From YalePressWiki

[edit] Chapter 4: A Misunderstood “Miracle”: The State and the Growth of the IT Industry in Ireland

Trying to induce private entrepreneurship in Ireland during the 1980s made me feel like I was part of the old lobster joke: A young man goes to the beach, he sees an old fisherman with two untied lobsters in a very low bucket. He comes to the fisherman and tells him, “Sir, your lobsters are not tied and the bucket is very low; you have to tie them or they will run away.” And the fisherman answers, “Dear lad, do not worry, these are Irish lobsters! As soon as one of them almost makes it to freedom the other jumps and drags him back down.”
—A former executive of the IDA’s Enterprise Development Plan

If one nation’s economic performance in the 1990s has inspired a revolution in international perception, that nation is Ireland. As late as 1995 the common perception of Ireland’s economic performance was one of a continuous failure, which doomed Ireland to be the perpetual basketcase of Europe (Guiomard 1995).^[1] Less than four years later, sick Ireland seemed to have become the roaring Celtic Tiger (Breathnach 1998, MacSharry and White 2001, O’Hearn 1998, Sweeny 1999). Such Irish software companies as Iona in middleware, Smartforce and Riverdeep in education, or Trintech and Baltimore in data security not only achieved global success but also made their founders and many of their employees rich when they publicly listed on NASDAQ. This success made software the first industry in Irish history to create such a virtuous cycle of entrepreneurship, and helped to transform even the Irish people’s perception of themselves.

Indeed, Ireland has become the most invigorated economy of western Europe, a model for the EU’s new entrants, with one of the world’s most successful IT industries. This achievement has been attributed largely to the Irish developmental agencies that, so it is argued, promoted this accelerated growth by utilizing a series of industrial and S&T policy initiatives. Moreover, this impressive feat has been achieved while both politicians and bureaucrats have strongly adhered to a neoliberal interventionist ideology, an ideology significantly different from Israel’s or Taiwan’s.

However, even a brief analysis of the Irish IT industry reveals dissonance between the operations of the foreign multinational corporations (MNCs) and those of the indigenous industry. On one hand, the MNCs, a cornerstone of Irish industrial policy since the 1950s, are thriving and delivering great economic growth. Dell, an American MNC, has just surpassed Microsoft, another American MNC, and become the biggest Irish exporter, singlehandedly responsible for 5.8 percent of Irish GDP in 2003.^[2] This is an impressive statistic, and shows that even with rising costs, Ireland is still central to the global strategy of leading IT MNCs. However, neither Dell nor Microsoft, together accounting for about 10 percent of Irish GDP, has significant high–addedvalue activities, such as R&D and product design, in Ireland. This leads many commentators to suspect that behind the impressive Irish GDP and trade figures lies even more impressive showmanship in the form of transfer pricing accounting, utilizing Ireland’s low corporate tax regime and the availability of a well-educated labor force.^[3]

On the other hand, in the indigenous sector, it is the software industry that has reached global prominence, not the hardware and electronic-manufacturing sector favored by the Irish FDI-based policy. Furthermore, while the state claims great credit in sponsoring the growth of the software sector, it has started to sponsor it systematically only after a few Irish-owned companies achieved global success. Moreover, these state programs appeared to suffer from a slight misconception in regard to their goals and methods, especially in the past four years.^[4] The result is that the industry’s growth has been brought to almost a complete halt since 2001, with almost no new companies managing to shore up annual revenues of more than $2 million.

Nonetheless, Ireland’s state-led transformation from a country that in the 1950s had the least-educated workforce in western Europe to one of the world’s top IT producers at the end of the 1990s is nothing but miraculous. Hence we have to suspect that this picture of dissonance is evidence of a misunderstanding— a misunderstanding of what exactly happens in the Irish miracle, and a misunderstanding of what was, and what should be, the role of the state in the development of the indigenous industry.

Utilizing our theoretical framework and comparing Ireland to Israel and Taiwan, in this chapter I offer an interpretation of this dual misconception. What are the causes behind this transformation? What enabled Ireland, the laggard of western Europe, to become the EU’s IT industry “shining light” in the time span of one generation? And what are the weaknesses, strengths, and generalizability of the Irish experience?

Throughout this analysis I make the following arguments: first, both the IT industry and the Irish S&T industrial policies evolved within a political system in which the state has had one overarching goal—job creation—a significantly different circumstance than in Israel and Taiwan. In both Taiwan and Israel the primary goal of their respectively very different S&T policies was the creation of new industries. Job creation was only one of the derivative aims of state efforts. This intrinsic difference in policy aims led Ireland to a specific process of state-industry co-evolution with markedly different outcomes in terms of the IT industry’s capabilities, structure, and position within the global production networks from those of Taiwan and Israel.

The focus on job creation led the Irish state, following an ideology of “neoliberal interventionalism,” to develop an FDI-based industrialization policy. The Irish neoliberal interventionalist ideology is based on two conflicting principles around which the Irish state has been formulating its economic and industrial policies since 1958. On one side, the state, in the levels of both government and the developmental agencies, possesses a strong free-market ideology. On the other side, it also sees a large role for government in enhancing the competitiveness of Ireland and strategically managing its growth. This ideology is radically different from those of Israel and Taiwan, where the state’s ideology, until very recently, was a variant of market socialism.

The state, led, from the late 1950s until the early 1990s by the autonomous governmental agency created specifically for the purpose—the Industrial Development Authority (IDA)—was focusing its attention on bringing MNCs to open manufacturing plants in Ireland and on upgrading the physical and educational infrastructure. Throughout this period the Irish state paid relatively little attention to the development of industrial R&D activities in general and indigenous IT industry in particular.^[5] Moreover, the state, true to its neoliberal ideology, did not apply pressure on the MNCs to embed with the local industry.^[6] This lack of interest in the particular needs of rapid innovation– based industry, coupled with only minimal attempts of the IDA to nurture, embed itself, and network with Irish-owned industry, hurt the early development of the Irish IT industry, in particular the hardware sector.

Even in the 1980s and 1990s, after the IDA and its successors, Forfas, IDA Ireland, and Forbairt/Enterprise Ireland, started their active involvement in the creation of the indigenous IT industry, they were neither technologically savvy nor deeply embedded within the IT industry, nor, for many years, did they view indigenous IT industry development as a task of creating rapid innovation– based industry. Industrial policy has been planned as a derivative of the goal to enhance employment through the creation and growth of enterprises broadly defined. Only in 1991, when industry pressure and the availability of EU funds led to the creation of a new specific subunit within the IDA—the National Software Directorate, whose employees were recruited directly from industry— did the state start to pay continuous attention to the particular needs of R&Dbased IT industry. Yet still the policy focused almost solely on software.^[7]

Accordingly, my second argument is that the case of the indigenous IT hardware industry in Ireland has been a clear case of stage one failure—the inability or unwillingness of the state to facilitate and participate in the growth of the indigenous hardware sector and its networks.

Interestingly, software was the recipient of focused state policy specifically because it was defined as a tradable service and not as an industry. During the 1980s, with Ireland once again in severe economic crisis, the IDA changed its policies in a few critical domains. One of the changes was to step up the indigenous industry–oriented activity with the creation of the Enterprise Development Program (EDP). The second change was to create a new focus on tradable services on both the indigenous and FDI-based sides of its operations, with software quickly becoming the prominent sector of this new policy.^[8] Seen as a service, however, these policies did not take into full account the particular needs of the software sector as an R&D-intensive industry. S&T industrial policies were enacted only after the industry achieved worldwide success. Therefore my interpretation of the role of the Irish state is much less sanguine than that of some writers, such as Sean O’Riain, MacSharry, and others, while at the same time unlike O’Hearn, I still view the state as the critical, and positive, actor in the Irish economic miracle (MacSharry and White 2001, O’Hearn 1998, O’Riain 2004, Sweeny 1999).

[edit] Historical Background and the Beginning of the IT Industry

After gaining independence, the Republic of Ireland passed through two almost complete U-turns in its industrial policy, and a subtler, but arguably at least as important, transformation since the beginning of the 1990s. Starting with its independence in 1921 and for the first decade of its existence, after conceding to England most of the industrial base, which was concentrated in the northern counties around Belfast, the Irish republic was almost solely an agriculture-based economy. Led by William T. Cosgrave of the Cumann Na nGaedheal party, the newly established republic followed an economic policy focusing solely on the agriculture sector, and was fiercely free trade in its economic ideology. Patrick Hogan, the then minister for agriculture, is famous for describing this policy as “helping the farmer who helped himself and letting all the rest go to the devil” (Haughton 2000). The main elements of this policy were free trade, parity of the Irish currency with the English pound sterling, low taxes, and low and modest government spending and intervention. This policy gave the Irish an almost unbeatable claim to the title of the most conservative revolutionaries in history.^[9]

In 1932, in the midst of the global depression, staging a major political transformation in Ireland, Fianna Fáil—the republican or nationalist party— won the elections.^[10] Fianna Fáil, led by Eamon de Valera, with Sean Lemass serving by his side as minister for industry and commerce, made the first complete reconstruction of Ireland’s economic policy and devised a highly nationalistic policy based on the ideal of an autarkic market. Free trade was abolished, high tariffs were put in place, and a proxy economic war was waged for a very high economic price against England over land annuities. While some debate still exists on whether these economic policies were truly damaging to Ireland or only slightly so, one fact remains clear. The economic gap between Ireland and the United Kingdom grew, and so did the gap between the North of Ireland and the Irish Republic. Moreover, net emigration continued at an alarming pace and an overall pessimism about the future of the republic as an independent state was deeply felt.^[11] At that time, at the request of Sean Lemass—who became prime minister (taoiseach) after de Valera—Ken Whitaker, Ireland’s most prominent civil servant and the general secretary of the Ministry of Finance, published in 1958 Ireland’s first comprehensive economic policy document, the celebrated Economic Policy (SO 1958a).

In this report, Whitaker devises the main points that continue to shape Irish economic policy to this day: export-oriented industrial policy tied with free trade and conceptualization of Ireland’s main economic threat as severe unemployment leading to alarming large net emigration.^[12] It is important to put Irish unemployment in context in order to understand the severity of the Irish situation, and the long-established national preoccupation with jobs. For example, even throughout the 1980s and until 1997 Irish national unemployment rates were higher than 10 percent, reaching 15 percent throughout most of the period, with unemployment outside the Dublin metropolitan area significantly higher. Moreover, these high rates continued even with exceptionally high rates of emigration throughout the same period. By 1991, for example, 20 percent of the Irish male aged 15–19 in 1986 had left the country, and in 1989 alone 1.1 percent of the Irish population emigrated in search of jobs (Walsh 1999).

The publication of Whitaker’s Economic Policy also marks the crystallization of the Irish neoliberal interventionalist ideology. The main goal of the first economic program and of all those that followed has been job creation. Since 1958 there have been three basic elements in Irish economic policy: provision of economic incentives and low corporate tax rates to induce industrial development and investment, transition to free trade, and—the cornerstone of Irish industrial policy until the mid-1990s—attraction of MNCs (mostly American) to locate export-oriented manufacturing activities in Ireland. The most important vehicle for institutionalizing this policy was the strengthening of the Industrial Development Authority, as an autonomous agency responsible for job creation through industrialization. The IDA mandate was to create more jobs with the aim of decreasing the cost per job created through time.

While not focusing on the MNCs sector in its inception, the IDA quickly adopted FDI-based industrialization as the main focus of Irish industrial policy. Using the yardstick of job creation, one can understand the lure of bringing large corporations that could promise the almost instantaneous creation of substantial and quantifiable numbers of jobs, a situation in which employment creation–tied grants make sense to both parties.^[13] The IDA rapidly grew in its importance and influence throughout the years and transformed itself into a full “one-stop shop” for attracting and locating MNCs.

In doing so the IDA utilized a double strategy. On one side targeting and approaching MNCs, employing a wide array of techniques to befriend their senior executives, the IDA operated as a marketing agency that sold “Ireland” to MNCs, especially in the United States. The IDA quickly established a network of foreign offices whose job was to create long-term relationships with companies (many times when these companies were still in their early growth phase) that would eventually result in these companies’ opening their European facilities in Ireland. The IDA quickly moved to target sectors and internally employed a set of targets based on the number of jobs created and companies moved in each sector.

