Table 5.11 come from the Annual Services Inquiry (ASI) which did not cover that sector. Accommodation & food services are largely absent from the table because the ASI is a survey of companies, and companies in that type of business are often not able to report how much of their sales were exports—meaning how much of their sales were to customers who are normally resident in another country—since they provide their services in Ireland to customers who come to them in Ireland. Hence the CSO needs to use separate surveys of foreign residents at airports and ferry ports to collect the data it requires on export tourism expenditure for the balance of payments, national accounts, input–output tables, etc. (Lawless and Studnicka 2017, Chapter 3).

Our own estimates (derived as explained in Appendix) of the value of indigenous financial services & insurance exports indicate that they accounted for about 18% of all indigenous services exports in 2005 and 2007. As regards accommodation and food services, the input–output table for 2005 indicates that exports of hotel & restaurant services were worth a similar amount to exports of transport services. We do not have an indigenous/foreign breakdown of the exports of hotel & restaurant services, but available data indicate that they were predominantly indigenous,³¹ so it seems possible that the indigenous component was broadly comparable in value to indigenous exports of transport services.

Combining these observations with Table 5.11, it seems that the overall composition of indigenous services exports at the end of the boom was approximately as follows: transport services 20%, hotels & restaurants 20%, financial services & insurance 18%, software 17%, professional, scientific & technical services 16%, and others 9%. Obviously, these figures cannot be regarded as precise, particularly in the case of hotels & restaurants and, to a lesser extent, financial services & insurance.

Whereas the role of transport services and hotel & restaurant services represented elements of continuity with the past situation before the boom, the prominence in indigenous services exports of software, financial services and the other professional, scientific and technical services were newer developments which largely emerged, or at least grew rapidly from small beginnings, during the boom.

These newer types of exports were much the same as the types of indigenous service activities that received a good deal of attention and assistance from state development agencies such as Forbairt or Enterprise Ireland since the 1980s, on the grounds that they had good potential to develop exports unlike many other services. As a consequence of the interest of these development agencies, surveys such as the Annual Employment Survey, the Annual Business Survey of Economic Impact and the regular surveys of business R&D (BERD), which were conducted by or for the agencies, generally covered a group of selected service activities in addition to all of manufacturing. This group of selected services was sometimes called “internationally traded services”, sometimes “software and other services” or sometimes “information, communication and other services”, when presenting results of the surveys concerned. The categories of services within this group included computer programming, consultancy & related activities; other IT & computer services; some elements of financial services; business services; education; and other services.

The surveys mentioned above indicate that these services tended to be relatively highly skilled. Thus, their pay levels were relatively high, with average payroll costs per employee generally being 20–30% higher than in indigenous manufacturing in 2000–2007.³²

In addition, the surveys of R&D show that Irish indigenous companies in these services were particularly R&D intensive. In 2003 indigenous companies in this group of services spent €131 million on conducting R&D, which amounted to 44% of all indigenous expenditure on R&D in manufacturing and this group of services combined.³³ For comparison, in the same year this group of services accounted for just 17% of sales and 19% of employment in indigenous companies in manufacturing and this group of services combined.³⁴ Expenditure on R&D had grown very rapidly in these indigenous services, by 26.5% p.a. in 1993–2001, valued in current prices, compared to 8.0% p.a. in indigenous manufacturing and 10.0% p.a. in foreign-owned manufacturing in Ireland.³⁵

These points concerning R&D are especially relevant to the software industry. In 1993–2001, R&D expenditure in indigenous “software & computer related services” increased by 35% p.a., and by 2001 it accounted for more than three-quarters of indigenous R&D in the group of selected services. The remainder of this section looks at the indigenous software industry in a little more detail.

5.2.2 The Irish Indigenous Software Industry

Some of the earliest software firms in Ireland began in the 1970s and more started in the early 1980s, but the substantial expansion of the industry really began in the mid-1980s as major foreign software MNCs started up in Ireland while the indigenous branch of the industry also grew quickly. In 1986, there were only about 1,800 people employed in the industry, with about 1,200 being in Irish indigenous firms and about 600 in foreign-owned firms. By 1991, employment in the industry was 7,800, rising to about 24,900 by 1999, so that employment grew by 15.6% p.a. in 1991–1999 (O’Malley and O’Gorman 2001; Crone 2002).

Throughout the 1990s, about half of employment in the industry was in indigenous firms and about half was in foreign-owned firms. Employment in indigenous companies grew by 16.8% p.a. in 1991–1999. The rate of growth of sales revenue in the indigenous industry was extremely high in 1991–1999 at 29.4% p.a., while the growth rate of indigenous exports was even higher at 37.3% p.a. The indigenous branch of the industry was becoming increasingly export-oriented, with exports rising from 40.7% of sales in 1991 to 62.0% by 1999. By 1995, 80% of indigenous software companies were exporting to some extent (O’Malley and O’Gorman 2001; Crone 2002).

The late 1980s and the 1990s was a period when software was a relatively fast-growing industry in other countries too. However, the sales and exports of the Irish indigenous software industry were growing a good deal faster than the international market for software, so the industry was gaining market share (O’Malley and O’Gorman 2001).

It was noted in Sect. 5.1 that there was also relatively strong growth in the indigenous branches of the high-technology manufacturing sectors at this time. However, indigenous development in the software sector was on a different scale. Indigenous companies accounted for only small minorities of employment in the high-tech manufacturing sectors, whereas they accounted for half the employment in the software sector. Indigenous employment of over 9,000 in software by 1997 was far more than the 4,900 in the indigenous computers, pharmaceuticals and instrument engineering sectors combined (O’Malley and O’Gorman 2001).

