* Director, Centre for Industrial Relations and Professor, Department of Economics, University of Toronto, and Visiting Scholar, North America Forum, Stanford University. Without implicating them for any of the conclusions, I am greatful for numerous dis cussions in this area with Mike Denny, Al Hollett, Doug May and Clark Reynolds.

What happens to the poorest regions of a country as that country undergoes trade liberalization and economic integration with other countries? Are the economic conditions in the poorest region likely to improve, both absolutely and relative to the rest of the country, or is there a risk that the region may become even more marginalized? Are the efficiency gains from trade liberalization likely to reduce regional inequality or to increase such polarization? Even if overall regional inequality is reduce d, is it possible that the very poorest may get "left behind" in the process of integration (analogous to a reduction in inequality but an increase in poverty)? If the poor region was already heavily dependent upon transfer payments from the re st of the country, is that situation sustainable or more likely to change because of greater economic integration? If the dependency on transfer payments is to change, how should that change best come about? Will global competition for business investme nt and the jobs associated with that investment put pressure on countries (and regions within countries) to compete for such investment, and if so will that lead to a downward harmonization of social polices with a particularly adverse effect on poorer re gions?

While an answer to these questions is clearly beyond the scope of any one study, this paper can at least shed light on these issues by an examination of one particular case study -- Canada's poorest province, Newfoundland. 1 The focus will not be on a conventional regional analysis of the impact of trade liberalization on employment in this particular region, although some of that will be discussed. Rather, the focus will be on the impact that increased economic integrati on will have on transfer and income maintenance programs (in particular, unemployment insurance) that are important to a region, as is the case with Newfoundland. Attention will also be paid to the feedback effects that such changes in transfer payments will have on the ability of the poorer regions to compete in the more integrated environment. Finally, analogies will be drawn between the poorer regions and less developed countries because there are insights to be had from the development literature, e specially with respect to issues of growth convergence and the pre-conditions for sustained development in an increasingly integrated world.

The focus on Newfoundland is justified for a variety of reasons. It is Canada's poorest province and hence highlights the implications of increased economic integration for regional poverty. It is dependent upon transfer programs, especially unemployme nt insurance (UI), and hence highlights issues associated with the design and implementation of transfers like UI. 2 It is a regional economy where "UI has been used as an integral part of a policy of regional development" (May and Hollett (1995, p. 119)) and hence it illustrates the issues associated with such a regional policy. It is heavily influenced by the fishing industry and hence highlights regional issues associated with resource dependence. It is a region that is currently struggling with its internal problems, and trying to design solutions that will reduce its dependency upon transfers in order to be more self-reliant in the new regional ordering. As indicated by May and Hollett (1995, p. 21, 22): "If it is in Atlantic Canada that the country's income security system can be seen most clearly as 'a social policy disaster on a grand scale', it is in Newfoundland -- and particularly in the Newfoundland fishery ...that the disaster has been most complete... If Atlantic Canada is viewed by some as the 'sickest' region in the country, then Newfoundland is the province in which the most extreme symptoms of the illness are to be found." Newfoundland is Canada's most Eastern Province (Figure 1). Although Labrador is on the mainland, bordering Quebec, Newfoundland itself is an island separated from the mainland. This adds to its isolation a nd increases its transportation costs.

As indicated in Table 1, Newfoundland has a population of 581,100. Its unemployment rate, at 20.2%, is almost twice the national average. Average weekly earnings at $525, are only slightly under the national average, suggesting that the high unemloymen t has not led to wage reductions. Out-migration exceeds in-migration so that net migration is negative. Relative to the rest of Canada, there is a disproportionately small portion of the workforce in manufacturing, and a large proportion in primary indu stries (especially fishing), and in public administration.