On the second front, the IDA was busily orchestrating the construction of the infrastructural and financial benefits Ireland could offer to the incoming MNCs. Ireland, like Taiwan, claims to have been the first state to open a duty-free export processing zone, around Shannon International Airport in 1958. Moreover, soon thereafter many of these tax incentives were applied to the whole republic. In 1958 the Export Profit Tax Relief provided 100 percent tax relief on export profit growth—that is, a full tax relief for new exportoriented facilities. In the beginning these policies were similar to Taiwan’s. However, in 1964 the neoliberal interventionalist ideology of Ireland sharply moved Ireland down a different path. All restrictions on foreign ownership and reparation of profits were eliminated, no pressures were put on MNCs to source locally, and the idea of demanding, or even asking, MNCs to transfer knowledge to local component suppliers was not even floated.

As part of making itself into a one-stop shop for MNCs, the IDA became the biggest industrial land developer and owner in Ireland, and took it upon itself to cater to the needs of the MNCs even after they had already moved to the country, representing their interests within the Irish state. Unlike local enterprises, MNCs coming to Ireland have been offered land, complete factories, and one centralized agency that represents the MNCs in all their dealings with other state and local agencies.

The IDA reached the zenith of its power in the 1980s. Not only was the IDA formally semi-independent (defined as a public agency, it was not part of the civil service), which allowed it to act as a bridge between other governmental ministries and departments, but its policy domain also included everything that related to industrial development broadly defined. Its strong political connection to MNCs and responsibility over local industry gave it an immense political clout. This was especially true in the period until the early 1990s, when the IDA was seen as effective in job creation—the national mission at the time.

The fact that the IDA was responsible for the recruitment of MNCs as well as one of the largest landowners in Ireland was especially important. In the context of the Irish electoral system—a single-transferable-vote system in which Ireland is divided into districts, each electing three to five members of the Dáil—the IDA was perceived to have significant influence on MNCs’ ultimate location decision within Ireland. Thus in terms of bringing the “pork” back home to specific constituencies, IDA’s decisions carried a lot of weight on the economic future of specific localities, and accordingly on the future of their elected officials. To further enhance its political muscle, the IDA was actively organizing the local managers of the MNCs into a coherent lobbying group around its policy initiatives (O’Riain 1999). The IDA also benefited by having internal cohesion and extremely high-skilled personnel. The agency recruited the top high school graduates into a long career pattern within it, in which they were also given opportunities to acquire higher education. Hence it was stuffed with highly skilled, very motivated, and extremely loyal personnel who were able to use the IDA’s semi-independence to make the agency into a crucial policy entrepreneur by utilizing the structural holes of the Irish bureaucracy to attain its own goals by connecting different organizations with weak social ties (Burt 1992, Granovetter 1973).

This policy turnaround, coupled with the MNC-focused industrial policy, followed by membership in the European Community since 1973 and entry into the European Monetary System in 1979, led to relative economic growth until the 1980s. However, while manufacturing output growth was quite stable throughout the years, employment growth stayed slow and even took a dip in the 1980s. Moreover, the macro-level figures hide a negative facet of Irish economic performance until the 1990s: the huge jobs turnaround, as a large number of jobs were lost, especially in Irish-owned firms.^[14] That is, at the same time many jobs were created a very high number of jobs were destroyed. Consequently, for many Irish the prospects of stable job tenure, and with it secure economic prosperity, remained low.

[edit] Building the Human Infrastructure: The Buildup of the Irish Education System

At the same time, Ireland was passing through a much more subtle transformation, a change arguably at least as important as any for the future growth of the IT industry. Within two generations Ireland went from being the western European country possessing the least-educated workforce to a nation with one of the most highly educated young workforces in the world.

Until 1966 less than 50 percent of the Irish population enjoyed education above the sixth grade. However, that situation quickly changed after the minister of education, Donogh O’Malley, following a heated debate started by a 1965 OECD sponsored survey, Investment in Education, decreed that secondlevel education was to be given free to all Irish children. While publicly Lemass conveyed utter surprise at the move, some argue that O’Malley had the blessing of Lemass, who hoped to circumvent a public debate in which the Ministry of Finance might delay or prevent this change from taking place. In an interview, Ken Whitaker, then the general secretary of the Ministry of Finance, strengthened this assumption, stating that he felt a need to protest only to find that Lemass not only was approving of O’Malley’s “surprise” decree but was quite prepared for it.^[15]

The OECD influence was also critical in the early 1960s in the establishment of the regional technical colleges, which have been recently upgraded to institutes of technologies (OECD 1965, 1969). This was the first in a series of changes that overhauled the Irish higher education system, first in the 1960s and continuing in more vigorous fashion in the 1980s after the IDA included education policy in its overall view of the supply side of Irish industrial policy.^[16] In the 1950s only 33 percent of Irish school leavers finished secondary education, and only 10 percent had any experience of third-level education.^[17] By the mid-1990s these levels had grown to 80 percent and 50 percent, respectively, and the percentage of under-twenty-eight-year-olds attending college in Ireland is one of the highest among OECD countries (Breathnach 1998).

The role of the IDA in this transformation was immense. In the 1980s, when its FDI-attracting focus changed from manufacturing activities to more technologically advanced activities in general and the IT industry in particular, the IDA realized that Ireland needed a much more technologically skilled workforce in order to attract the MNCs it targeted. Using its influence, the IDA first managed to bridge the gap between the universities and the government that had delayed the improvement of the higher education system, especially research, for more than a decade. This feud between, on the one hand, UCD and Trinity, the two prominent Dublin-based colleges, and, on the other, the Irish government, began in the 1960s. The government tried to force the merger of the two institutions and to establish a firmer control over the merged entity. The plan was quickly abandoned, but animosity between the political elite and the two leading Irish universities remained until the 1980s. In retrospect, this bungled move by the government was one of the most important to the creation of the Irish software industry. Immediately after the government announced the plan, under which Trinity’s engineering faculty was to disband and move to UCD, Trinity moved to counter it by creating and rapidly expanding a new engineering department of computer science in 1969. UCD followed suit by expanding its own computer science program located in the sciences. These two departments became crucial to the development of the Irish IT industry in later years.

Second, the IDA managed to accelerate changes that upgraded the whole education system. Within a decade and a half the regional technical colleges became full-fledged academic establishments, the National University of Ireland system was expanded, and a few more higher education institutions became full-fledged universities (White 2001).^[18]

There were major differences between the Irish education system and policies and those of Israel and Taiwan. Education in Ireland was seen as an economic supply-side issue, and consequently the system’s main aim was to create highly skilled labor that could work for the incoming MNCs.^[19] Indeed, even today, if a regional college wishes to open a new degree course, it has to show that the industries around it not only want these skills but also approve the specific curricula offered. Many of the managers of the large MNCs whom I interviewed, such as those of Intel, Digital, or Ericsson, talked at length about the involvement of their corporation in the development of the nearby colleges’ course of studies.^[20]

In one important respect, quality and quantity of academic research, Ireland is still far behind Israel, Taiwan, or most OECD countries, though very recently a small positive change appears to have occurred.^[21] In the 1970s the state supplied very minimal resources for academic R&D, about £27 million (IEP), and during the financial crisis in the 1980s even that amount dipped sharply to almost nothing. EU grants became the main source of capital. Only in the mid-1990s, with the looming future expansion of the EU, did policy makers and politicians realize that Ireland must upgrade its R&D capabilities if it was to compete with the eastern European countries for MNCs.

These developments triggered the establishment of Science Foundation Ireland (SFI) and the Programme for Research in Third-Level Institutions (PRTLI) in 1998.^[22] SFI has been built on a similar basis to the American NSF. Indeed its first director general, William Harris, an American of Irish origin, was formerly a director of one of the NSF divisions. SFI was envisioned as a five-year targeted effort to create high-quality research in the areas of biotechnology and ICT, but this focus was somewhat expanded. The SFI also continues another recurring theme of Irish economic policies: there is an intentional bias to award grants for teams that are led by foreign-based internationally renowned researchers. The argument behind this partiality has been that Ireland needs to attract high-quality researchers if it wants to conduct top-level research. While Ireland has the advantage over Israel and Taiwan of being a highly developed English-speaking country with a high quality of living that can attract foreign talent, others argue that this move symbolizes the inability of the Irish state to trust its own citizens. As of 2006, SFI’s budget for its first period of operation, 2000–2006, was 646 million euros (SFI 2006).

PRTLI is the twin effort of SFI, and its main aim is to enable the building of the necessary research infrastructure within the university system. PRTLI allocates its funds on a competitive basis. As of 2004 the PRTLI had three multiyear funding cycles. It was envisioned as a state-industry initiative. As of 2006, not including private contributions, 605 million euros had been allocated in three investment cycles starting in 1999 (HEA 2006).

By establishing these two programs in 1998 the Irish state indicated its seriousness about transforming the academic research infrastructure. However, there are still many caveats as to the continuous state support for SFI and its medium- and long-term influence. Moreover, even when finally moving to upgrade the Irish research apparatus, the Irish state still employs an FDI orientation— for example, in its insistence on the participation of foreign-based scientists. Hence questions remain as to whether the goal of the SFI is a genuine attempt to change the Irish economy capabilities toward rapid innovation– based industrialization or just the last stage of supply-side policies to create MNC-attracting human capital.

Thus, although the development of the Irish education system is unprecedented in Irish history and has greatly helped to propel Ireland along the track of high-skills IT industrial development, it appears that the strategic view behind these developments treated the education mainly as a way to produce high-skilled labor, not innovation and research. Only in the past five years have policies with a vision of making the Irish higher education system more research oriented been implemented.

[edit] Seeds of the IT Industry: From the 1960s to the Early 1980s

Throughout the same period the seeds of the future IT industry in Ireland were planted. In both the public and the private sectors, computerization activities were taking place that would eventually give birth to the Irish-owned IT industry. The official beginning of computerization in Ireland was in 1958. In that year Suicra—the state-owned Irish Sugar Company—decided to computerize its operations and brought the first computer to Ireland.^[23] Throughout the 1960s and early 1970s, big public and private companies and organizations were the main loci of professional IT training and knowledge diffusion, especially as formal third-level education was beyond the means of most young high school graduates at the time, and most universities didn’t offer computer-science degrees.^[24] Barry Murphy, the first director of the National Software Directorate and before and after CEO of a few leading Irish-owned software companies, such as Insight, Openet Telecom, and Netsure, echoed the claims of many that these institutions are the unsung heroes of the industry: “Those big companies were the only institutions that were actually training people at the time; these companies were the likes of the banks and insurance companies. They gave superb training to cohorts of young people, which they all lost in five to seven years’ time. These companies made an immense contribution to the Irish software industry” (interview, November 6, 2000).

The early Irish IT companies were established in the 1960s and 1970s. Most, due to the high cost of computers at the time and the capital constraints in Ireland, concentrated either on consulting and services or on outsourcing of IT services’ business models.^[25] The first Irish software company was System Dynamics, founded in 1968 by a team of former IBM Ireland employees and focused on software consultancy services. System Dynamics was also important for three other reasons: first, it spawned a few of the biggest Irish software consultancies, such as Chaco and Delphi, which in turn spawned others, such as Enovation.^[26] Second, System Dynamics’ founder Tom McGovern was a key actor in raising the profile of the Irish software industry and creating its identity as a distinct sociopolitical group within the Irish industrial landscape. Third, McGovern, in a way similar to other successful early IT entrepreneurs, such as Michael Peirce of Mentec, played a crucial role in the early financing of other, more product-oriented, companies. As other sources of finance did not exist, and as the Irish banks were not even willing to extend overdraft facilities to many software companies until the early 1990s, such a role must not be underestimated.^[27]

The Irish IT companies in that period, unlike the MNCs, had to surmount extensive difficulties in order to commence and maintain operations. Not only was there no source of funding, nor entrepreneurial financiers such as the Israeli Tolkowsky and Discount Investment, but banks were not even willing to extend loans or simple credit facilities to indigenous IT companies. The result has been that those few individuals still daring to be entrepreneurs typically needed to mortgage their own houses to secure working capital. As a result of these constraints, most of these early IT companies stayed financially fragile and collapsed at the slightest change to their monthly cash flow. Moreover, unlike the MNCs, local new technology–based firms (NTBFs) were not offered land and offices in the IDA’s industrial estates, and many property owners turned out to be as conservative as the banks, unwilling to rent space for software companies.^[28]

Slowly but surely, the first Irish companies moved on the backs of their customers to product development. Timing and the movement of the MNCs into Ireland were important catalysts in this development. In the 1970s the MNCs led the progress in Ireland away from the mainframe platform into minicomputers. The opening of the new and smaller Irish MNC subsidiaries coincided with the rise of minicomputer technology; as a result, Ireland became one of the world’s heaviest per capita users of minicomputers.