In the period after about 2000, the growth of the indigenous software industry slowed down a good deal, although its growth did continue at a more modest pace. The data on the industry that are available for the 2000s are not precisely comparable to the earlier data quoted above for the 1990s, since the 1990s data originally came from the National Software Directorate whereas the available data for the 2000s come from the Annual Employment Survey and the Annual Business Survey of Economic Impact. Nevertheless, it is evident that the indigenous software industry grew a good deal more slowly in the 2000s.

In 2002–2007, employment in Irish indigenous “computer programming, consultancy and related activities” increased by 1.3% p.a., which was a little faster than indigenous manufacturing but slower than some other indigenous services.³⁶ In 2000–2007, sales of the same sector increased by 5.2% p.a. while its exports grew by 6.3% p.a., in current values.³⁷ Again, this was somewhat faster than indigenous manufacturing but slower than some other indigenous services.

It was found in Sect. 5.1 that a significant cause of growth slowing down at this time in the indigenous branches of the high-technology manufacturing sectors was because many of the Irish companies concerned were being taken over by foreign-owned companies and consequently disappearing from the indigenous ranks. There is evidence that a similar trend was beginning to have a substantial effect in the software sector in the late 1990s.

In 1998 and 1999, industry data from the National Software Directorate made a three-way distinction between Irish-owned, foreign-owned and “takeover” companies—“takeover” companies being those that were originally Irish-owned and then subsequently acquired by foreign owners. The number of takeover companies was quite small at just 11 in 1998 and 12 in 1999 compared to over 600 Irish-owned firms. However, the takeover companies were relatively large and highly export-oriented, so that their employment amounted to 19% of employment in Irish-owned companies in both 1998 and 1999, their sales amounted to 20% of sales of Irish-owned companies in 1998 and 17% in 1999, while their exports amounted to 28% of exports of Irish-owned companies in 1998 and 23% in 1999 (Crone 2002, Table 1).

Unfortunately, data are not available to follow this through into later years. However, it is relevant to note that Crone (2013) presented brief profiles of eight of the most prominent indigenous software companies to have been established in the period between the mid-1990s and 2001/2002, and he mentioned that five of the eight had been acquired by foreign-owned companies by October 2009.

When Irish firms were taken over by foreign firms they dropped out of the “Irish indigenous” category, which weakened the trend in statistics for that category. That, in itself, was not necessarily always a bad thing for the Irish economy, since it was possible that the new parent company might aim to promote the continuing growth and development of the Irish subsidiary in Ireland. However, it was somewhat more likely that the new parent company would be focused essentially on promoting its own growth by gaining access to the Irish subsidiary’s technology, marketing or other assets. In that case, growth or even maintenance of activity in the Irish subsidiary could become at best a secondary consideration that would not receive much support.

Ó Riain (2004, p. 123) quoted journalist John Sterne’s analysis of eighteen takeovers (domestic and international) in the Irish IT industry in the late 1980s:

IT acquisitions in this country, with just a couple of exceptions, have meant job losses rather than gains and reductions, not increases, in the functions and responsibilities of organisations here. Every so often, though, an agreement is struck which builds on what has gone before, strengthening instead of weakening the country’s technology base. (Irish Computer, March 1992)

As noted above, the growth of the indigenous branch of the software industry in the late 1980s and 1990s occurred alongside the rapid growth of the foreign-owned branch of the industry. There were some major differences between the two branches. Much of the foreign-owned branch consisted of large companies which were engaged in large-scale production of software packages or products for mass markets. Typically, these were US MNCs which initially developed their products in the USA. They adapted or “localised” their products in Ireland to make them suitable for many different markets in Europe and elsewhere, and then produced and exported them. Since many of these were leading companies in the world software industry—such as Microsoft, Oracle and Lotus—Ireland became the second largest exporter of software in the world in the 1990s.

Irish indigenous companies in the software sector were generally much smaller and their activities were different. It is helpful to make a distinction between software products and software services. Software product companies developed a software programme which could be copied many times and sold to many customers, whereas software services were provided uniquely to each customer as required, rather than involving repeated selling of copies of a standardised product. Such software services could include bespoke (or once-off, customised) development of programmes for individual customers, system integration, consultancy and technical training. The main foreign-owned software companies in Ireland were predominantly focused on products, whereas indigenous companies had significant involvement in both products and services.

In addition, the indigenous companies’ products were generally more specialised and more focused on narrow niche markets compared to the major foreign-owned companies, so their products were sold in much smaller quantities. For example, indigenous companies’ products aimed to address the specific requirements of different types of customers in financial services, dairy processing, distribution, drinks, chemicals and many other sectors.

It seems that the indigenous software industry was engaged equally in products and services in the 1990s (O’Malley and O’Gorman 2001; Crone 2002), although it is difficult to be precise about this since many firms were involved in both to some extent. Indeed, since many of the industry’s products were specialised for a limited range of customers, selling such products often required the companies concerned to provide some related services, such as an element of additional bespoke development, installation, training and after-sales support.

Another difference between indigenous and foreign-owned firms was that foreign-owned companies had much higher sales and exports per employee than indigenous companies. Although indigenous companies accounted for close to half the industry’s employment, they accounted for less than 20% of sales and 15% of exports in the late 1990s (Crone 2002). However, this did not mean that indigenous companies had exceptionally low sales and productivity per employee, since it was the foreign-owned companies in Ireland that were unusual in having such extremely high sales per employee. This arose because much of the value embedded in the output of those companies was generated, not by their subsidiaries and employees in Ireland, but elsewhere in the value chain by the R&D and other activities in other parts of the same MNCs (O’Malley and O’Gorman 2001).