The disproportionately high number of workers in the public administration category reflects the fact that the state is a key player in the Newfoundland economy. As will be outlined in more detail subsequently, this reflects a complex interplay between governments at all levels - federal, provincial and local (the latter especially through community based development projects). This interplay, however, is not so much based on the three levels of government co-operating to establish an infrastructure th at will sustain a dynamic, self-sufficient economy, but rather, it is based more on trying to maximize transfers and shift costs to other levels of government. As outlined subsequently, an artificially sustained economy has resulted, which will likely ex perience huge adjustment costs if it is compelled to adjust to the global market place.

The paper begins with a brief background description of the unemployment insurance (UI) system in Newfoundland, as it existed in the early 1990's. Emphasis is placed on how it has affected regional development, especially with respect to work incentives , human capital formation, and private and public investment decisions. The potential impact of increased economic integration on such income maintenance programs is then analyzed, with particular emphasis on the issue of downward harmonization and inter regional competition for investment and jobs. Alternative income maintenance reforms that have been proposed will then be outlined. The paper concludes with a discussion of the lessons that can be learned from the Newfoundland experience.

UNEMPLOYMENT INSURANCE AS INCOME MAINTENANCE

As the name implies, unemployment insurance was originally conceived of as an "insurance" program to provide short-term income support for individuals who become temporarily unemployed. Over time, however, it has evolved into a more permanent income maintenance program. It is poorly designed as an income maintenance program, however, for various reasons: it is based on the individual and not the family as the unit of account; it is based on earnings and not measures of assets or wealth; and i t is based on being unemployed as the only measure of "need", which means it does nothing to assist the working poor.

The unemployment insurance scheme discussed here is the one that has been in place in Canada throughout most of the 1980s and into the early 90s with some modifications. 3 It provided eligible recipients with an income replacement rate of 60 percent of their earnings in their former job. The eligibility or qualification requirement was 10 to 14 weeks of previous insurable employment, depending upon the regional unemploym ent rate, with the less stringent 10 week requirement for regions of high unemployment (9% or more) and the more stringent 14 week requirement for regions of low unemployment (6% or less). The benefit duration period was also variable, depending upon the individual's previous work history and the unemployment rate of their region. Specifically, for qualified individuals, the basic benefit period was one week for every week of previous insurable earnings, up to a maximum of 25 weeks of benefits. In addi tion, individuals could receive extended benefits of 1 week for every two weeks they worked in excess of 26, to a maximum of 13 weeks of additional benefits. Furthermore, they could receive regionally extended benefits of four weeks of additional benefit s for every percentage point that the unemployment rate in their region was above 4.0 percent. The maximum regional extension was for an additional 32 weeks of benefits. The maximum duration of benefits from all three components (basic, individual's pre vious work period, and regionally extended) was 50 weeks (52 less a two-week waiting period when no benefits were paid).

"Repeaters" who received UI in the year prior to a new claim had longer qualification periods. If the regional unemployment rate was 12% or more, however, such repeaters were exempt from the longer qualification period. Self-employed persons were not covered because of the moral hazard problem; they might "declare themselves" to be unemployed so as to collect UI. Self-employed persons in the fish harvesting industry who are boat-owners, however, have been exempt from this exclusio n.

Clearly, there are numerous design features of Canada's unemployment insurance system that provided disproportionate benefits to high-unemployment regions like Newfoundland. Specifically: (1) The qualification period to be eligible was the lower bound o f 10 weeks rather than the upper bound of 14 weeks; (2) The benefit period was extended by an additional 32 weeks 4 ; (3) repeaters were not subject to the longer qualification period; and (4) self-employed fish harvesters were covered (i.e., they were exempt from the self-employment exclusion).

The potential impact of these regional design features is illustrated in Table 2, which contrasts the potential UI benefits for a person in a low-unemployment region, and a high-unemployment region like Newfoundland. A one-year time horizon is used, and it is assumed that the individual goes on UI as soon as eligible. If the persons did not go on UI, their annual income would be 52w -- that is 52 times their weekly wage, or w.