Because minicomputers were a new platform, they did not have an established base of software—and were incompatible with software employed at the MNCs’ foreign headquarters. As a result, many local companies managed to get development projects that could easily be packaged (a new concept at the time) into software products. The more successful of these packages quickly secured a few more sales in Ireland and the United Kingdom. However, a second and much more important transformation of the local software industry happened when the computer manufacturers, such as Digital, IBM, and ICL, invited some Irish companies to market their products together as bundled solutions.

Kindle, one of the first Irish software companies to reach successful constant overseas sales, is an example of this model of organizational development. Kindle, founded in 1981 as Triple A Systems, first developed banking systems on the ICL platform.^[29] ICL sold its computers around the world, especially in former British colonies, and did not have many banking packages that ran on its platform. It added Kindle to its official list of ICL software vendors as a banking package specialist. Soon thereafter, Kindle got orders from all of the former British colonies. In the second part of the 1980s, Kindle converted its products to other platforms. Kindle is now owned by a British company and is a five hundred–person software house with a global sales and support network. However, its main locus of activity and R&D is no longer in Ireland.

The early 1980s brought an overall maturation of the Irish software industry. This process coincided with the beginning of the changes in the IDA view of the indigenous IT industry and the tradable services industries. With the establishment of the Enterprise Development Program by the IDA, the agency started to become more proactive in assisting local IT NTBFs, though on a very low scale and with relatively small amounts compared to the aid given to MNCs. Furthermore, the IDA and many leading politicians developed a relationship of mutual distrust with the indigenous IT industry, a state of affairs that continued to the mid-1990s.

Apart from Kindle, two other software companies whose rise and fall changed the way the IDA and subsequently Enterprise Ireland managed their R&D grant schemes were Insight Software and Real-Time Software (RTS). The two companies, in particular RTS, were different from earlier Irish-owned companies in being product based from their inception. They were also representative of the particular profile of the Irish software industry at the time: the products of both were based around Maapics. Maapics was the main software suite for manufacturing facilities offered by IBM as part of its own foray into the minicomputer arena.^[30] Both companies grew out of the relative strengths of the Irish industrial base in the early 1980s—manufacturing facilities that were world leaders in utilizing the new minicomputer technologies.

Insight was originally the name of a package for financial reporting and analysis built to work on top of the Maapics package by Vector Software, a product-oriented venture of AMS. Insight was sold by AMS, which also financed all the development efforts from its IT consulting revenues. AMS started to directly sell the Insight packages around the world in 1980 and changed its own name to Insight Software in 1983. The company reached record product sales for an Irish company in the early 1980s, and by 1988 was poised to sell its more advanced software package for the IBM AS/400. However, the company started to suffer severe cash flow problems. These were caused by a series of factors: (a) the need to financially rely solely on its own revenues, (b) its use of a business model that necessitated a costly international direct sales operations, (c) the fact that Insight, with Vector acting as its R&D center, still maintained both the product line and the consultancy business, and (d) the added difficulties in moving into the higher priced software package on a newly launched platform. In the end the founders were glad to sell Insight to the Hoskyns Group, a publicly listed British software house, in 1988.^[31]

Real-Time Software was given more attention by the IDA, mainly because it was perceived to be the first technology-based spin-off to form from a subsidiary of foreign MNCs, and as such was celebrated as a political success story. The company was founded to develop software packages and tools around the Maapics software. Its seed finance came from it cofounders, one of whom then proceeded to sell his services as a consultant to finance the initial product R&D. Starting by offering multicurrency modules, RTS developed a whole suite of financial packages. By 1985 the company had revenues of a few million dollars and a chain of foreign offices. However, it still did not manage to secure investment anywhere in Ireland, and even though RTS managed to raise a relatively small amount of capital in the United States and the United Kingdom by the end of 1985, it was clear that its better-funded American rivals were winning market share. In 1986 the company suffered heavy losses that brought its founders to sell it to the American company MSA. Soon thereafter MSA itself faced difficulties, closing down its Dublin operations in 1988.

In the same period, many politicians as well as other parts of the IDA aired concerns that the EDP was investing too heavily in risky indigenous software businesses without generating enough employment. In the three years between 1986 and 1989 these concerns were strengthened when many of the leading software companies went bankrupt. In addition, as the result of the foreign merger and acquisition of RTS, Insight, and CBT, some of their founders made what was considered in Ireland at the time a small fortune. That these entrepreneurs made this fortune by selling their companies—perceived to entail the destruction of jobs instead of the creation of jobs—triggered severe resentment. Both the public and the politicians, only too willing to spend millions on MNCs, saw the newfound riches of the software entrepreneurs who benefited from relatively minuscule IDA grants not as a sign of success but as socially unfair gains. Moreover, the growing number of bankruptcies created a wariness about the chances that a local industry would ever be able to achieve continuous and sustained growth.^[32] In the end a decision was made that state aid should be constructed to share the upturn of businesses it invested in and not just their risks. Direct pressure from the then new minister of finance, Albert Reynolds, brought the IDA to introduce a small, later to be proved critical, change to the conditions of its grants: starting in 1988 the Irish state started to take equity stakes in return for financial aid.

The tales of these three companies also show the most significant difference between the Irish and the Israeli software industry. The Israeli software industry grew within an innovation system already focused on R&D. As a result, many of the products of its software companies were software development tools. The Irish software industry followed a different track, developing software applications for particular business domains, such as finance, insurance, or manufacturing control. Only with the rise of university research in the early 1990s would Ireland manage to grow successful software companies with a focus on developing software technology itself. However, when one such company, Iona, was listed on NASDAQ, it epitomized the new international success of the industry and completely changed its perception in Ireland.

An almost forgotten fact about these early years of the Irish IT industry is that the first IT companies that truly reached sustained global success were hardware, not software, firms. In the mid- to late 1980s the Irish IT hardware industry briefly enjoyed unprecedented success. The two prominent companies of that period were Lake Electronics and Mentec, and their development stories also explain why there were only a few Irish hardware companies until the mid-1990s.

Harry Lynam set up Lake Electronics in the end of 1976. Lynam, a physicist, worked for many years in the engineering part of the Post and Telecommunication Ministry, where Jim Mountjoy, another important entrepreneur, also gained his first experience. In 1976 Lynam become fascinated by microprocessors and decided to sell his house and move with his family to his inlaws to secure the seed finance for Lake Electronics. Luckily for Lake, in the same period the IDA established the EDP, and Lake became one of its first clients. An early employee of Lake described the effect of this state help as critical:

More than anything else at that period in Ireland [the IDA’s backing the company produced a positive] psychological effect. We were frowned on, seven adults with families leaving good secure jobs to start some adventure in a technological area nobody understood. The fact that the esteemed and all-powerful IDA thought that we were making something useful gave us the mental strength to continue. (interview, July 9, 2001)

That injection of capital from the IDA also transformed the way in which the banks treated the company, and together with a first project for Lynam’s former employer, the Ministry of Post and Telecommunication, Lake managed to develop its first product: a private-branch exchange (PBX).^[33] Lake secured a critical alliance with British Telecom (then a state-owned enterprise) in 1979. Building on its first major international sale, Lake soon expanded worldwide and by 1983 grew to become a two hundred–person–strong company with revenues of £10 million. However, the company suffered from lack of management skills and encountered recurring cash flow problems. The result of these unstable times was the sale of the company in 1988 to Landis and Gyr, a Swiss engineering MNC.^[34]

Thus 1988, the year in which the remains of RTS’s Dublin operation closed down and Insight Software and CBT’s first incarnation were bought by Hoskyns, became the year in which the first generation of international successful Irish IT companies suffered heavy setbacks and the image of the indigenous sector was tarnished.^[35]

Mentec’s creation story has many similarities to Lake’s. Mike Peirce, who was a lecturer on Trinity’s engineering faculty at the time, founded Mentec in 1978. Peirce’s expertise was in the use of computers in manufacturing. In 1978, together with one of his graduate students, he went on a study trip in Japan to see the use of robots in manufacturing. The trip proved to be a needed catalyst, and Peirce and his student-turned-partner mortgaged their assets to secure seed financing in the form of a bank loan. Fortunately for them the IDA’s EDP agreed not only to match their funds but also to guarantee 50 percent of a £100,000 bank loan. The fact that the IDA agreed to assist the company at the time was critical, according to an interviewee working for Mentec at the time:

We could absolutely not get started without the IDA; the banks were unwilling to talk with us before the IDA chipped in. As a matter of fact, we probably would not have even tried. When we floated the idea of opening a company, almost all the people we knew told us that we are crazy and that to leave secure respectable careers at Trinity is madness. Many reminded us that we have families and children to think of. (interview, June 10, 2003)

With the IDA’s backing, Mentec approached the banks. However, the banks insisted that Mentec buy “real assets” worth £50,000 to secure the 50 percent of the loan not covered by the IDA. Hence, Mentec’s first activity was the building of its own offices. The banks were pleased, but Mentec was left with no working capital and no way to finance R&D activities.

At the time the biggest IT MNC in Ireland was Digital Equipment of Massachusetts, which had large facilities around Galway. Mentec first focused on projects to develop custom hardware solutions around Data General’s computers, but switched allegiance to Digital in 1979. Digital, unlike the other MNCs at the time, had an active and developed OEM partnership program and was looking for suppliers to develop bundled solutions for its PDP computer line. In order to help its smaller suppliers, it gave them complete and open documentation of its systems. Mentec soon became Digital’s most important OEM supplier in Ireland. Focusing on product development, Mentec created its core product in 1982, a single-board computer-controller based around Digital’s J-11 CPU, to be used in manufacturing control systems. This basis led to many other developments around Digital’s and Mentec’s own chips that were fabricated by Texas Instruments. In its prime, Mentec had revenues of £10 million and employed more than two hundred people in Dublin and a few score more in the United States and England. At one point, sixtyfive of Mentec’s employees worked on R&D, a unique feature in the Irish IT industry of those days. Mentec also benefited from becoming an official partner of Ireland’s foremost MNC, and the IDA gave it more than a few R&D grants throughout the years.^[36]

The stories of Mentec and Lake Electronics exemplify the difficulties that Irish hardware companies had over and above the significant obstacles that software companies were facing. Not only is hardware product development a much costlier endeavor than software, but the opportunities for consulting projects, especially those that can ultimately lead to product development, were scarce. In addition, at that time both software and hardware companies had to face a hostile social environment that either saw technological entrepreneurs as risk takers when they failed, or resented them when they succeeded. Thus, whereas in Israel and Taiwan the pioneers of the IT industry were seen to advance the national dream—soldiers in the war for national growth and glory—in Ireland, under the political context of job creation, the same pioneers were seen as unstable and untrustworthy gamblers.^[37] Banks and other financial institutions not only were unwilling to invest in the industry but also refused to extend even the minimal necessary financial services to IT companies. Only those founders willing to risk everything they had by mortgaging their own property managed to secure enough capital to open their businesses. Accordingly, the financial backing of the IDA, even if limited, was absolutely critical.

Even in the relatively simple issue of finding office space, indigenous Irish IT companies have been facing severe difficulties and discrimination compared with the MNCs and traditional businesses. Regulation and tradition in regard to commercial space in Ireland have led to a system in which private property owners prefer to lease almost solely for multiyear terms. Furthermore, almost none of them were willing to lease their offices to NTBFs, which they perceived as too risky. Thus, if on the one side the IDA was busily developing purpose-built industrial estates for its MNC clients (hence, probably lowering the profitability of developing industrial commercial spaces for private property developers), on the other side the indigenous companies faced immense difficulties in securing even a small office space. Many of my interviewees as late as 2002 still argued that finding office space was one of their most difficult obstacles, even for well-funded companies.

The experiences from the creation of the IT industry in Galway after the closure of Digital’s facilities in the early 1990s are representative. The founders of both Storm and Toucan, two of the most successful companies to be created by former employees of Digital after it closed, quickly found out that even with state and regional agencies backing their efforts, and even at a time when their success was deemed crucial for the economic future of their communities, no one was willing to rent them office space. In the end the only way in which they managed to secure space was by the IDA’s allowing both companies to build their own premises in one of its industrial estates, a solution that forced both firms to spend a large percentage of their limited resources. Last but not least, while throughout this period MNCs were offered a wide array of tax incentives, the indigenous IT companies, like all Irish businesses, had to pay the full, high, corporate taxes rates.