As regards the size of indigenous software companies, in 1991 they employed an average of 13.1 people, rising to 14.7 by 1998. In 1991, there were 14 companies, or 4.8% of the total number of indigenous companies, that employed over 50, and they accounted for 33% of total employment in the indigenous software industry. By 1998, the number of companies employing over 50 increased to 34, or 5.4% of the total, and they accounted for 43% of total employment.³⁸ Thus there was an increase in company size, but at the same time, there was nothing on the scale of the largest companies in the USA or Western Europe where some software companies employed thousands.

This did not necessarily have to be a major problem for indigenous software companies if much of the market continued to be quite segmented and if the Irish companies were appropriately specialised in selected segments. O’Malley and O’Gorman (2001) made the point that even in the USA there was a large number of successful small companies and the size structure of much of the software industry was similar to the indigenous industry in Ireland, with only the top few per cent of US companies being far larger than any of the Irish companies. However, small size did present a substantial barrier to competing in significant areas of the market which were dominated by large firms.

The software industry was one of the more successful stories in Irish indigenous growth during the boom and this success was explained by several favourable influences in combination.³⁹ First, a suitably qualified and high-quality labour force was available for the industry at a time when shortages of relevant skills occurred quite commonly in other countries. The education system in Ireland was turning out graduates in computer science and software engineering at a rate that was relatively high compared with most other OECD countries. Initially, much of the motivation for rapidly increasing the supply of computing graduates was to take advantage of the perceived opportunity to attract foreign software MNCs to Ireland if the right type of skills were made available, but it soon emerged that indigenous software companies were also capable of benefiting from the skills concerned. In addition to the formal education system, the skills of the labour force were further developed by staff acquiring specialised expertise on the job and by companies’ staff development activities.

Demand conditions in the Irish market were also helpful for the development of many indigenous companies because they were often selling to relatively successful and sophisticated companies in Ireland in sectors such as pharmaceuticals, chemicals, drinks, dairy products and financial services. In many cases the customers concerned were foreign-owned MNCs. Irish software companies often found that their interactions with such local customers were beneficial for the development of their business and helped to prepare them for export success.

In addition, there were a number of industries in Ireland that were somewhat related to the indigenous software industry and had a helpful influence of some importance. One significant type was industries that helped to develop and improve the pool of labour skills which the indigenous software industry could draw on, such as the foreign-owned software, computer hardware and telecommunications equipment industries. Such industries constituted a relatively large concentration of information technology activities in a small economy. Another type of related industry was those in which indigenous software entrepreneurs had previously worked and gained relevant experience. These naturally included other indigenous software companies, as well as other indigenous and foreign-owned companies in information technology activities.

It can be seen from these remarks that indigenous software companies were benefiting in several different ways from being part of a type of cluster or agglomeration of similar or related industries (O’Malley and O’Gorman 2001; Crone 2002, 2013).

The state also played an important part in assisting the development of the industry. A fundamental aspect of this was the role of the education system as mentioned above, while software was also one of the selected service sectors that were eligible for support under industrial policy measures. O’Malley and O’Gorman (2001) found that four-fifths of the indigenous software companies who they interviewed had received financial assistance under such measures, and just over half of those said that it had been important or very important for their company’s development. Most of their interviewees had also received non-financial assistance, such as marketing information and assistance with developing management skills and business planning. About one-third of their interviewees said that state development agencies had influenced their strategy or goals, mostly by encouraging and assisting them to focus on developing software products for export markets.

Crone (2002) made a number of similar points about the role of state assistance, commenting that “in the late 1980s and early and mid-1990s, when there was no significant private venture capital industry in Ireland, the State agencies were the dominant external supplier of finance to indigenous software firms”. He noted that these agencies offered both grant aid and equity investment, and provided softer forms of assistance with marketing, management development and training, while the state also created a specialised set of supporting institutions for the industry.

Ó Riain (2004) presented an extensive analysis of the role of state assistance in promoting the indigenous software industry. This included quantitative data and analysis on the extent and actual effectiveness of financial support, as well as more qualitative analysis of the way in which the relevant state institutions worked.

He noted that the state agencies did not try to direct particular firms to take very specific steps such as developing specified technologies or markets. However, they did influence firms’ decisions indirectly—“primarily by attempting to create specific kinds of firms: those that are oriented toward learning, R&D, and ‘high-value-added’ competition” (Ó Riain 2004, p. 91). The agencies focused on encouraging firms to develop software products for export markets—because products were easier to export than services and because the small size of the Irish market made exporting essential. To influence firms towards software products for export, the agencies were more receptive to such firms when they applied for grants, and they focused some state supports on the problems of product exporting (p. 97). Furthermore,

state grant-giving practices … promoted a general company development program including marketing, management development, training, and R&D. The precise form this took was flexible, depending on the company, but the state agencies required that such efforts at company development take place. (O Riain 2004, p. 100)

Seen in terms of “carrot and stick” policy measures, this type of policy was all carrot with little or no stick, but the carrots were often offered subject to conditions requiring the recipients to act in ways that would contribute to long-term economic development.

Finally, the question arises why the growth of the indigenous software sector was a good deal stronger than the growth of the indigenous branches of the high-tech manufacturing sectors (although indigenous growth in those other sectors was also relatively fast except by comparison with the software industry). One important reason was probably because of differences in the structure of these industries internationally, since the other industries concerned tended to be more highly concentrated in relatively large firms, offering fewer opportunities for new or small firms. By comparison, the software industry presented lower entry barriers and it offered more significant scope for new or small firms to develop in specialised niches serving segmented markets.