The first panel indicates the situation for a person who works 10 weeks in the year, thereby having labour market earnings of 10w in either region. In the low-unemployment region they would not be eligible for regular UI since they would not have worked their 14 week qualifying period. Nor would they be eligible for regionally extended UI benefits. As such they would receive no transfers from UI and their annual income would be their 10 weeks of labour market earnings, or 10w. This amounts to 19% [i. e., 10w/52w] of what their annual income would have been if they had been able to work all year.

In contrast, the person in the high unemployment region would receive their labour market earnings of 10w. Since their 10 weeks of work would qualify them for UI, they would also receive 10 weeks of regular UI (one week for every week of previous work). This amounts to 60% of their previous wage for 10 weeks [i.e., 0.6w(10)]. They are also eligible for an additional 32 weeks of regionally extended UI (4 weeks for every percentage point their region is above the standard of 4.0%, to a maximum of 32 wee ks). Their total UI benefit period is therefore 42 weeks --10 of regular UI and 32 of regionally extended UI. Their total UI benefits is 60% of their former weekly wage for 42 weeks [i.e., 0.6w(42)]. Their annual income consist of 10 weeks of labour ma rket earnings plus 42 weeks of UI benefits [i.e., 10w + 0.6w(42)], which is 68% of what their income would have been had they worked all year [i.e., 52w]. Clearly, this is much higher than the 19% of the annual income for the person in the region with a low-unemployment rate. This stark contrast occurs, of course, because the person in the latter region did not qualify for UI.

The bottom panel illustrates the more realistic scenario where they both qualify for UI, for example by working 14 weeks. The person in the low unemployment region gets 60% of their weekly wage for the 14 weeks they receive UI [i.e., 0.6w(14)]. This is their total UI benefits since they are not eligible for regionally extended UI. Their total annual income consists of 14 weeks of labour market earnings, plus 14 weeks of UI [i.e., 14w + 0.6w(14) = 22.4w]. This is 43% [i.e., 22.4w/52w] of what their an nual income would have been had they worked all year.

In contrast, in that one-year period, the person in the high-unemployment region receives 14 weeks of labour market earnings plus 38 weeks of UI for the rest of the year [i.e., 14w + 0.6w(38) = 36.8w]. 5 This amounts to 71% [i.e. 36.8w/52w] of what their annual income would have been had they worked all year.

Clearly, in whichever scenario is used, persons in the high-unemployment region can receive substantially more in UI benefits, and this is likely to replace a much higher portion (e.g., 70% compared to 40%) of what their income would have been had they w orked all year. This occurs not only because they are obviously more likely to be unemployed, but also because the design features of UI reduce the qualifying period and extend the benefit period. In addition, although not shown in this illustration, th ey are less likely to be penalized for being a repeater, and (in the case of fishery workers) they are eligible even though they may be self-employed.

The scenario described above creates incentives to adapt towards an annual work cycle known as "lotto 10-42" whereby individuals work for 10 weeks to qualify for 42 weeks of UI (10 of basic UI and 32 of regionally extended UI) in high-unemploym ent regions like Newfoundland. There are strong pressure on all parties to adapt to this norm. Workers who take jobs for longer than 10 weeks have been dubbed "scabs" for taking the job of someone who could otherwise use it to build eligibilit y for UI. Employers are under strong pressure to provide a series of jobs that last 10 weeks (and only 10 weeks) so as to rotate as many persons as possible through the jobs to build eligibility for UI. They are also under pressure to maintain high wage s in those jobs because the transfer payment is a percentage of the former weekly wage, although employers may also "auction off" the jobs because they provide the eligibility for UI. As indicated by May and Hollett (1986, p.193) "a small percentage of more desperate individuals 'buy' stamps (eligibility for UI). To do this, the individual pays an 'employer' to put him/her on the payroll for the minimum eligibility period."