[edit] State Policies and the Discovery of IT and Tradable Services: The Evolution of State Policies in the 1980s and Early 1990s

On the IDA’s side, the late 1970s were the years in which the agency started to appreciate the potential of the IT industry. If in the 1950s and 1960s the IDA was recruiting MNCs from a broad array, it became much more focused in the late 1970s, especially with the “growth without jobs” socioeconomic crisis of the 1980s. The IDA’s biggest coup in the 1970s was attracting Digital Equipment. Digital, then and throughout the 1980s one of the world’s top IT companies, opened its facilities in Galway in 1971 and soon became the biggest MNC in Ireland. Together with Ericsson, which has large facilities in Athlone, Digital has also been one of the first MNCs to move more advanced R&D operations to Ireland.

By the late 1970s the IDA started to focus on the tradable-services industries and on high-technology electronics. This process was soon accelerated when it became clear in the 1980s that manufacturing MNC subsidiaries preferred to increase their investment in capital equipment, not in expanding their headcount. Thus, while overall production grew, the number of jobs did not.

In 1975 the IDA established the International Services Program (ISP), which, through political lobbying, extended to the service sectors the incentives formerly given only to the manufacturing sectors. In the beginning software was not considered to be the main target, but by 1981 it became one of the ten designated sectors of internationally traded services to which tax incentives were fully extended. The IDA pushed for and succeeded in getting incentives, approval, and recognition for a rainbow of tradable-services industries, from English-language schools to software, finance, and film. These incentives and the actions of the IDA to both bring MNCs and create local enterprises in these areas developed the institutional framework in which the perception of service industry firms gained in reputation and legitimization.

In 1978 the IDA institutionalized its support to indigenous companies by creating the Enterprise Development Program (EDP). The main goals behind the creation of the EDP were to (a) increase employment, (b) lower Ireland’s reliance on FDI by growing indigenous companies, (c) create a set of suppliers to attract more MNCs, and (d) change the attitude against entrepreneurship and local enterprises in Ireland. The focus was on all sectors, both traditional and technological.

The same officials who were running the MNC-focused International Services Program were also given the responsibility over the EDP. Thus, unlike the case of the indigenous hardware industry, which was continuously treated as an extension of the already established MNC-oriented policies, the birth of indigenous software industry coincided with changes within the developmental agencies that institutionalized their interest and focus around the industry.^[38]

The two programs, ISP and EDP, were not focused on the IT sector in particular or on NTBF promotion in general, but on promoting any kind of entrepreneurial activity in Ireland.^[39] However, processes related both to state activities and to the activities of private entrepreneurs, coupled with changes in global demand for IT, propelled the software sector as the first in which indigenous companies achieved worldwide success. This success, in turn, refocused most of the attention of the newly created state agencies on the IT sector, specifically enhancing the software industry and aiding its continuous growth since the mid- 1990s. Indeed, an executive of the EDP program strongly insisted:

There was no aim to create science-based industry at the time; the program [EDP] evolved in a case-by-case process. Slowly it became clear that services are the main beneficiary, and that in reality we were mainly dealing with software. This is why we quickly added software to the list of sectors to get the full tax benefits and why later policy was constructed around software. But we had no idea of creating a technological industry in Ireland at the time. We just aimed to create more enterprises and jobs. (interview, March 14, 2001)

A similar response was given by all other former executives of the EDP who were interviewed. The EDP became crucial not only in the funding of companies but also in actively convincing potential entrepreneurs to start their business. At least as important, the EDP’s actions granted some legitimacy to technological entrepreneurship and risk taking. The latter might have been the most important accomplishment of the EDP, considering Ireland’s ultraconservative business and financial landscape at the time. In addition, on the FDI-oriented policy side, the new policies and the actions taken by the ISP culminated in major software companies’ opening subsidiaries in Ireland. A consensus evolved that the future of Irish industry did not rely solely on the manufacturing sector.

The first major software MNC that arrived in Ireland, transforming it into the world capital for software localization, was Lotus in 1985.^[40] In contrast to the indifferent reaction of the political establishment to the activities of local companies, the political and social elite attended the opening ceremony of Lotus’s Irish operation, and the company gained the goodwill of both the Irish developmental agencies and leading politicians. John Sterne argues that the main reason why Lotus was greeted with such esteem is specifically because it did not conduct any R&D activities in Ireland but concentrated on localization, packaging, and logistics. Therefore Lotus was seen as a real manufacturing company with sensible well-understood operations (Sterne 2004, pp. 101– 108).^[41] Following Lotus almost all of the top software MNCs, such as Oracle, Symantec, Corel, Novell, and Microsoft, opened localization, packaging, and logistics operations in Ireland during the 1990s. Microsoft soon became Ireland’s top exporter, and the sales attributed to its Irish subsidiary alone would make Ireland the world’s biggest software exporter.

In the 1980s the commitment of the state expressly to the software industry was further strengthened with the establishment of the National Software Center (NSC). The NSC was the first failed attempt to create within the IDA a subunit stuffed with industry experts to promote the growth of the software industry. In 1983 two competing proposals for a National Software Center were put forward. One, by a coalition of the Science and Technology Board, the Higher Education Authority, and representatives of the indigenous industry, envisioned a nonprofit center providing the industry with a wide range of services. The second proposal, of the IDA, was focused on the MNCs sector and envisioned the NSC as a self-financed operation giving services to the industry for a fee. Unlike the Science and Technology Board and the Higher Education Authority, the IDA did not bother to consult with the indigenous industry (O’Riain 1999, 2004, Sterne 2004).

In another example of the IDA’s power at the time, although the local industry supported the competing plan, the NSC was created on the basis of the IDA’s proposal.^[42] Brian Dugan was recruited from a position of vice president in Standard and Poor’s in New York to become the NSC’s first director. The NSC’s first board of directors also consisted solely of MNC representatives. However, the NSC quickly focused more on the indigenous industry. By 1987–1988 the NSC had become one of the main institutions in which the state and the indigenous software industry were collaborating. Nevertheless, the NSC’s need to become self-financed ultimately led to conflicts as well as perceived and real competition with industry. In addition, its focus on the indigenous sector cost it its IDA support. By June 1988 the NSC had closed.

Before it closed the NSC managed to complete the first survey of the industry. The result of this survey shows how different from the Israeli, and how similar to the Taiwanese, the Irish software industry was in the 1980s. If in Israel the leading companies at the time were focusing on solutions to R&D and software programming processes, in Ireland most of the companies either imitated already established packages of foreign companies on new platforms or focused on industry-specific applications, with little in terms of R&D or innovation. For example, in 1984 there were twenty-four companies selling packages for insurance agents, eighteen with software for lawyers’ offices, and twelve selling applications for television rental stores (Sterne 2004, p. 17).

By 1988 the political landscape in Ireland had changed, and the further development of the indigenous software industry was very much on the agenda. As soon as the NSC was closed, two proposals for centers to replace it were put forward. The first, focusing on skill development, resulted in the ultimately ill-fated PAT university-industry linkage centers established with EU funds. The second, developed together by the Ministry of Industry and Commerce and the indigenous industry, resulted in the creation of the National Software Directorate within the IDA in 1991.

In a symbolic move, very different from the IDA’s choice of an American to be the NSC’s first director in 1984, an Irishman, Barry Murphy, was chosen to become the NSD’s first director. Murphy, Insight Software’s CEO and a member of the board of the ISA’s predecessor, the Irish Computer Services Association (ICSA), has been one of the most vocal and prominent leaders of the indigenous industry.

The NSD became crucial on three fronts. It was the first organization to chart the size and scope of the software industry in Ireland on a comprehensive and regular basis, finding it, in the words of Murphy, “to have more companies than anyone thought at the time” (interview with Barry Murphy, November 6, 2000). Following that initiative, the NSD became the main center devoted to collecting, processing, and publishing data on the software industry. Second, the NSD became the main promoter of the movement toward the product-development business model and away from the consulting-business model. Last but not least, using EU financing, the NSD became one of the main initiators of the high tech–oriented venture capital industry in Ireland.

At least as important, however, was the effect of the NSD within the Irish developmental agencies. Unlike any other sector, software companies now had a direct line and a voice within the Irish developmental agencies, first the IDA and then Forbairt and EI. With the NSD pushing for more resources to be channeled toward the indigenous software sector, and with its people sitting on the grant-giving committees of the EI, software soon became the pilot sector around which Ireland’s science-and-technology industrial policies, focusing on the indigenous industry, were planned and implanted.

[edit] The Restructuring of the Developmental Agencies

Even with the IDA’s new programs and focus in the 1980s, the “growth without jobs” crisis, coupled with the failure of fiscal expansion policies, left Ireland in a dire situation. Emigration reached new heights, with 1.1 percent of the population leaving Ireland in 1989 alone, many of those from the young and highly educated segment of the population. This turn of events, together with a growing public resentment of the IDA’s focus on MNCs, started subtle but important changes in Ireland’s industrial policy, which, in the end, refocused it around the high-technology sectors in the 1980s and 1990s.

The latest restructuring of Irish industrial policies around the hightechnology industries started during the latest Irish economic crisis. In the early 1980s, major policy and social upheavals started the realignment process of Ireland’s institutional system along the path of NTBFs-oriented industrial development. In the sociopolitical arena resentment toward what was seen as the excessive focus of IDA on MNCs with much smaller amounts of resources channeled into the indigenous industry, coupled with the severe crisis of Irish industry, culminated in the Telesis report.

The report, commissioned by the National Economic and Social Council (NESC), hence with substantial political support behind it, argued for an almost complete renovation of Irish industrial policy. The report concluded that Ireland’s economic growth should have a “double engine” of FDI and Irishowned companies. The report also argued for the building of national champions, and contended that the level of organizational and management capabilities of Irish firms was too low for them to succeed without a hands-on industrial policy. Thus the report called for a break with the traditional neoliberal interventionalist industrial policy. Nevertheless, the report itself did not give any conditions whatsoever as to how national champions and winners should be selected, or to how the state should employ direct intervention to strengthen the management and organizational capabilities of Irish-owned firms.

The Telesis report was published in February 1982, and a heated debate about industrial policy followed. However, in the 1984 White Paper on Industrial Policy, the Telesis report recommendations were mostly ignored (O’Sullivan 2000b). Nevertheless, the report’s long-term impact was larger than it seemed at the time. It both started and lent legitimacy to a long process of refocusing industrial policy around the indigenous industry. This process culminated in another committee report, the 1992 Culliton report.

The Culliton report’s most important recommendation was that the state should direct its assistance into fixing general financial market failures—that is, into helping companies, in all sectors of the economy, that were deemed too risky by the existing conservative financial institutions to be granted finance.^[43] Moreover, the Culliton report envisioned a restructuring of the development agencies’ organization.

As a result of the Culliton report, two waves of bureaucratic reshuffling occurred. In the first, under the Industrial Development Act of 1993, the IDA was restructured into two main agencies: Forbairt took charge of the indigenous industry development and the Irish Science and Technology Board (Eolas); the IDA, still influential but with reduced power, was renamed as the Industrial Development Agency Ireland and given the mandate over the MNCs and FDI-related activities. In addition, a strategic, coordination and advisory agency—Forfas—was created. In 1998 the reshuffling was completed as Forbairt merged with the Irish Trade Board and parts of FAS (the training agency) into one agency with capabilities and responsibilities for promoting Irish-owned industry in both Ireland and abroad. The merged agency was renamed Enterprise Ireland.

This period also strengthened Ireland’s particular neoliberal interventionist ideology. These bureaucratic reconstruction efforts progressed side by side with the establishment of a neocorporatist regime by the new right-center government. In 1987, in a stark contrast to both the United States and the United Kingdom, when a new center-right government gained power in Ireland from a center-left one, it opted not only not to crush the labor unions but also to enhance their power by structuring an Irish-style neocorporatist framework— the social partnership agreements. This framework led to a series of ever more comprehensive three-year agreements, which are an integral part of the Irish industrial and economic policy to this day (Hardiman 2000, 2002).

Thus Ireland, led by an ideologically center-right government, possessing an industrial base consisting mainly of American MNCs, and under strong pressure from the IMF and the World Bank, decided on and successfully implemented a neocorporatist framework. Moreover, the Fianna Fáil government accomplished this feat at exactly the same time that their ideological contemporaries and partners, Thatcher in the United Kingdom and Reagan in the United States, were busily dismantling union power. Thus the end of the 1980s and the beginning of the 1990s brought major policy changes to Ireland. These policy changes, taken together as part of a major restructuring of the Irish industrial policies, have been very different from those taken in Taiwan and Israel in the same period.