Also, the rapid growth of FDI in Ireland probably had a greater positive influence on indigenous firms in the software industry than on those in the other sectors. Foreign MNCs in a range of different sectors helped to generate sophisticated and rapidly growing domestic demand for indigenous software, no doubt to a greater extent than the demand generated by foreign MNCs for products from indigenous firms in high-tech manufacturing sectors. In addition, foreign MNCs—in the software sector as well as other sectors—were employing people with software programming skills and helping to strengthen the high-level skills of the software workforce through on-the-job experience in R&D and through further training. The scale of this effect was probably greater in software than in the high-tech manufacturing sectors, as indicated by the fact that software firms accounted for a highly disproportionate share of total R&D in all foreign MNCs.⁴⁰

5.3 The Indigenous Contribution to the Boom

This section considers to what extent Irish indigenous companies contributed to causing the boom.

In view of the importance of exports—and especially net foreign earnings—as determinants of economic growth, an obvious way to look at this is to consider how much did indigenous companies contribute to the growth of exports and net foreign earnings. In that regard, the indigenous contribution over the whole course of the boom looks considerably less than the contribution of foreign-owned companies. Between 1985 and 2007, the current value of exports from Irish indigenous companies increased by an estimated 7.5% p.a. compared to 12.6% p.a. for foreign-owned companies. Net foreign earnings are more important than exports as determinants of economic growth, and the difference between indigenous and foreign-owned companies was smaller but still significant when measured in terms of net foreign earnings. The current value of net foreign earnings for indigenous companies grew by an estimated 7.4% p.a. in 1985–2007 compared to 10.4% p.a. for foreign-owned companies.

Table 5.7 shows how the balance between indigenous and foreign-owned companies changed as a result of these different growth rates. Indigenous companies accounted for about 30.0% of exports in 1985, declining to about 13.4% in 2007. At the same time indigenous companies accounted for an estimated 44.2% of net foreign earnings in 1985, declining to about 30.4% in 2007. Thus, the indigenous contribution to net foreign earnings remained much greater than it appeared to be in terms of exports, but the indigenous share of net foreign earnings was declining due to faster growth among foreign-owned firms.

In two respects, however, the indigenous contribution to the boom looks more significant than these numbers suggest—first concerning the timing and, second, concerning the degree of change or improvement in performance compared to the period before the boom.

As regards timing, it is noticeable that the indigenous contribution was relatively important in the late 1980s and in the 2000s whereas the foreign-owned contribution was very dominant in the 1990s. Thus, the indigenous contribution was influential at the time when the economy was pulling out of the lengthy recession of the 1980s and embarking on a prolonged period of rapid growth, and it was influential again later in maintaining a relatively high rate of growth well into the 2000s.

This point concerning the late 1980s was seen in Sect. 5.1 above where it was shown that Irish indigenous manufacturing exports grew a little faster than exports from foreign-owned manufacturing in 1986–1990, while they also grew significantly faster than manufacturing exports from the EU and the OECD. This was important at that time since indigenous companies accounted for almost 40% of manufacturing net foreign earnings (Table 5.7). There is no adequate indigenous/foreign breakdown of exports from the services sector in the 1980s, but services exports were relatively small then.

In 2000–2007, indigenous exports and net foreign earnings grew faster than foreign-owned exports and net foreign earnings, as seen in Table 5.7. That was a period when there was a very marked deterioration in trends in foreign-owned manufacturing, particularly in electronics, so that the contributions of indigenous companies and foreign-owned services were essential for sustaining a relatively high rate of economic growth.

The other way that the contribution of Irish indigenous companies looks significant is when one considers what was it that changed, compared to the period before the boom, resulting in a substantial rise in the rate of economic growth. Fast growth among foreign-owned firms was clearly a very prominent feature of the boom, but the increase in growth rates compared to earlier trends was greater among indigenous firms.

O’Malley (1998) examined this issue in terms of employment trends in the periods 1980–1988 and 1988–1996. He found that employment in foreign-owned manufacturing declined by 0.9% p.a. in 1980–1988 and this improved to growth of 2.3% p.a. in 1988–1996—an increase by 3.2 percentage points. In indigenous manufacturing, employment declined by 3.2% p.a. in 1980–1988 and this improved to growth of 0.8% p.a. in 1988–1996—an increase by 4.0 percentage points. Thus, the rate of growth in 1988–1996 was higher in foreign-owned industry than in indigenous industry, but the increase in growth compared to the period before the boom was greater in indigenous industry. Consequently, by 1996, employment in indigenous industry was 33,100 higher than it would have been under a continuation of the 1980–1988 trends, while employment in foreign-owned industry was 22,400 higher than it would have been under a continuation of the earlier trends.

It is not possible to do the same sort of calculations for exports because the necessary data on exports by nationality of ownership are not available for years before 1986, but there must have been a somewhat similar effect with respect to exports. It is evident that the value of exports from foreign-owned industries was already growing fast before the boom because sectors that were predominantly foreign-owned (measured in terms of output) had fast growth of exports at that time, whereas other sectors had much slower export growth. For example, O’Malley and Scott (1987) identified a group of export categories that came from predominantly foreign-owned industry sectors. The value of those exports was growing more than twice as fast as other industrial exports in 1980–1986. Consequently, when Irish indigenous manufacturing exports then accelerated to the extent that they were growing slightly faster than exports from foreign-owned manufacturing in 1986–1990, this was a major change from trends before the boom. It means that the change in the performance of indigenous exports was probably the key change that started the boom in the late 1980s.