Even the behaviour of the ultimate "paymaster" - the federal government - has been altered to conform to this norm of 10 week jobs. The lack of viable economic activity in the region has meant that there was a scarcity of even such 10 week job s, especially relative to the increased demand created for them by the fact that women and younger people were increasingly entering the labour market to get such jobs so as to build eligibility for UI. Political pressure led the federal government to &q uot;create" such 10 week jobs through the job creation component of their training and labour adjustment programs. These programs were dubbed "make-work" projects in the recipient regions. 6

These projects initially were of some value to the community, but over time diminishing returns set in, and it became harder to find projects that were valued by the community outside of the jobs they created. Given the importance of the fishing industr y in that province, many of the newly created jobs were in building and operating fishplants and wharfs in the fishing communities. Not surprisingly, the fishing industry expanded (fostered also by expectations in the growth of the cod stocks that turned out to be wrong). The exemption for the self-employed to be covered by UI if they were fishery workers, encouraged entry into the industry. Entry was easy, since education and regulatory requirements were minimal, as were capital requirements (essentia lly a boat and some gear) which were also often subsidized. It was an ideal occupation for accumulating the 10 weeks of work to qualify for UI, since that period of time coincided with regular seasonal fishing patterns. As well, fishing was supported by an infrastructure of wharfs and fishplants - themselves established and operated to create more short-term jobs to facilitate qualifying for UI.

The ease of qualifying for UI in the fishing industry was also facilitated by the fact that $100 worth of catch per week was deemed the equivalent of one week's work. Therefore the 10 week eligibility period could be satisfied by $1000 worth of catch ov er a 10 week period. With reasonable luck, $1000 worth of fish could be caught in less than a week. Fish harvesters could then take turns dividing such a catch amongst each other, with each selling it to the fishplants in $100 lots to qualify for a week of UI. In such circumstances, it would be possible to qualify for a year of UI on the basis of less than one week of work, and certainly much less than the 10 week qualifying period (May and Hollett, 1995, p. 65).

Not surprisingly, the expansion of fishing that occurred in these circumstances (in what otherwise would be considered a declining industry) contributed to the overextraction and depletion of the fish stocks as a renewable resource (Cashin, 1993). As su ch, a moratorium on cod fishing was established to facilitate renewal of the cod stocks. This led to the problem of what to do about all of the unemployed who were directly or indirectly dependent upon the fishing industry. The problem was compounded by the fact that many had left school early to enter the industry (thereby not acquiring a general education), and they obviously have industry-specific human capital in a declining industry. Furthermore, fishing in such communities is not just a job - it is a way of life, even if one that has been artificially sustained. Moreover, individuals in that industry are not exactly the type who relish being re-trained to "flip hamburgers."

As part of the solution, the federal government focused on the retraining of fishery workers affected by the depletion of the fish stocks. The training was to be geared in part to the federal government's offshore Hibernia oil project in Newfoundland. That project, in turn, was not likely to be economically viable at current oil prices, but it was an attempt to diversify the Newfoundland economy away from fishing. Unfortunately, many of the jobs that were created in the building and operation of the o il platforms were filled by workers "imported" from other countries such as Norway due to insufficient human capital infrastructure in Newfoundland. Furthermore, much of the equipment building was shifted out of the province, again because of a lack of facilities within the province to do the work.

The cornerstone of the federal government's new initiative was the five-year, $1.9-billion Atlantic Groundfish Adjustment Strategy (TAGS) to provide adjustment assistance for some 40,000 fishery workers (an average of $47,500 per worker). Unfortunately , that strategy also was beset by problems, again related to the lack of "real" jobs in the economy. A recent report from Newfoundland indicated that "The federal government spent $892,000 on a program that probably will result in 22 unemp loyed Newfoundland fishery workers [average of $40,500 per person] being trained for 62 weeks to become unemployed divers ... [the training] was as much as eight times as expensive as that offered in Ontario ... those fishery workers who are scheduled to graduate from the course in about two weeks will need a lucky break to get a job in a stagnant market for offshore divers ... Many fishery workers, however, complain that their training is not properly targeted towards employment opportunities and that th e main beneficiaries of the federal government programs are the companies running the courses" (Globe and Mail, May 22, 1995, p. A1, A2).