[edit] The Rise of the Software Sector and the Continued Stagnation of the Hardware Sector: Development and State-Industry Co-Evolution in the 1990s

By the latter half of the 1980s, the success of the Irish product-oriented software houses, together with the PC revolution, started to change the Irish IT industry’s landscape, and a growing number of software companies were founded around specific product ideas. Nevertheless, as the stories of the first few years of even the most successful Irish companies like Iona, Aldiscon (now Logica-Aldiscon), and Smartforce (formerly CBT and now part of Skillsoft) attested, the industry’s main problem remained the acute lack of capital.

An example of the difficulties of the indigenous companies in these years is the story of Glockenspiel. Glockenspiel is also important to the development of the Irish software industry for three reasons: first, it was the first globally successful Irish company that utilized the path common to the Israeli software industry—development and sales of software tools to developers. Moreover, Glockenspiel pioneered in Ireland a particular strategy for doing so: the quick implantation of the latest international standards of a particular software development technology, with the company becoming the first to the market offering a workable product based on these standards. This is the exact strategy executed later by Iona, Ireland software industry’s flagship company, and since then imitated by many of Iona’s spin-offs. Second, Glockenspiel became famed for operating a high-technology firm in one of the roughest neighborhoods in Dublin’s inner city. This, while irritating some, also gave the indigenous software industry in general, and Glockenspiel in particular, some political goodwill. The strategy later became an integral goal of the Irish state IT development schemes.^[44] Third, the story of the rise and fall of Glockenspiel portrays the ambivalent way in which the IDA was treating the indigenous industry. The IDA’s treatment of the company when it faced crisis became a focal point for the growing resentment of the agency by the software industry.

John Carolan, an Irish software programmer with extensive knowledge and networks with research institutions and developers outside Ireland, founded Glockenspiel in 1984. At that time, Carolan became involved with the new technologies of object-oriented programming and the new C++ language standard. Glockenspiel utilized his knowledge of the C++ standard and became the first software company worldwide to develop a C++ compiler for the PC platform. In its first years Glockenspiel financed its product development mainly from C++ consulting.^[45]

In 1988, the same year in which many major Irish IT companies succumbed to financial difficulties, Glockenspiel reached a point at which most of its revenues came from export sales of its products. Very quickly Glockenspiel taught the Irish software industry the lesson already well learned by the Israeli one: software development tools which are sold directly to engineers and developers need no localization and can be sold all over the world without costly modifications. Moreover, the customers for these products, being engineers and programmers, are more sophisticated than the average consumer; hence products that are too complex and crude in their user interface to be sold to the consumer market can still be sold to these highly skilled customers. Building on its C++ expertise, Glockenspiel started to offer a wider range of RAD tools for C++ development. The company’s deep technical knowledge enabled it to develop award-winning products, and both IBM and Microsoft promoted their sales.

However, in Ireland, this technology development–oriented business model was so different from any other Irish company ever to exist that it led both the IDA and the financial institutions to treat the company with suspicion and mistrust. In 1991 Glockenspiel achieved record sales and was the second top indigenous software exporter after CBT. In 1992, however, its American distributor went bankrupt. Almost immediately its bank in Ireland informed the company that it had to pay its overdraft facility in full, and the IDA declined to extend any aid. Shortly thereafter Computer Associated bought Glockenspiel out of receivership, and the IDA, in a move that instilled further distrust within the indigenous industry—and that directly contradicts the logic of solving market failures by risk sharing—demanded that every grant the company ever received should be paid back in full (O’Riain 1999, Sterne 2004).

By the early 1990s the software industry had notably grown, and a number of product-oriented companies had attained global success. For the first time, Irish software companies managed to successfully cross the Atlantic and thrive in the American market. Furthermore, the Irish software industry became much more technologically savvy. One of the most important factors in enabling this development had nothing to do with the Irish state but a lot to do with the Irish university sector and the EU.

Starting at the end of the 1980s, and with growing importance in the 1990s, the Irish colleges and universities, in particular Trinity, started to participate in EU research framework schemes. In a short period of time, Trinity’s computer science department became one of the largest recipients in Europe of ESPRIT grants for computer science. These grants not only enabled Trinity to expand its computer science research activities but allowed several research groups to form and then spin out as already established companies with working products and strong IP.^[46]

The most important software company to rise out of these particular institutional settings was Iona. Iona had its origins when Trinity received approval to offer the first undergraduate computer science degree course in 1979. Trinity needed to expand its faculty quickly and enrolled as lecturers many of its graduate students, among them both Chris Horn and Sean Baker. At the time there was little money for research in Ireland. However, in 1980, one of the senior lecturers in the department, Neville Harris, came back from a sabbatical at Stanford, bringing with him in his baggage four Sun Microsystems processors and a 3mb-per-second Ethernet card to establish the Distributed System Group (DSG). In 1983 the coming of the ESPRIT program suddenly changed the research landscape in the department. ESPRIT allowed for research groups to apply for funding on a multiple-year basis. In 1984 the DSG group, then consisting of four lecturers, managed to get its first multiyear project. This allowed the group to expand, and joined by Chris Horn, who came back from Brussels after being part of the EU team forming the ESPRIT program, the group quickly managed to get enough funding to make it the biggest and richest computer science research group in Ireland. By 1985 the DSG group grew to encompass six lecturers and about forty graduate students. At that time the group was already run almost as an independent financial entity, a unique and novel experience in the Irish higher education system.

In 1991 a group within the DSG led by Chris Horn, Sean Baker, and Annari O’Toole decided to open Iona, a campus company based around their knowledge of distributed systems.^[47] By that time, Horn and Baker were already active in a new international standard group, the Object Management Group (OMG). The founders saw a window of opportunity utilizing a strategy similar to Glockenspiel: being the first company to offer products that implant OMG’s new middleware standard.^[48] However, in 1991, no one in Ireland was willing to invest in a company with such extensive R&D-based business plans. From 1991 to 1993 the company was financed from consulting revenues, including the fees the founders earned for teaching C++ and system analysis, from two ESPRIT projects, and from some savings. This financial situation allowed Iona to start product development only in June 1992. By that time Iona had also managed to get an IDA employment grant of £150,000, for which the IDA took a 7 percent stake of the company.^[49]

By June 1993 Iona had become one of the first companies to offer a product implanting the new CORBA middleware standard with C++ compatibility. This raised the interest of Sun Microsystems. In December, Sun offered to invest $600,000 (USD) for 25 percent of the company. This offer from a leading MNC completely changed the reputation of Iona in Ireland, transforming it from an untrustworthy R&D-based IT operation into a legitimate and promising company. It was here that the development agencies’ misconception of the software industry again proved to be a hurdle. The inward-looking development agency, by then restructured as Forbairt, as well as Trinity, refused to let its stake in Iona be diluted by Sun’s investment. Only when the founders threatened to close down the company did the two institutions back down and the IDA agree to match Sun’s funding to keep its stake from being diluted.^[50]

Quickly the new relationship with Sun, much as its alliance with the Israeli Checkpoint or as Amdoc’s alliance with SBC had done, opened the doors of the U.S. industry for Iona. By 1995 both Boeing and Motorola were using Iona’s products, and in February 1997 Iona became the second Irish company, after CBT, to go through an IPO on NASDAQ. Since by that time Sun had decided to sell its entire 25 percent stake in Iona, the IPO became the fifth-largest software IPO on NASDAQ up to that point.^[51] The sums involved and the extensive publicity helped to transform the perception of the indigenous software industry both within and outside Ireland.^[52] Software had become the jewel in the crown of S&T industrial policies in Ireland. This was especially true as EI’s growing portfolio of Irish IT companies’ stocks soon made it the most profitable Irish venture capital organization. Nevertheless, as we shall see, the growing fixation of EI on profitable investments has by now become an obstacle to the industry’s continuous growth.

The DSG group and Iona also offered a development paradigm for the evolution of other research groups, many of them turned into commercial spinoffs in Trinity’s computer science department. Some examples of these have been Wilde Technologies, another spin-off of the DSG group focusing on software design; Prediction Dynamics, which develops financial modeling software and was founded by Padraig Cunningham and John Carney, based on Carney’s Ph.D.; and Havok, which is the commercial reincarnation of the Graphic Users Group. In addition, EU programs and the incentives they supplied to organize the research into semi-independent applicative technology research groups gave rise to some of the relatively few successful Irish hardware companies in the late 1990s, such as Haptica and MV Technologies. As Trinity, even under the constraints of minuscule research budgets, has spun off many other companies both before and after, including another two of the only six Irish software firms ever to go public on NASDAQ, Baltimore and Trintech, one wonders why in Ireland the universities’ research efforts are seen to have failed to positively affect the growth of the industry. This is especially remarkable in that other universities, including University College Dublin, Dublin City University, and University College Cork, were also breeding grounds for many software companies throughout the years, if not to the same degree as Trinity.

Iona itself was also critical for the development of the industry in another way. Iona has become the educator of experienced managers and entrepreneurs in a whole cohort of computer science students. These students joined the company before its IPO and passed through its rapid growth phase. Many of them left the company after 1999 to establish their own companies. There are now more than a dozen direct Iona spin-offs. In the case of one of these, Cape Clear, not only did Iona invest, but two of Iona’s founders, Colin Newman and Annari O’Toole, left Iona to join and manage the spin-off.^[53]

Accordingly, by 1994–1995, the time of the restructuring of the development agencies and the NSD initiative to develop a high tech–oriented VC industry, the landscape of the indigenous software industry was already transformed. Smartforce/CBT had become the first Irish software company to become public on NASDAQ, Iona had secured major partnerships with Sun, Motorola, and Boeing, and by 1995 Aldiscon was the market leader in the new mobile phone technology of short text messaging (SMS).^[54] In short, by the time the industrial policy restructuring process had begun in 1994, the indigenous software industry was already established as the leading and most successful export-oriented sector in Ireland.^[55]

Many product-oriented firms had to support their R&D efforts by offering services and consulting, a fact that not only slowed their R&D efforts but limited their ability to develop large-scale, complex, R&D-based products. This situation started to change only after the reconstruction of Irish industrial policy and the refocusing of a more significant portion of that policy on the indigenous industry. The formal creation of Forbairt, the forerunner of Enterprise Ireland, in 1994 increased state support for Irish-owned software companies.

In the hardware sector, however, there was no significant improvement. The two prominent hardware companies that became major exporters in the 1990s were Silicon and Software Systems (S3) and Parthus, both IC design houses. The circumstances involving their establishment were unique, making them the exception rather than the rule. S3 was founded in 1986, and it has been managed since as an independent company by Maurice Whelan, a professor at Trinity. However, S3 was established as part of efforts by Whelan, a former researcher for Philips Electronics, and by the IDA to persuade Philips to open a subsidiary in Ireland. These efforts culminated in Philips’s agreeing to open S3 as a joint venture with Whelan, with Philips owning 90 percent of its shares.

Parthus was established as Silicon Systems in 1993 when Brian Long, then the chief design engineer of Digital in Ireland, decided not to move with Digital to Scotland after Digital’s operation in Galway closed. Instead, Long, with the help of the IDA, which also matched him with Peter McManamon, Parthus’s financial cofounder, co-established Parthus. The IDA’s help in the launch of Parthus was part of the agency’s effort to reconstruct the IT industry in Galway after the closure of Digital, by far the biggest employer in the region. However, what allowed the company to start operations and grow was a combination of financial backing and aid from Michael Peirce of Mentec; seed financing and a stream of orders from STMicroelectronics, obtained through Long’s industry contacts; and Long’s and McManamon’s savings. Only after securing these financial resources did Parthus obtain a matching grant from the then newly established EI. Parthus was started as a service IC design house, mainly for STMicroelectronics, which supplied Parthus with more than 65 percent of revenues in its first three years. However, luckily for Parthus, the contract with STMicroelectronics stated that the IP developed was to be owned by Parthus.

In 1998, building on the already established reputation of the Irish software industry, Parthus changed its business model to an IP licensing, utilizing what has been one of the biggest investments by Goldman Sachs in an Irish IT company, $16 million. In 2000 Parthus became the first and only Irish hardware company to be publicly listed on NASDAQ when it double listed on the London Stock Exchange and NASDAQ. In 2002 Parthus was bought by an Israeli company, DSP group, and was merged with DSP’s own IP licensing division to be refloated as Parthus-Ceva; in 2003 the name was changed to Ceva.