5.4 The Role of Industrial Policy

As was outlined towards the end of Sect. 2.2 in Chapter 2, a series of changes were made in industrial policy beginning in the mid-1980s, often with the intention of providing more focused and effective assistance for the development of Irish indigenous industry (including selected internationally tradeable services). There is a certain amount of evidence indicating that such policy measures were quite effective in significant respects, at least for some time, although the findings of this chapter have shown that ultimately the overall outcome for indigenous development was no more than a partial or qualified success. This section first outlines some of the evidence that policy measures were effective and then considers why the overall outcome was not satisfactory.

O’Malley et al. (1992, Chapter 3) examined employment trends in the period up to 1990 in existing industrial companies (i.e., leaving aside new start-ups) that were assisted by grants under industrial policy measures, and they compared those trends with trends in non-assisted industrial companies. They found that the grant-assisted companies consistently had much stronger employment trends than the non-assisted companies—for example growth of 5.7% p.a. for those awarded grants in 1987 compared to −1.8% p.a. for non-assisted companies, or 11.5% p.a. for those awarded grants in 1988 compared to −1.1% p.a. for non-assisted companies. They also cited survey evidence from the Department of Industry and Commerce (1990, Chapter 11) showing that most of the companies who received grants for investments said that the support received was a major factor in their investment decision. Thus, they concluded that it was probable that state financial support usually had a beneficial effect in producing growth.

O’Malley et al. (1992, Chapter 3) noted that it was a stated aim of policy after the mid-1980s to focus attention more selectively on building on existing relatively promising companies, rather than spreading assistance too thinly among large numbers of small firms. As evidence that increasing selectivity was applied in practice, they found that the cohorts of existing firms that were awarded grants declined in size each year between 1984 and 1990. Those that received grant approval in 1984 employed 25,900 whereas those receiving grant approval in 1990 employed 13,200.

At the same time, the employment trends improved very substantially over time in each succeeding cohort of grant-assisted existing companies, e.g., growth rates in the period up to 1990 at 2.3% p.a. for the 1985 cohort, 5.7% p.a. for the 1987 cohort and 11.5% p.a. for the 1988 cohort. O’Malley et al. (1992) considered whether this improvement over time could have been caused by the increasingly favourable economic environment during this period, or whether it could have been an automatic effect of applying increasing selectivity over time.

As regards the increasingly favourable economic environment, they noted that there was also improving growth over the same period in non-assisted firms, and this presumably was an effect of the strengthening economic environment. However, the improvement was far greater in the case of grant-assisted firms, so much of the improvement in their performance could not be explained simply by the economic environment.

As regards the possibility that the improving growth in grant-assisted firms might have been an automatic effect of applying increasing selectivity in grant assistance over time, they noted that increasing selectivity would be expected to result in better average growth rates among grant-assisted companies by means of weeding out less promising grant applicants. However, increasing selectivity per se could not have resulted in rising absolute employment increases in succeeding cohorts, especially when succeeding cohorts of grant-assisted firms were declining in size each year. But, in fact, there were rising absolute employment increases in succeeding cohorts of grant-assisted firms, from net increases of 1,300 or less per year for the 1984 and 1985 cohorts to between 3,200 and 3,700 per year for the 1988, 1989 and 1990 cohorts—despite the declining size of cohorts and declining expenditure on grants. Thus, they concluded that grants were not only awarded more selectively by refraining from aiding weaker firms but they were also awarded to the more promising firms in the context of industrial policy measures which were becoming increasingly effective.

O’Malley et al. (1992, Chapter 3) also noted that it was an aim of industrial policy after the mid-1980s to focus attention more on supporting development of Irish indigenous companies. They found that the pattern of rising absolute employment increases in succeeding cohorts of grant-assisted companies was most pronounced in indigenous companies. Employment increases per year in the period up to 1990 rose from less than 500 for the 1984 and 1985 cohorts of grant-assisted indigenous firms to between 1,700 and 2,100 for the 1988, 1989 and 1990 cohorts. The corresponding rise in foreign-owned industry was smaller—from 800 or less to between 1,200 and 1,600. Again, this was consistent with stated policy goals and suggested that the policy was effective.

Further evidence that policy measures were effective was already discussed above in the section on “R&D and Innovation”. Studies were quoted there which found that policy measures had positive influences on R&D and innovation success.

In addition, the section above on the “Irish Indigenous Software Industry” referred to studies which found that industrial policy measures made a significant contribution to the development of that sector. A Sunday Business Post report in 2003 on 62 indigenous technology start-ups featured the same theme:

If you ask their bosses about Enterprise Ireland’s cash-funding policies, they are unanimous in swearing that the money has made a big difference. … Even though most of the entrepreneurs admit that the money is by far the most important thing they rely on from the state agency, they stress that it is not the only element: … “They have excellent contacts all over the place.” … “They have knowledgeable local resources in virtually every part of the world.”⁴¹

Some reports on industrial policy measures, including some reports from the Industry Evaluation Unit, paid attention to the issues of deadweight and displacement. In this context, displacement means the degree to which output from an assisted firm displaces output from another existing firm. Deadweight means the degree to which increased output or employment in an assisted firm would have happened anyway in the absence of assistance under policy measures.⁴² Lenihan (2003/2004) and Lenihan et al. (2005) drew together the findings of literature on this issue in Irish industrial policy, concluding that estimates of displacement were generally low, from usually around 3 or 4% up to 12%. Estimates of deadweight were a good deal higher, mostly at around 45–60%.