The previous scenario was portrayed in a somewhat apocolyptic fashion, highlighting the worst features of UI in this environment. Certainly, a more positive spin could be put on many elements of the program. It does encourage people to work at least fo r certain periods to build up eligibility for UI. Some design features, like the extra week of UI eligibility for every additional week of work beyond 26 weeks, does encourage additional work to get the additional eligibility. Institutionalized workshar ing, by encouraging persons to share a series of short-term jobs supplemented by UI, may be an equitable policy in an economy that is demand constrained in that there is insufficient demand to generate enough full-time jobs. 7 The income transfers may also reduce outmigration to other provinces with their own severe unemployment problems. The TAGS program, while expensive, seems designed to enable people permanently to leave fishing, although many regard it as more of a fina ncial bridge (in the hope that fishing returns). Most importantly, and this will be emphasized in a later section, there is general recognition within Newfoundland that the existing system is not workable in the long-run; it is not in their own interest to have an artificially sustained economy.

IMPACT OF UI ON BEHAVIORAL RESPONSES

The previous discussion highlighted that, because of UI, the parties clearly have an incentive to adapt their behaviour, especially in response to the special regional provisions. In this section, the theoretically expected behavioral responses are outl ined in more detail, and the empirical evidence on the actual impact is presented, emphasizing the regional dimensions. The behavioral responses include: work incentives, unemployment, wages, human capital formation (education, training, mobility), entre preneurship, household production, and proposals to change the transfer systems themselves. While little direct empirical evidence exists on the impact of UI on these elements of behaviour in Newfoundland per se, there is some indirect evidence, as well as evidence from elsewhere.

In theory, UI should reduce the incentive to work in labour market activities since the cost of being unemployed is reduced and the income support enables one to afford not to have to work. UI can also encourage otherwise marginal labour force participa nts to enter the labour force to build eligibility for subsequent UI. Empirical evidence from Canada generally supports these predictions, indicating that UI encourages individuals to move from being employmed to being unemployed, and to remain so for a longer period of time. 8 The evidence suggests that the Canadian unemployment rate in the 1970s was increased by about 1 to 2 percentage points (when the rate was around 8 percent) as a result of the liberalization of UI that occurred in 1971. Riddell (1979) found the effect to be almost four times as large in Newfoundland compared to Ontario (i.e., increasing the unemployment rate by approximately 8 and 2 percentage points respectively).

Evidence also confirms the expectation that UI fosters short-term, seasonal jobs that are prominent in many regions, since the unattractiveness of such jobs is now at least partially offset by UI income-support that can provide a stable annual income-pat tern. 9 Large "spikes" in the distribution of employment at 10 to 14 weeks (the minimum period required to qualify for UI) have also been documented, reflecting the importance of the regionally extended benefits, as some individuals enter the labour force just long enough to qualify for UI, and others reduce their weeks worked given that it is supported by UI. Card and Riddell (1993) suggest that this may have accounted for some of the divergence of the unemployment rate between Canada and the U.S. that arose in the 1980s.

That UI has been institutionalized into the regular work patterns of many individuals who regularly alternate between work and UI is evidenced by the fact that about 80 percent of UI claimants had previously received UI, with almost 50 percent having fi ve or more previous claims (Corak, 1992). The prominence of repeaters is especially high in provinces like Newfoundland where UI has been institutionalized into the work patterns and the job creation patterns of the community.