Thus the stories of S3 and Philips, Parthus and STMicroelectronics, as well as the earlier relationship of Mentec and Digital, show that in the context of the Irish innovation system, with its dearth of both capital and a semiconductor industrial ecosystem, such as Taiwan’s, only with a rare combination of long-term support of an MNC and of the development agencies could a successful hardware NTBF grow. With the development agencies focused as they were on the indigenous software sector, treating the indigenous hardware sector only as a derivative of the MNC-focused policy, it is not surprising that stories of successful Irish hardware companies are few and far between.

[edit] The Creation and Limitations of the Local VC Industry

The venture financing situation improved significantly after 1995. NSD, by then a part of EI, led an initiative to distribute EU-backed finance in an attempt to spur the establishment of high technology–oriented venture capital funds. This effort, coupled with the proof for profits by the publicly listed companies and some high-profile M&As after 1995, culminated in a small but vibrant local VC industry. By 1999, for the first time in the history of the industry, local software entrepreneurs could reasonably expect to find enough investment capital to start up a product-oriented company.^[56]

The first government initiative to create a VC industry in Ireland was not at all aimed at the high-tech sector. In 1994 the government, realizing that financial institutions in Ireland were overly conservative and unwilling to grant working and growing capital to new enterprises in any industry, used a policy of veiled threats to persuade the big pension funds to make available up to £100 million to venture capital very broadly defined.^[57] Using this finance, three VC funds were established: Delta, Act, and ICC. The three invested most of the funds in later-stage development of businesses, mostly in traditional companies, and not in IT.

In 1995 the NSD initiated the first attempt to spur a high technology–oriented VC industry. The NSD VC initiative, using EU money from the Operational Program for Industrial Development, 1994–1999, distributed a43.9 million to establish sixteen funds under a scheme in which half of each fund’s finance was granted by the state and the other half was raised in the private market.^[58] The state position in all these funds was that of a regular general partner. This created the ironic situation whereby in Ireland—where, unlike in Taiwan and Israel, the state specifically chose policy vehicles with the aim of limiting its direct intervention in the market—representatives of EI, often the same people, sit on the boards of the VC funds, the boards of companies in which EI invest, and EI investment committees that choose companies in which to invest.

The Irish state, much like the Taiwanese, saw VC as a missing pool of finance and not as a separate industry with specific capabilities to be developed. Hence, while one of the declared aims of the policy, as for Israel’s, was to enlarge not only the pool of finance but also the pool of financiers, the state opted to cooperate with local investors, who lacked specialized high-technology VC knowledge and background, and not with foreign VC-skilled institutions.^[59] The first fund to be established was the ICC software fund I. ICC fund I started operation in 1996, and Maurice McHenry, an old NSD and EDP hand, relocated from the NSD to become the fund’s manager. Aside from one large and recent VC fund (Cross Atlantic) and a few semi-institutionalized funds of successful entrepreneurs (Mentor, Oyster, and Island are the prominent ones), all the VC funds operating in Ireland were created as part of the NSD initiative.

In 2002 Enterprise Ireland announced another VC initiative, this time more regionally and sectorally oriented and fully funded by the Irish government. Under this new initiative the Irish government distributes 95 million euros to ten funds, most of which are the new funds of the same management companies of the first initiative, with the hope of stimulating the commitment of a total of 500 million euros to these funds’ management.^[60] It is the ironic result of the inherent conflicts of Ireland’s neoliberal interventionalist ideology that the state’s attempts not to interfere with the market’s price-setting mechanisms ended with the only nation of our three cases to fully subscribe to free-market principles, Ireland, becoming the one case where the state is not only the biggest VC in itself, owning shares in all of Ireland’s promising IT companies, but also the biggest investor in the VC industry. Accordingly, the state also has a claim to a large stake of the VC industry’s investments and a significant role in shaping its decision making.

Figure 4.1. Capital Raised by Irish VCs, 2000–2004. Sources: PwC 2000, 2005, 2006). Note: There is no distinction between Private Equity and VC in Ireland, hence only an estimated average of 70 percent of the figures above has been allocated to technology-based investment.

The VC initiatives of the Irish government through Enterprise Ireland and the NSD bore some fruit. However, as can be seen in figure 4.1, by the end of 2006, especially in comparison with Israel, the Irish VC industry is smaller, less professional, and still intimately linked to established Irish financial institutions. Whereas in Israel many of the VCs are former technological entrepreneurs, in Ireland most of the VCs are former accountants and management consultants. Moreover, the Irish VC industry is less internationally connected than the Israeli one, with the sources of its financing and its connections concentrated in Ireland and Europe. Thus, especially in the software industry, where the United States is the main market, the VC industry in Ireland is still lacking.

On the other hand, it might be that some of these features, principally the close connection with local institutional investors and the fact that Ireland possesses very developed institutional pension funds, made it attractive to Irish firms to double list on both the NASDAQ and the Dublin (or another European) stock exchanges. This double-listing option is one that all the Irish IT companies listed on NASDAQ followed, and it probably diminishes the need to transfer more and more activities to the United States.^[61] The greatest concerns regarding the Irish VC industry today are (a) its ultraconservatism and lack of will to invest in seed-stage companies and (b) its continuous reliance on the state for both its investment decision making and financing.

[edit] From the Late 1990s to the Current State of Affairs: The Strengths and Weaknesses of Ireland

In the late 1990s the Irish software industry enjoyed its greatest success so far. Six software companies—CBT/Smartforce, Iona, Trintech, Baltimore, Datalex, and Riverdeep—listed on NASDAQ. Together with a few large-scale M&As, these high-profile success stories helped to change the perception of the industry both within and outside Ireland. In 1998–2001, the new generation of NTBFs also enjoyed the advent of the VC industry and could raise enough financing to utilize product R&D–focused business models. In addition, many established companies, such as SoftCo and Fineos, transformed their business model into a product-based one, often using VC financing to accomplish it.

Figure 4.2. Software Sales per Employee, Ireland and Israel. Sources: IASH 2003, NID 2006.

However, as can be seen from figure 4.2, the average sales per employee in the Irish indigenous sector, while vastly improving from a low of $45,000 in 1991, is still only 40 percent of the average of the Israeli IT industry. This seems to imply two points: (a) from an optimistic perspective, there is still a lot of room for improvement, but (b) a larger percentage of the Irish industry is still focused on the less profitable activities of service, consulting, and bespoke development.

Furthermore, as can be seen from table 4.1, the composition of the Irish indigenous software industry deteriorated after the late 1990s. By comparison, the Israeli industry was hit hard, but its companies managed to regain high levels of growth and, in tandem with the U.S. industry, has since 2004 successfully gone back to IPOs on NASDAQ. However, in 2007 only three Irish companies are still listed on NASDAQ, and a new generation of NASDAQ high flyers is not in sight.

Enterprise Ireland has also been intensifying its own grant-giving mechanism, as can be seen in table 4.2. Between 1994 and 2003 Enterprise Ireland directly gave companies an annual average of £16 million.

However, the grant-giving mechanisms of Enterprise Ireland highlight the different underlying objectives of the Israeli and Irish industrial policies, and exacerbate the main obstacle that still exists for many Irish entrepreneurs: the acute lack of very early– and early-stage financing for NTBFs. EI financial aid packages for new companies consist of both a hodgepodge of grants—some of which are the remnants of older grant schemes mixed with some new ones— and equity-based investment. In order for companies to get special tax benefits and qualify for R&D grants, EI needs to approve them as “fast-growing startups,” a classification devised by the NSD in the early 1990s. Fast-growing startups are companies that are seen as capable of reaching sales of 2 million euros within a few years. Only a limited number of companies can be approved, however, and only limited resources are given to the NSD and EI to conduct the necessary analyses. Hence a bottleneck develops specifically in this critical development stage, at which the firms are in their most dire need for state aid.

This situation is further complicated by the equity base of the EI grantgiving mechanisms. EI, in one of the most paradoxical twists of the Irish neoliberal interventionist ideology, claims on the one hand that the state can and should take large stakes in private companies. However, adhering to neoliberal principles on the other hand, EI wishes to avoid a situation in which a government agency determines the market capitalization evaluation of a private company. This creates a situation in which the firms seeking EI aid must find private-market investors that would resolve the evaluation process. Only then would EI join the investment round on the same valuation basis. Many investors, though, agree to invest only in companies that have already received EI’s seal of approval. This seeming Catch-22 situation is even more complex when we realize that EI is itself the single largest financier of the Irish VC industry. Hence, in many cases, EI actions make the Irish VC industry even more conservative in its investment decisions than it already is.

This process, in effect, makes it impossible for most early-stage IT start-ups to get sufficient financial aid from the state. As early-stage investment is riskiest and the uncertainty highest, EI investments do almost nothing to solve the most intense market failure inherent in new technology development. In addition, it can be argued that EI financial aid packages, as they are organized now, help more in lowering the risk for the VCs than they help NTBFs to secure larger amounts of capital than they could without this venue.^[62] These shortcomings are readily admitted by EI. Accordingly, EI’s then-CEO Dan Flinters mused: “I agree with some of what you say, especially with the fact that we are still lacking a good mechanism to finance seed and early-stage companies, on which we are working, but considering our goals and resources, I think we have the best-working scheme” (interview, February 5, 2002).

Indeed, considering that the goal of EI is not to generate the maximum amount of R&D but to maximize the number of successful indigenous businesses that supply the highest number of jobs for the lowest cost to the taxpayer, EI mechanisms are sufficiently well suited. Nevertheless, the continuous crisis of the industry since 2001, coupled with the fact that there have been almost no seed investments by Irish VC companies since then, raises the question of whether EI policies and grants answer the needs of the indigenous industry, or whether they have in fact become an obstacle to its growth. This sentiment was voiced not only by many companies but also by former EI executives. One can indeed defend the Irish record by arguing that since 2001 there has been a global crisis of the IT industry in general and VC financing in particular. Nevertheless, this does not explain why Ireland fared so much worse than either Israel or Taiwan, especially when Ireland has just culminated a major policy effort, including VC creation, whose main aim was specifically to set in motion more NTBF formation, not less.

A founder of a successful company in the 1980s reflects on EI’s changing policies: “When we opened our first company, EI, then the IDA, was crucial. They also proved very, very helpful throughout the years and early 1990s. Now they just [have] become too conservative. They are no longer that helpful, especially to new young firms” (interview, April 18, 2001).

A few founders of younger start-ups concurred:

In the past EI were very helpful to young companies. Now, after they suddenly made money, the only thing they do is to play “king of the lake” games. Meaning that if you are already very successful, EI will come and grant you both money and help. When you actually need them, though, they just ignore you.
The part of EI that deals with helping you to set offices and meetings abroad is superb; not so within Ireland. In the beginning we were desperately looking to get grants from them, and they turned us down. Now after our success and after getting some VC money, they [EI] are running after us trying to convince us to take their money. They think they missed our boat and that they could have made money on us. . . . I really wonder if EI is at all capable of making decisions to invest in truly innovative, hence risky, new IT companies.
For two months I was trying to get money from EI, and then I finally understand that this is a waste ofmoney and time to go to EI. To get good money [large grants] from EI you must already have a good VC deal in your packet.
The funny thing is that I found out that EI is much more aggressive than the VCs themselves asking for a big chunk out of companies. This, and they are so utterly slow and bureaucratic. I truly do not understand why they do it, because on the other hand they are the nicest and friendliest people I ever met. I think that the real problem is that they become omnipotent: you now must have EI approval in order to get anything in Ireland and they are talking with you under so many hats that sometimes I wonder if the person in front of me knew which role he is supposed to be playing now. (interviews with company founders, January 7, March 13, March 20, and April 20, 2001).

Indeed, the situation has become so severe that one former IDA executive who took part in planning the equity-based grant-giving mechanisms argued for a radical change of EU grant-giving mechanisms:

At the time it was a good idea, we had two problems we needed to deal with. First, there was political pressure after politicians became concerned when, on one hand, many software companies who got grants went bankrupt, and, on the other hand, a few were sold to MNCs, with the founders making a large profit. The second was the need of technology companies that do R&D to get a large quantity of capital up front, which we could not do with employment grants, especially as new software start-ups just do not create that many jobs. Thus equity-based grants seemed like an ideal solution for both and could be worked within existing regulations. I also must say that in the beginning, when we did not make any profit, it worked like a charm. The problems started when EI holdings in a few companies were suddenly worth millions. The more EI turned into the most successful VC organization in Ireland, the more profits became a yardstick, with the result that investment decisions are becoming more and more conservative and profit-oriented. EI is now so obsessed about making money and does not care enough about the overarching goal that the state should have—the development and growth of the industry. (interview, March 10, 2001)

However, the most telling example of the acute lack of true seed capital might be the political battle the ISA led to prevent the cancellation of the Business Expansion Scheme (BES) and the Seed Capital Scheme (SCS). BES and SCS are tax schemes under which a group of private investors pool capital of up to 250,000 euros to invest in exchange for stocks in a specific enterprise under special tax discount procedures. The two, in particular the older BES, have provided a crucial source of capital for all newly established enterprises, especially IT NTBFs. Almost all of the Irish IT companies have used the BES as their main, and sometime sole, source of seed financing.