To be clear, a level of deadweight at, say, 50% means that half of the expansion seen in assisted firms was attributable to the assistance in the sense that it would not have happened without the assistance. The other half would have happened anyway, without assistance. Consequently, the policy measure concerned did have real positive effects, but those effects might potentially have been achieved at less cost if it was possible to identify in advance some of the assistance that was going to be unnecessary. Although a level of deadweight at around 50% may seem high, the positive effects of industrial policy measures were probably still crucial in generating the growth seen in indigenous industry.

For example, employment increases between 1988 and 1989 in new and expanding indigenous manufacturing firms amounted to 11.5% of 1988 indigenous employment. At the same time, employment reductions in declining or closing indigenous firms amounted to 9.8% of 1988 indigenous employment (Census of Industrial Production 1989, Table 5). Consequently, there was net employment growth of 1.7%. The same source also indicates that assisted firms accounted for a large majority of the increases, although the data on this are incomplete. If we take it that at least 60% of the increases came from assisted firms, their increases amounted to at least 6.9% of 1988 employment. If the level of deadweight was 50%, about half of that 6.9% would have gone ahead without assistance but the other half would not have gone ahead without assistance. Consequently, if there had been no assistance under industrial policy measures, total increases would have been about 8.1% of 1988 employment rather than 11.5%, so that employment reductions at 9.8% of 1988 employment would have outweighed the increases resulting in net decline by about 1.7%.

If we take the example of 1994–1995, when growth was a good deal stronger, a similar calculation indicates that actual indigenous manufacturing employment growth of 5.3% would have been only 1.2% in the absence of assistance under industrial policy measures.

It can be concluded that policy measures were quite effective and had substantial results in some significant respects. However, earlier findings of this chapter showed that ultimately the overall outcome for indigenous development was not very satisfactory. Specifically, the trends in indigenous manufacturing looked strong at first in the late 1980s, including exports. Then in the 1990s export trends continued to be good in the high-tech and medium–high-tech sectors, but overall indigenous manufacturing exports looked relatively weak by international comparisons. In the 2000s, export trends also became weaker in the high-tech and medium–high-tech sectors. In indigenous services, the trends mostly remained better for longer and exports continued to grow relatively fast by EU standards. However, when we focused on the software industry, there were signs of growth slowing down a good deal in the 2000s after an earlier period of extremely rapid growth.

In general, it seems that the industrial policy system had considerable success in promoting the growth of small and medium-sized firms particularly in the high-tech and medium–high-tech sectors, and in enhancing R&D and innovation. But such success continued only up to a certain point. Few of the companies concerned in the higher-tech sectors became large, and the more prominent ones often became takeover targets for foreign MNCs, so that the overall growth momentum tended to weaken over time.

In order to achieve more lasting and sustainable success in indigenous development, it would have been necessary for industrial development agencies to invest much more substantially in developing the scale and capabilities of some promising companies. For example, equity investment in companies by the agencies was generally limited to a 10% share, but that would have needed to be a good deal larger in selected cases. A related problem was that agencies had no way of preventing foreign takeovers and there needed to be a way of deterring that trend, whether by having larger shareholdings or by some other means.

Before concluding this section on industrial policy, it should be mentioned that a commonly stated goal of policy in the 1990s and later was to develop clusters of linked and related industries, with a view to enhancing competitiveness and growth. In the context of Ireland, this issue often involved both foreign-owned and Irish indigenous industries, and the relations between those two groups. This issue is not discussed in the present chapter as it will be more convenient to consider it in the next chapter on foreign-owned companies in Ireland.

5.5 Conclusion

Against a background of prolonged weakness before the boom, there was a substantial improvement in the growth and development of indigenous manufacturing in the late 1980s. Its performance became more uneven later, with some strong points as well as some weak points which grew more evident over time. In indigenous services, the trends mostly remained better for longer, as exports continued to grow relatively fast by EU standards up to the end of the boom.

Industrial policy was partly responsible for the improvement that occurred, since a series of changes were made in policy for indigenous companies beginning in the mid-1980s and the new policy measures were quite effective in some significant respects. However, the overall outcome for indigenous development was ultimately not very satisfactory. There was considerable success in promoting growth of small and medium-size firms particularly in the high-tech and medium–high-tech sectors, and in enhancing R&D and innovation. But the success did not continue beyond a certain point. Few of the companies concerned in the higher-tech sectors became large and many of the most prominent ones were taken over by foreign MNEs, so the growth momentum tended to fade over time.

The indigenous contribution to growth over the whole course of the boom was a good deal less than the contribution of foreign-owned companies, when assessed in terms of growth of exports and net foreign earnings. Consequently, indigenous companies’ share of exports declined from about 30.0% in 1985 to about 13.4% in 2007, while their share of net foreign earnings declined from about 44.2% in 1985 to about 30.4% in 2007. Thus, the indigenous contribution to net foreign earnings remained much greater than it appeared to be in terms of exports, but the indigenous share of net foreign earnings was declining due to faster growth among foreign-owned firms.

The indigenous contribution to the boom was particularly significant at certain times. It was relatively important in the late 1980s and in the 2000s, whereas the foreign-owned contribution was very dominant in the 1990s. Thus, the indigenous contribution was influential at the time when the economy was pulling out of the lengthy recession of the 1980s and embarking on a prolonged period of rapid growth, and it was also influential again later in maintaining a relatively high rate of growth well into the 2000s.