There is also evidence indicating that regional mobility has been reduced by UI, especially the regionally extended benefits, because there is less incentive to leave high unemployment regions and move to low unemployment regions. 10 This in turn has fostered the continuation of the substantial variation in regional unemployment rates that prevails across Canada. This is consistent with the general empirical evidence in Canada that migration tends to be from the poorer (i.e., low- income, high-unemployment) regions, to the wealthier (i.e., high-income, low-unemployment) regions, and that the process of migration fosters convergence in such factors as wages, income and unemployment rates. 11 Conversely, income maintenance programs that offset the income and unemployment disparities reduce such migration. In-migration into Newfoundland is also encouraged by higher UI. 12

Evidence also exists suggesting that transfers like UI have slowed the regional convergence of productivity and wages that is otherwise occurring across Canada. 13 Milne and Tucker (1992) for example, find that the convergence in labour market earnings was slowest in the decade of the 1970s, a decade that saw substantial convergence in post-transfer income in large part because of the liberalization of UI over th at period. In essence, social policies like UI that can promote equality of post-transfer income can slow down the market forces that tend to foster more equality in the labour market earnings component of such income.

The impact of UI on wages is theoretically indeterminate and difficult to disentangle from other forces that affect wages. UI may reduce market wages by reducing the compensating wage premium that has to be paid for the risk of unemployment, and by enab ling employers to pay lower wages in return for offering jobs that provide eligibility for UI. On the other hand UI may increase market wages because of the reduction in aggregate labour supply created by the adverse work incentive effects, and because h igher wages maximize transfers from UI. Direct evidence to resolve this question does not exist. Earlier studies, however, emphasized that wages in Newfoundland were similar to those in the rest of Canada, in spite of the high unemployment, and that thi s wage rigidity contributed to the sustained high unemployment. 14 It is unknown, however, the extent to which the wage rigidity was influenced by UI or other factors such as a high degree of unionization, federal government pay that tends to be uniform across regions, or a historical tendency to regard "comparabi lity" with the rest of Canada as a right guaranteed by Confederation.

There is less direct evidence on the impact of UI on other human capital decisions notably education and training. Concern has been expressed in Newfoundland, however, that the availability of UI has discouraged younger persons from staying in school by raising the opportunity cost of schooling and lowering its benefits. The availability of UI has raised the opportunity cost of education, since the student is forgoing the opportunity to collect UI while in school. It has also lowered the benefits from higher education to the extent that it fosters unemployment, which reduces the period of employment over which to amortize the benefits of higher education.

UI also discourages training for similar reasons, although recent reforms have allowed recipients to collect UI while in training programs. Informal on-the-job training is also reduced, in part because employment periods are reduced and they may be spor adic. It is difficult to imagine much training occurring in sporadic 10 week jobs.

Self-employment (outside of fishing) is discouraged, in large part because the self-employed are not eligible for UI. Entrepreneurship is deterred because the business has to "compete" with UI for labour, and in fact has to be geared as much t o the UI system as to regular business activities. The underground economy is encouraged as incentives are created to work "for cash" or "in-kind" while collecting UI. 15 Household production (including the building and maintaining of homes) is encouraged while collecting UI. 16 While this is viable economic activity, and it contributes to a high-level housing stock, its high level is artificially sustained by UI rather than being a non-distorted decision between market work and household work. 17

Perhaps of greatest concern to Newfoundlanders is that the system of UI dependency and "lotto 10-42" work patterns are being institutionalized and regarded as the norm for new generations of younger workers. There is a realization that such an artificially sustained economy is not building the infrastructure (work norms, human capital, sustainable jobs, and viable investment) that is necessary to compete in the global economy. Incentives are alive and well; they are simply geared to working t he margins of the UI system, rather than sustainable economic activity.

IMPACT OF INCREASED ECONOMIC INTERDEPENDENCE

Given the way in which the Newfoundland economy has evolved, especially with respect to its dependence on transfer payments like UI, what impact will increased economic interdependence (e.g., from the Canada - U.S. FTA, and from NAFTA) have on poorer reg ions like Newfoundland?

The direct impacts are unlikely to be large. 18 Newfoundland is not exactly an economy prepared to develop world product mandates in particular items for which it has a market niche for export to the large North American markets. Given its remoteness as an island off the East coast of Canada, it may re-orient itself more in a North-South rather than East-West direction (May and Hollett, 1986.p.14), especially because the East-West orientation is somewhat artificially sustained by confederation. Nevertheless, because of its isolated island nature, t ransportation costs remain high. As such, it is unlikely to expand substantially relative to major cities and population centers in the rest of Canada that formerly "had a lock" on much of the domestic market under more protectionist regimes. 19 While major direct effects of economic interdependence are unlikely, the province obviously will have a comparative advantage in certain areas related to fishing and its forward and backward linkages, and possibly off-shore oil production.