The two schemes were supposed to phase out in the end of 2003. The Ministry of Finance argued that with Irish technology-oriented VC industry having the largest-ever amount of capital under management, along with EI help, the two schemes, which need a special approval from the EU competition commission, should be scrapped. However, with the Irish VC industry and EI in effect reaching a deadlock over seed investment, the importance of the two schemes to the industry rose instead of declining after 2000. After extensive lobbying by the ISA, the two schemes were prolonged until 2006 and their maximum award was raised to 1 million euros. It is symbolic of the current missteps and misunderstandings of the Irish state-industry co-evolution processes that in a newspaper interview the then-chairman of the ISA, Cathal Friel, who has also been a top manager in the Irish VC and corporate finance industry, described the schemes as the only source of funding available for start-ups in the software sector, apart from EI aid (O’Halloran 2003, Smyth 2004a, Taylor 2003).

Hence it seems that even with all the new initiatives and the greater amount of finance designated for investment in the IT sector, the actual situation of the Irish IT industry in 2004 is not much different from that of the Irish IT industry in the beginning of the 1990s. This, again, hurts the IT hardware industry more than it hurt the software industry. The main reason for this disadvantage is that the hardware product R&D phase is both longer and more expensive; therefore the risk associated with hardware R&D is higher. The reactions of two leading VCs and hardware entrepreneurs to questions about the lack of investment in the hardware sector are revealing:

It is obvious why there are almost no new hardware companies. The risks are a lot higher than in the software sector, so we prefer not to invest in them.
There are no hardware followers to us because there is no money; you need a lot more money in hardware than in software, and when there is almost no money, period, then software companies, which can also generate revenues through consultancy and bespoke development, will be the only ones that survive.
Hardware investment is just too risky for Irish VCs: the development phase is long with no revenues, the costs are an order of magnitude higher than software; this is just too risky. Even in software we dislike investing in companies with no source of revenues. You want to know what was the best VC investment of the year in my opinion? It was this lovely company, three ex-Iona founders, they already generated 3 million euros in consultancy fees, and now they talk about a product. They do not really know what the product is going to be, just that it is going to be in middleware in the mobile domain. You see? Three trusted veterans, a constant stream of revenue, and low risk that they will go belly-up. These are the kind of investments we like to take part in, not some haphazard idea for some hardware product development. (interviews June 10, 2003; February 5, 2004; February 4, 2002)

This crisis in new company formation coincides directly with the culmination of the most extensive initiative to expand the Irish technology VC industry. Moreover, it coincides with the most extensive efforts ever to create significant academic R&D activities, with the aim of fostering the continuous growth of the IT industry in Ireland. In sum, it seems as if the coevolutionary process of state-industry relations in Ireland is now out of step. With EI’s centrality in each and every point of the system and with its new zeal for profit-generating investment, one must wonder whether the future development of the indigenous software industry in Ireland is now in danger of suffering from a stage two failure—the inability of the state to relinquish its own power over the sectoral industrial system.^[63]

The crisis in the formation of new firms is not the only one the Irish indigenous software industry is facing. Internal reports by the ISA strategic vision group in late 2003 showed that only twenty-four of the total of seven hundred Irish software companies reached the very low barrier of 2 million euros in annual sales. These findings suggest that even established software firms find it very difficult to expand their activities under the current conditions. Moreover, fewer than 30 percent of surveyed companies in 2004 were even willing to look for VC funding; this is a sharp decrease and another indicator of the growing mistrust between industry, investors, and the state agencies (ISA 2003, 2004, McManus 2002).

The one sector of the Irish IT industry to which this crisis is irrelevant is the MNC sector. As can be seen in table 4.3, MNCs not only extended their R&D activities in Ireland by buying Irish IT firms for their technologies but also, in the past five years, moved to open their own in-house R&D operations. Both Intel and IBM announced the opening of new R&D facilities, and the MNCs’ participation in the new Science Foundation Ireland schemes is growing, compared with the almost insignificant participation of the local software industry.

Accordingly, it seems as if without some changes in policy the end result of the new S&T industrial policy initiatives, aimed at least partly to diminish Ireland’s overreliance on the MNC sector, might end up increasing it (ISA 2004, IT 2003a, 2003b, Lillington 2003a, 2003b, Smyth 2004b).

In this chapter I have analyzed the development of the IT industry and the co-evolution of state-industry relations in Ireland. In particular, I have inquired into the different ways in which the state acquired its technological and scientific skills, the industry’s evolution, and the interdependencies between state policies and the industrial development. Unlike Israel or Taiwan, the Irish state’s almost sole development focus for many years was job creation. The state has also been following a much less active strategy with regard to the indigenous IT industry. While vastly improving the educational, physical, and communication infrastructures, Ireland’s industrial policy focused mainly on bringing foreign companies to open manufacturing facilities within its borders. Some thought was given to the ultimate goal of embedding these firms within the Irish industrial system, but there have not been, until very recently, coherent attempts to do so. As a result, many of these companies eventually left Ireland.

In contrast to Israel and Taiwan there were limited resources devoted to R&D in Ireland. Even when the IDA established a program to promote local industry in the late 1980s, its focus was on job creation and not R&D. This especially hurt the hardware sector. Furthermore, during the 1980s the hardware industry had to compete fiercely for the limited pool of skilled engineers against MNCs who were recruiting for both their Irish and global facilities. With meager financial resources and unable to offer secure long-term employment, the Irish hardware industry was at a severe disadvantage. This failure of the state to engage with the specific needs of the Irish hardware industry and to embed itself within it—even when the first successes of the indigenous industry, Mentec and Lake Electronics, were both hardware companies—is a case of stage one failure.

This failure, and the particular way in which state and industry finally did embed more successfully around the software industry, had much to do with the way in which the Irish bureaucracy is structured, specifically in the way knowledge in gathered and employed. For many years the Irish bureaucracy development agencies followed the ideal-type Weberian structure, with internal recruiting, training, and career development procedures. However, it has an English flavor, including a strong distinction between the social science– humanities–law “general educated” employees, who are being groomed to management, and the technical-engineering “expert educated” personnel, whose careers do not lead them to high management positions (Rose 1981, 1986). Since the 1960s there has been only one secretary general of a ministry who joined the civil service technical personnel career track with an engineering background.^[64] Accordingly, Ireland opted to follow the third strategy of acquiring technology- and industry-specific skills—creating domain-specific subunits stuffed with industry insiders—when these skills were deemed necessary.

For these reasons the Irish state, focused as much as it was on FDI, was unable to allocate enough resources and attention to devise specific policy initiatives and embed itself within the industry to spur its growth. It was only a decade later, after a few software companies achieved some commercial success in foreign markets, with the software industry using all its political muscle, and with the IDA already changing its focus of attention, that Ireland developed a coherent IT industry–focused industrial policy. Even then, the industrial policy was geared toward the indigenous software and not hardware sector.

Only in the late 1990s did Ireland start to pay close attention to the assistance its local companies might need abroad, and even then the state limited its role to that of a promoter of companies in various foreign markets and did not attempt to structure specific relationships between local and foreign firms. With regard to foreign financial markets, the state, while happy that some IT firms publicly listed on foreign stock exchanges, has also encouraged them to list locally. In addition, while employing and channeling EU grants, the state focused on local financial institutions as the main loci for its various initiatives to create a local venture capital industry.

Toward the end of the 1990s, with the efforts of EI and the establishment of SFI and PRTLI, coupled with the successful state actions to create and grow a local, and locally financed and managed, technologically oriented VC industry, it seems as if the Irish IT industrial innovation system has been finally transformed. However, the particular way in which this system evolved and the devotion of the Irish developmental agencies to the two contradictory ideals of the neoliberal interventionalist ideology still prevent the indigenous industry from enjoying most, if not all, of these newfound riches. With their resources and knowledge and their new interest in moving more R&D activities to Ireland, the MNCs are much better poised to utilize the growing academic R&D infrastructure in Ireland. Already American MNCs are by far the largest patent issuers in Ireland, and the inability of the indigenous industry to successfully grow firms leaves it extremely vulnerable vis-à-vis the MNCs in two fronts. Hence, without extensive changes in the indigenous-focused science-and-technology industrial policy, the new move of Ireland into highlevel R&D, hailed as a way for Ireland to become less dependent on FDI, would end up making the Irish economy even more dominated by MNCs than before.