In fact, the improvement in performance in indigenous industry in the late 1980s was probably the most important change that got the boom started at that time. The value of exports from foreign-owned industries was already growing fast before the boom, whereas the trend in indigenous exports was much weaker. Consequently, when indigenous manufacturing export growth then accelerated to more than match the growth of exports from foreign-owned industry, that was a major turnaround that changed the trajectory of the economy.

Appendix: Data and Estimates for This Chapter

5.1.1 Table 5.1

In Table 5.1 the data for 1993, 2000 and 2006 come from the Census of Industrial Production (CIP). There was a break in the CIP data series between 1990 and 1991 because of a change in the industry classification system from NACE 70 in 1990 to NACE REV.1 in 1991. This involved virtually no change in the definition of total manufacturing employment, but as explained in O’Malley (1998), it did apparently increase the Irish indigenous share of the total by about 1.6 percentage points, to judge from the 1990–1991 trend in the IDA/Forfás Employment Survey. Therefore, the shares of Irish indigenous and foreign-owned companies in total CIP manufacturing employment in 1988 were adjusted by this amount to obtain the 1988 figures in Table 5.1, in order to produce consistency between the 1988 and 1993 figures in Table 5.1. Since the CIP did not include data distinguishing firms by nationality of ownership in the early 1980s, the figures for 1980 in Table 5.1 were derived by applying the 1980–1988 change in indigenous and foreign-owned manufacturing employment from the IDA/Forfás Employment Survey to the 1988 figures in Table 5.1.

There was another change in the industry classification system in the CIP from NACE REV.1 in 2001 to NACE REV.1.1 in 2002. This involved virtually no change in the definition of total manufacturing employment and it also involved virtually no alteration to the Irish indigenous share of the total, to judge from the 2001–2002 trend in the Forfás Employment Survey. Therefore, the CIP data for 2006 were used in Table 5.1 without requiring any adjustment to produce consistency with the data before 2001.

The CIP expanded its coverage somewhat in 2007 compared with 2006 which led to a discontinuity in the series. Since the expanded coverage brought in more relatively small firms, these additional firms were probably mostly indigenous, giving an artificial boost to apparent indigenous growth in 2006–2007. Consequently, the figures in Table 5.1 (and Fig. 5.1) finish in 2006 rather than 2007.

5.1.2 Figure 5.1

In Fig. 5.1, the CIP figures for 1988–1990 were adjusted in the same way as the 1988 figure in Table 5.1, for the reasons outlined above with respect to Table 5.1. Figure 5.1 terminates with 2006 rather than 2007 for the reason mentioned above concerning Table 5.1.

5.1.3 Table 5.2

The change in the industry classification system in the CIP from NACE REV.1 in 2001 to NACE REV.1.1 in 2002 resulted in no visible difference for manufacturing sectors at the NACE 2-digit level which is presented in Table 5.2. Therefore, the CIP data for 2006 were used in Table 5.2 without requiring any adjustment to produce consistency with the data before 2001.

5.1.4 Dairy Products Exports in 1986–1990

It was mentioned in footnote 6, in the “Exports and Net Foreign Earnings” part of Sect. 5.1, that in 1986–1990 there was particularly rapid growth in the exports of the dairy products sector. This sector was largely Irish-owned and it accounted for about one-fifth of all indigenous manufacturing exports. Unfortunately, there is room for doubt about the accuracy of the data on exports of dairy products. The doubt arises because the Central Statistics Office warned that one should be cautious about using export data on the food industries because respondents to the CIP might vary in the extent to which they interpreted sales into EC Intervention and to An Bord Bainne as exports. The reason why their interpretation could vary was because products sold by producers to EC Intervention or to An Bord Bainne were being sold to a purchaser in Ireland who would be expected to sell them on later to export markets. Since dairy products exports were quite an important component of total indigenous manufacturing exports, this doubt about data on dairy products exports raises the question of whether total indigenous manufacturing exports really grew as fast as they appeared to in 1986–1990.

While it is not possible to give a very precise answer to this question, it is worth bearing in mind that the data on the total output of dairy products should be just as reliable as any other data. The only reason why the export data might be inaccurate is because of possible flaws in the allocation of the output data between exports and domestic sales. Such inaccuracies could result in a false appearance of rapid export growth during a particular period if there were a large increase in misallocation of output data towards exports during that period.

Given this situation, it is significant that the output of dairy products really did grow very rapidly in 1986–1990, by 19.1% p.a. measured in current Irish pounds. The growth rate of exports of dairy products, at 21.6% p.a., was not a great deal higher. Thus, even if there had been no increase in the proportion of output being reported as exported, the rate of growth of exports would still have been high. Just in case the faster growth of exports reflected an increased misallocation of output data to exports, we can consider what would have been the outcome if dairy product exports had grown at only the same rate as dairy product output (i.e., export growth at 19.1% p.a. instead of 21.6% p.a.), so that the percentage of output being exported remained unchanged at 56.6% in both 1986 and 1990. In that event, total indigenous manufacturing exports would have grown by about 11.4% p.a.⁴³ measured in current Irish pounds, instead of the recorded rate of 12.2% p.a. This would still have been close to the 11.9% p.a. growth rate of exports from foreign-owned industry in Ireland. Under the same assumption, total indigenous manufacturing exports would have grown by about 17.5% p.a. valued in current US dollars, which would still have been faster than the growth of exports from the EU-15 at 15.2% p.a. or the OECD at 14.2% p.a.