While the direct effects of economic integration are unlikely to be large, the indirect effects may be substantial. The indirect effects pertain to the viability of transfers like UI under increased economic integration. Certainly, there will be an inc reased demand for adjustment assistance programs to facilitate the labour adjustment from the declining import impacted sectors to the expanding export led sectors. Nevertheless, the demand will be for active adjustment assistance programs (e.g., retrain ing, labour information, and mobility assistance) that facilitate adjustment in the direction of market forces. More passive programs that focus on income maintenance (e.g., UI) will fall out of favour because by providing income support for the unemploy ed they tend to discourage the reallocation of labour from declining, high-unemployment sectors to expanding low-unemployment sectors. This is especially the case with respect to the regionally extended benefits that are prominent under Canada's UI. It is true that UI can provide the necessary income support to facilitate and finance a period of job search, but this is likely to be swamped by its negative incentive effects as previously discussed.

Pressures to retrench on UI and special regional support will come from other sources related to increased economic interdependence. Transfer programs and special regional support programs could be interpreted as "unfair subsidies" and hence s ubject to countervailing duties if they artificially lowered the cost of particular exports. UI does not appear to be targeted to specific sectors as such, although it does "subsidize" seasonal industries and sectors that do not stabilize their employment pattern. In effect, employers in such industries do not have to pay a compensating wage premium to compensate for the risk of unemployment. Overall, however, UI likely raises wages because of the reduction in labour supply created by the adv erse work incentive effects, and because higher wages maximize transfers from UI. The special treatment of fish harvesters is more targeted, but the extent to which it "subsidizes" fish exports is unclear. It creates the potential for a gaming strategy for employers to get "cheap labour" by offering 10 week jobs which bring entitlements to extended UI benefits, but the obvious way for workers to "bid" for those jobs by working for low wages (or even negative wages!) is made difficult by the fact that UI benefits are positively related to such wages. Besides, unfair subsidies are not at issue when the fish stocks are depleted.

The more likely indirect way in which increased economic interdependence will lead to a retrenchment of UI will be through the increased pressure for governments to compete for investment and the associated jobs by reducing such transfer costs. With redu ced tariff barriers, firms are now more able to relocate to low-wage countries (or to countries with other comparative advantages) and to export back into the higher wage countries.

This is especially the case with the new "flexible factories" and "footloose industries", inter-connected by sophisticated transportation and communications systems (compared to the old fixed worksites like steel and auto assembly). I nternational financial capital is also incredibly mobile and prepared to shift instantly to arbitrage opportunities, including those created by lower transfer costs.

In such circumstances, firms are more likely to locate in regions that do not have high transfer costs, including transfer costs that arise from UI. It is true that since UI is not experience- rated, firms do not pay higher UI payroll taxes if they are in regions like Newfoundland where their employees are more likely to be on UI. Nevertheless, firms may be reluctant to locate in regions where they have to compete with the UI system for labour or where the work norms are geared to UI and where education and training has been discouraged by UI. More importantly, firms in the rest of Canada may be reluctant to pay the payroll taxes and corporate income taxes that support UI in other regions, especially if their competitors in other countries do not pay s uch costs, and if they now have the viable option of relocating some of their production activities elsewhere. Firms do not have to exercise that option; they only have to threaten to exercise it.

Under such threats of plant relocations (threats that are now more credible under increased interdependence), governments will have to compete more aggressively for business investment and associated jobs. One way to do so will be to trim regulatory cost s and transfer payments like UI, especially if they also interfere with the reallocation of labour from declining to expanding sectors. In essence, the federal government in Ottawa will be under pressure to reduce UI, and especially the regional aspects that do not benefit employers elsewhere in Canada. The Newfoundland government will also be under pressure to reduce its dependency on UI so as to attract the investment that is necessary for sustained development.