[edit] Notes

1. ↑ As late as 1989 Ireland lost 1.1 percent of its population to emigration. Indeed, even today the Irish population is less than 50 percent of the eight million before the great famine of 1845–1850.
2. ↑ At around $8 billion, Dell’s 2003 annual revenues alone equaled the combined value of the Irish tourism and agriculture sectors.
3. ↑ Transfer pricing refers to the internal trade accounting techniques that MNCs use in order to record most of their profits within units or subsidiaries situated in the most favorable corporate tax regimes.
4. ↑ One significant result of such missteps is that in 2001, with the Irish VC industry providing record levels of capital, there were only three seed investments, which together equaled $750,000 USD, in the whole of Ireland. Hence at the same time that state-led efforts increased the amounts of IT-designated funding to their highest levels ever, the formation rate of new IT firms dropped to almost zero.
5. ↑ The autonomous IDA, building on its ability to lure MNCs and its domain over what was widely perceived as a key issue—job creation—grew to wield immense power of policy formulation and implementation over all issues of industrial and economic development until the 1990s. Hence, between the 1950s and the early 1990s, the IDA was a true pilot agency with sway over the actions of many other agencies (Chibber 2002, Johnson 1982).
6. ↑ As will be discussed later, in the early 1990s the structure of the developmental agencies was changed, and after several reorganizations IDA, or IDA Ireland, is now the acronym of the Industrial Development Agency Ireland and no longer of the Industrial Development Authority. IDA Ireland is an autonomous government agency responsible solely for FDI industrial policy. Hence it is a very different and much less powerful agency from the original IDA in its heyday. Currently the structure of industrial developmental agencies consists of Enterprise Ireland (EI), which is responsible for the indigenous industry, and IDA Ireland, which is responsible for the FDI-based industry. Both agencies are nominally under the auspices of Forfas, which is responsible for strategic planning in regard to industrial development. The three are semiautonomous public agencies under the Ministerial Department of Enterprise, Trade, and Employment.
7. ↑ The National Software Directorate was moved into Forbairt and then Enterprise Ireland after the two agencies were created. It is now called the National Informatics Directorate. An ill-fated predecessor, called the National Software Center, operated on a much smaller scale within the IDA from 1984 to 1988.
8. ↑ Consequently, as will be described in detail, the indigenous hardware industry, never the focus of targeted policies, continues to face severe disadvantages vis-à-vis both the indigenous software companies and the hardware MNCs.
9. ↑ For more on the Irish economic history see Haughton 2000, Kennedy 1989, O’Gr’ada 1997.
10. ↑ Two center-right parties have dominated the Irish political system: Fianna Gael and Fianna Fáil. The main difference between the two has been that Fianna Fáil, the Republican Party, until very recently did not officially accept the political separation of the Irish island. Since 1932 Fianna Fáil has managed to secure its position as the largest party and has been in power for most of the period. This makes Ireland a one-partydominated political system for most of the period of its early industrialization, similarly to Israel and Taiwan.
11. ↑ By 1961 the republic population was only 2.8 million, compared with 3.1 million in 1921 and 8 million before the great famine of 1845–1850.
12. ↑ Emigration was so severe that the Irish government organized special committees on the subject, and some observers, most notably John O’Brien in the 1953 edited volume The Vanishing Irish, questioned the viability of Ireland as an independent state (O’Brien 1953). In the 1958 economic program which followed Whitaker’s report, the government described Ireland’s main problem: “Production has not been increasing fast enough to provide employment and acceptable living standards for the growing number of our people; large-scale emigration has been accompanied by a high level of unemployment. Emigration will not be checked nor will unemployment be permanently reduced until the rate of increase in national output is greatly accelerated” (SO 1958b, italics added).
13. ↑ The IDA, and Enterprise Ireland after it, have been using employment grant mechanisms as one of their main vehicles for direct financial aid. The agency and its client corporation agree on the number of jobs to be created, and the agency awards the corporation a specific amount of capital per new job. The almost obsessive attention with which job creation and destruction have been followed in Ireland is evident in reading Irish newspapers. The exact number of jobs created or discontinued still appears in the title of all articles dealing with the fortune of foreign (and many local) companies, even if the main thrust of the article deals with other issues.
14. ↑ For an assessment and historical description of Ireland’s industrial policy, see O’Sullivan 2000a, 2000b. For a critique of Ireland’s FDI-oriented industrial policy now and then, see O’Hearn 1998.
15. ↑ Interview with Ken Whitaker, September 3, 2003.
16. ↑ For a detailed account of the development of the Irish higher education system, see White 2001.
17. ↑ Higher education in Ireland was such a rare opportunity in the 1960s that virtually none of the civil servants in the Ministry of Education dealing with higher education policy had a university degree (White 2001, pp. 41–42).
18. ↑ In addition to opening the new facilities and institutions, the state has also been using its control of university slot allocation to steer students toward engineering and technology courses.
19. ↑ Exposure as interns to various workplaces is still often an integral part of the education in the RTCs/RITs and other, nonuniversity, parts of the Irish higher education system.
20. ↑ For a paper trying to measure the economic effect of the expanding education system, see Durkan et al. 1999.
21. ↑ See NSB 2000, figures 6-55–6-61.
22. ↑ Interview with Don Thornhill, chairman of the Higher Education Authority, February 4, 2002.
23. ↑ For accounts of the development of the Irish software industry, see Arora et al. 2001, O’Riain 1997, 1999, Sands 2005, Sterne 2004. For an account based on cluster theory, see O’Gorman et al. 1997. For an interesting collection of papers on the specific experiences of software industries in a few emerging countries, see Arora and Gambardella 2005.
24. ↑ For some years, Cara, the computing subsidiary of the national airline, Aer Lingus, was the biggest software company, data processing center, and exporter in Ireland. With its ticket-ordering system and remote terminals, Cara was the first to introduce distributed software systems into Ireland and later also successfully sold hotel management systems in Europe.
25. ↑ Some of the local managers of MNCs’ subsidiaries claim that the existence of these IT outsourcing-services companies in Ireland granted the Irish subsidiaries a comparative advantage versus other foreign subsidiaries when competing for more activities in the corporate headquarters (interviews with former managers of the Irish subsidiaries of telecommunication MNCs, July 17, 2001).
26. ↑ Michael Purser, an employee of System Dynamics who left to become one of the first computer science faculty members at Trinity, founded Chaco. Chaco was created partly to help Purser’s students find employment. The company was considered the most technologically oriented of the Irish software companies in the 1970s before it merged with Baltimore, another company that Purser cofounded, together with Jim Mountjoy. A few more service-based companies that later transformed into product companies appeared in that period. The most notable of these were GC McKeown and AMS.
27. ↑ McGovern’s two most important investments were made in conjunction with his relationship with Jim Mountjoy, the cofounder of Baltimore in 1984 and Euristix in 1990.
28. ↑ Because many of the commercial space problems relate directly to the Irish-British tradition of renting commercial property only on long-term leases (five to twenty-five years), even in the 2000–2003 period many interviewees complained about the unwillingness of landowners to rent their offices to newly established software companies.
29. ↑ ICL was a leading British computer company that merged into Fujitsu.
30. ↑ The Maapics package was introduced to Ireland together with IBM’s System/34. IBM launched System/34 in April 1977 as its own approach to minicomputers. System/34 was a low-cost distributive data processing system modeled on the IBM 5340 system.
31. ↑ In 1988 Hoskyns also bought the first incarnation of CBT—Computer Based Training—Systems. With the sale of Hoskyns to CapGemini in 1990 the remaining product-based activities of the former Insight were discontinued. Barry Murphy, then CEO of Insight, left to become the first director of the NSD. CBT was established by Patrick (Pat) McDonagh in 1983. Its first product was a training kit for money market dealers. The domain knowledge, as well as some of the financial backing, came from Dermot Desmond. Desmond was the first private financier to invest significant amounts of capital in the software industry. (His best-known software investment was the backing of Fran Rooney’s transformation of Baltimore into a data security company in the late 1990s.) In 1987 McDonagh sold the company to Hoskyns only to buy it back in September 1991 after CapGemini bought Hoskyns. Under the management of Bill McCabe the new CBT concentrated its effort on the American, and not the European, market. Within three years CBT, selling software user–training products, mainly for Lotus Notes, reached revenues of $18 million before becoming the first Irish software company to list on NASDAQ in 1995. CBT later changed its name to Smartforce before merging with Skillsoft in 2002, and is now trading under the Skillsoft name. Using the money he earned from CBT’s public listing, in 1995 Patrick McDonagh financed Riverdeep, another educational company selling multimedia school courses. Riverdeep also publicly listed on NASDAQ in 2000, making McDonagh, a schoolteacher by training, the most successful software entrepreneur in Ireland.
32. ↑ As one interviewee from the former IDA remarked, “In that period [1986–1989] we joked that in Ireland the best way to create a small Irish software company was to establish a big one and wait” (interview, March 10, 2001).
33. ↑ A PBX is a private telephone network used within a specific organization. Users of the PBX share a number of outside lines for making telephone calls outside the PBX.
34. ↑ Lynam also became seriously ill in 1988 and retired from the industry. In 1990 he came back and established, together with his wife and son Henry, Klas, a telecommunication hardware company.
35. ↑ Landis and Gyr ran into difficulties in 1991 and closed down its Irish subsidiary. Out of its Irish operations two new Irish-owned IT companies were created by MBOs. The first was Lake Communication, which is still managed by several members of the old management team of Lake Electronics who stayed on with Landis and Gyr. The second was Peregrine Software, a financial data-security company that flourished for a few years before being bought in 2000 by Trintech, one of the three Irish software companies that are still listed on NASDAQ in 2006.
36. ↑ Mentec and its founders, especially Mike Peirce, became crucial to the continued development of the Irish IT industry. Peirce invested in and was the chairman of Parthus, an Irish IC design house established after the demise of Digital’s facilities in Galway, which became the only Irish hardware company ever to be listed on NASDAQ. Peirce and Mentec not only encouraged a few spin-offs, such as Eurologic and AEP, but also became their key early investors. In addition, in 2001 Peirce, together with Peter McManamon, the former director given to Mentec by the IDA, opened Mentor Capital, a small VC fund.
37. ↑ John Sterne, a historian of the Irish software industry, remarked that even as late as 1995 a minister of the state, Pat Rabbitte, decided that it would be politically expedient to use a joint conference of the Irish and Massachusetts software associations to chide the Irish software industry for “paying too much for their employees” (Sterne 2004, pp. 86–87).
38. ↑ During the interviews, executives of the EDP all remarked that electronics was always considered manufacturing. As a result it was not firmly on the new agenda of indigenous enterprise building and new tradable-services industries.
39. ↑ Interviews with former executives of EDP and ISP, March 10 and July 16, 2001, and with the first CEO of EI, Dan Flinters, February 5, 2002.
40. ↑ Many smaller foreign software companies arrived in Ireland before Lotus. However, the move of Lotus, at a time when it was considered one of the premier software MNCs, had the effect of legitimizing Ireland as the “place to be” for American software MNCs. The effect was similar to those of DEC’s move to Galway in the 1970s and Intel’s in the 1990s.
41. ↑ Localization is the programming activity of taking a software working in a specific culture or language and translating it into different cultures or languages.
42. ↑ O’Riain, reporting on the view of the indigenous industry of that period, cites its position as reflected in its main publication of the time, Irish Computer: “If the interests of the indigenous computer services industry companies are safeguarded and promoted, and if there is a proper structure within which they can influence the new body in the interest of their members, then it [NSC] will have the association’s support [ISA, then called ICSA]. If they are not met, in the word of a member of the ICSA executive committee, it will be ‘total opposition’ ” (O’Riain 1999, p. 214, italics added).
43. ↑ Again the contrast with Israel is apparent. In Israel the state focused its energies solely around market failures associated with R&D activities, not on general financial market failures.
44. ↑ The latest of which is the project to create an R&D-based multimedia-digital hub in the Liberties (around the old Guinness brewery), which brought MIT’s Media-lab as its anchor tenant.
45. ↑ A complier is a program that decodes instructions written in a specific programming language, such as C, Pascal, or Basic (so-called higher-order languages) and produces an assembly- or machine-language program that can then be executed by the computer.
46. ↑ The continuous importance of ESPRIT and other EU framework programs to the research activities in Ireland is evident when one realizes that as late as 1998 more than one-third of the research budget of Trinity, at the time Ireland’s university with the biggest research budget at a20.3, came from these programs (Grimes and Collins 2003).
47. ↑ Colin Newman was Iona’s fourth founder.
48. ↑ The specific standard was ORB, which stands for object request broker. ORB is the component in the heart of the common object request broker architecture (CORBA) middleware standard. Middleware is the general name for software whose main task is to allow programs written and operated on different platforms to exchange information and work together.
49. ↑ A former executive in Iona argued, when interviewed, that at that period in order for an indigenous company to be approved for the lower tax regime (automatically given to MNCs) the IDA needed to give its consent. The only way to get the IDA approval was to agree to a grant scheme in which the IDA took 7 percent stake in the company.
50. ↑ Financial resources for R&D in Ireland were so constrained even after the IDA approval of its second grant to Iona in 1994 that Iona hired a former graduate student of Horn and Baker’s to a full-time position of applying for ESPRIT funding.
51. ↑ Sun had two reasons for selling its stake: (a) by that time it had decided to abandon the CORBA middleware standard and concentrate its efforts on its own self-developed Java technology, and (b) Sun Microsystems preferred as a corporation to avoid the legal risks inherent in holding large stakes and board directorship positions in other public companies.
52. ↑ Iona had one other key influence. Led by former lecturers at Trinity, the company was the first Irish company to hire entire cohorts of Irish computer science students to conduct R&D work in Ireland. In this Iona helped to transform the perception of Irish computer science students about what should, and can, be done in Ireland.
53. ↑ However, later the relationship between Cape Clear and Iona turned sour; Iona took Cape Clear to court in 2002 over hiring issues (Daly 2002).
54. ↑ In 1996 Aldiscon was also the first company to secure major investment from the EI-sponsored ICC software fund. In 1997, in the middle of an IPO process, Aldiscon was sold to Logica for £90 million in cash and is now the division that generates most of the revenues and profits for Logica-Aldiscon.
55. ↑ At the time, the software industry was the only successful indigenous sector in Irish history apart from agriculture and tourism.
56. ↑ Good statistics are lacking, but using multiple resources we can ascertain that starting in 1999 a major upgrade of the Irish VC funds occurred, with actual VC investment in IT of 62 million euros in 1999 and a162 million in 2000. Sources: M.O.P 2000, PWC 2000, and the Irish Venture Capital funds’ Web sites.
57. ↑ Unlike the United States, in Europe almost any capital involved in the growing or the restructuring of a business falls under the rubric of venture capital. In effect, the European use of the term VC is similar to the American definition of the private-equity firm.
58. ↑ In the end only fifteen funds actually started operations.
59. ↑ On the distinction between the two VC industry creation policy types, see Avnimelech and Teubal 2003a, 2006a, 2006b.
60. ↑ Interview with Denis Marnane, March 7, 2002. For more information on the distribution of the first EI VC initiative, see EI 2000.
61. ↑ In 2001 Israel had acknowledged that problem and changed some of its regulations to make it easier and more attractive to Israeli firms to double list. However, it remains to be seen how widespread will be the effects of those measures.
62. ↑ As EI finance is conditional on VC funds’ agreement to invest, and as Irish VC funds usually make their investment conditional on the NTBF securing EI financing, EI packages are in a sense complementary finance, which otherwise the VC fund would need to supplement.
63. ↑ One measure of the new passion for profits and EI’s self-congratulatory mode, which is in part created by the adherence to the neoliberal interventionalist ideology, is the farewell interview that Dan Flinter, EI’s first CEO, gave the Irish Times before his departure in September 2003. Flinter was quick to point out that Enterprise Ireland realized about 250 million euros in 1999–2003 while investing “a sinfully small amount!” Flinter also forcefully defended the status quo, arguing that EI activities are getting the balance exactly right (O’Keeffe 2003).
64. ↑ As with Taiwan, and unlike Israel, the top civil servants in each ministry, including the secretary general, or the general manager, are chosen from within the ranks of the civil service.

This site made possible through the generous support of the Alfred P. Sloan Foundation