It is mathematically possible that the true rate of growth of dairy product exports was even lower than the rate of growth of output, so that domestic sales grew faster than exports, but that seems highly unlikely in reality. It is difficult to see how the rapid growth of sales of food items such as dairy products could have been led by particularly rapid growth of sales into a mature domestic market that had no population growth at that time. It is far more likely that rapid growth of sales of such food products had to be led by rapid growth of sales into new or expanding export markets.

5.1.5 Net Foreign Earnings

Appendix in Chapter 4 explains the procedure used to derive estimates of net foreign earnings for each sector—with “net foreign earnings” being defined as the value of the sector’s exports minus the value of the imported inputs that are contained in the exports, minus the profit outflows that arise from production of the exports. This section outlines how that procedure was extended for this chapter to estimate the net foreign earnings of indigenous and foreign-owned companies in each sector.

As explained in Appendix in Chapter 4, the sectoral export data were taken from the official input–output tables for the year concerned, while the input–output tables also provided the data that were used for each sector’s imported inputs. For this chapter, those numbers had to be broken down into exports of Irish indigenous companies and foreign-owned companies, and imported inputs used by the Irish indigenous companies and foreign-owned companies. Appendix in Chapter 4 also outlined how each sector’s profit outflows were estimated. Since the outflows were of relevance only for foreign-owned companies, those numbers were used unchanged when estimating net foreign earnings of foreign-owned companies for this chapter.

To break export data down into indigenous and foreign components, the general principle, in the case of manufacturing sectors, was to use data from the Census of Industrial Production (CIP) on exports by sector and nationality. The export data for each sector from the input–output tables were divided into indigenous and foreign components in proportion to the indigenous and foreign shares of the sector’s exports in the CIP.

While that was a general principle, and it could be applied for some sectors in some years, there were also some sectors in some years where the necessary data on exports by sector and nationality were not available in the CIP. In those cases, CIP data from the closest year that had the necessary data were used to calculate exports as a percentage of gross output for the sector and nationality concerned. Then those percentages were applied to gross output data for the relevant sector and nationality in the year required.

In the case of 2007, no input–output tables were published for that year. The estimates of indigenous and foreign manufacturing exports for 2007 were derived by applying the growth seen in those two categories in the CIP in 2005–2007 to the estimates derived for 2005 as outlined above.

The method used to break export data down into indigenous and foreign components in the case of services was mostly analogous to the method for manufacturing, with the Annual Services Inquiry (ASI) taking the place of the CIP. Since the ASI had no data by nationality of ownership before the 2000s, we did not attempt to make estimates of services exports by nationality before that time.

The ASI had no data on exports by nationality of ownership before 2008, although it did have data on turnover by nationality for earlier years in the 2000s. Therefore, we used data on exports as a percentage of turnover by sector and nationality from 2008, and then applied those percentages to turnover data by sector and nationality from earlier in the 2000s, in order to obtain the required export estimates for earlier in the 2000s. This meant that we were assuming that about the same percentage of output was exported in 2000 and 2005 as in 2008. To give some assurance that this was a reasonable assumption, it can be noted that the Annual Business Survey of Economic Impact (ABSEI) indicates that exports generally were a stable percentage of sales in 2000–2008 in the selected internationally traded services covered in that survey, at 38–42% for indigenous companies and 92–94% for foreign-owned companies.

The ASI did not cover financial services and insurance. Instead, we used some figures from Accenture (2010) on exports from Enterprise Ireland client companies (i.e., Irish-owned companies) in financial services, and on the scale of employment in international financial service activities in the main Irish banks. Together with export growth trends in Enterprise Ireland client companies from the ABSEI, this enabled us to estimate that indigenous companies accounted for about 10–11% of the financial services exports in the input–output tables for 2000 and 2005. With additional guidance from export growth trends in the Balance of Payments, our estimate for 2007 was similar. The estimates for this sector would be less reliable than for other service sectors.

As regards imported inputs, Appendix in Chapter 4 already explained how the value of imported inputs that were contained in each sector’s exports was derived from input–output tables. For this chapter, those numbers had to be broken down into imported inputs contained in Irish indigenous companies’ exports and imported inputs contained in foreign-owned companies’ exports.

For that purpose, data from the ABSEI on sales and imported inputs, by sector and nationality, were used to calculate imported inputs as a percentage of sales for each sector by nationality of ownership. Then those percentage figures were applied to the estimates of exports by sector and nationality, derived as outlined above, to obtain estimates of imported inputs contained in exports, by sector and nationality. Finally, the data that were derived directly from the input–output tables on imported inputs contained in exports, by sector, were divided into indigenous and foreign components in proportion to the indigenous and foreign shares of our estimated imported inputs contained in exports by sector derived using the ABSEI data.

An exception was made in the case of financial services because it was found that there was a particularly low representation of foreign-owned financial services companies in the ABSEI. Instead, since our export estimates had indicated that foreign-owned companies accounted for almost 90% of all financial services exports, we simply took it that imported inputs as a percentage of all financial services output derived from the input–output tables were sufficiently representative of foreign-owned financial services companies. As indigenous financial services companies were quite well represented in the ABSEI, we used the same method for them as for all other sectors.

For 1985, the data used to estimate imported inputs by nationality for the manufacturing sector were taken from an earlier forerunner of the ABSEI called the Irish Economy Expenditures Survey. Results of that survey were reported in Industrial Development Authority. (1985). The Irish Economy Expenditures of the Irish Manufacturing Sector. Dublin: IDA, and the results relevant to our purpose here were reported in O’Malley (1989, Table 7.3).