RESPONSES

Although it is difficult to know the extent to which they were induced by greater economic interdependence, many of the responses predicted above have occurred. The federal government has retrenched on UI and has reallocated funds from passive income ma intenance aspects to more active training and mobility aspects. More extensive reforms are also being proposed, including the possibility of making it harder for "repeaters" to institutionalize UI into regular patterns of work and UI.

Interjurisdictional competition for business investment and the associated jobs appears more prevalent. The most recent example is New Brunswick's attracting an amalgamation of UPS jobs from other parts of Canada, notably Toronto, Vancouver, Winnipeg, H amilton and Windsor (Financial Post, June 11, 1995). The defeat of the NDP party in Ontario by the Conservative party is also consistent with attempts to provide a more "business friendly" environment, especially as that party is repealing NDP initiatives in the area of labour law, affirmative action and workers' compensation.

The Maritime provinces themselves have been more actively pursuing their own regional alliance (Feehan, 1993), especially under the threat of being isolated from the rest of Canada if Quebec separates. Part of their agenda is to remove internal barriers to trade and to create a larger internal "common market" as a precondition for external competitiveness. Certainly, if they hope to develop the economies of scale and agglomeration externalities that facilitate external competitiveness, this s eems a pre-requisite.

External migration poses a difficult issue for provinces like Newfoundland, especially since there is still the memory of the forced relocation associated with the "resettlement" of whole outport communities that did not have adequate health an d the education facilities. The dilemma is enhanced by the fact that "the best" and most educated may leave. The latter also posses a problem for the strategy of improving the education system, since the most educated are most likely to leave. Of course, internal migration within Canada is not formally restricted, although occupational licensing and other barriers may prevail. The issue is more whether such out-migration should be more actively encouraged, for example, through mobility allowa nces. The political realities are such that it will be more easily encouraged indirectly by reducing the artificial barriers like regionally extended UI.

The most important and promising responses to the issue of UI under increased economic interdependence is coming from within Newfoundland itself. The House (1986) report was an internal Royal Commission study, and was very critical of the UI/Make Work s ystem and the perverse incentives it created. Of particular concern was the fact that younger persons were building up a dependency on UI, regarding it as an integral part of their lifestyle rather than as short-term support. The Powell (1988) Task Forc e report also expressed concern over the growing reliance on UI, while recognizing that the worksharing it encouraged was a natural response to the lack of jobs in the economy.

In 1992, the Cabinet of Newfoundland and Labrador proposed an Income Supplementation Scheme (ISP) that basically would involve replacing the existing UI scheme with a combined guaranteed annual income and earnings supplement program. The ISP would have three basic components: (1) a basic income guarantee of $3000 per adult, $1200 per youth, and $1500 per child; (2) an earnings subsidy of 20% for earnings between $500 and $10,500; and (3) a "tax-back" of 40%, applied to ISP receipts for recipie nts whose family income is above $15,000 (subsequent proposals in 1994 reduced the taxback rate to 30% but applied it to a lower threshold level of family income of $12,500).

The basic guarantee was designed to provide a minimal level of income for all persons in a family (unlike UI, which provided support only for those with some work history, and irrespective of family size). The earnings supplement was to augment the inco me of the working poor, and to encourage work incentives for persons of low earnings (unlike UI, which did nothing to assist the working poor). The tax-back was designed to "claw-back" some of the transfer to those with higher family income (ag ain, unlike UI, which could go to families of high income).

In essence, the proposal argues for taking the same amount of transfer payments that were coming in, especially from UI, and reallocating them so as to prevent the perverse incentives of UI. The main reforms to UI are the equivalent of removing the reg ionally-based benefits: (1) increasing the eligibility period for UI from 10 weeks to 20 weeks; (2) reducing the normal maximum benefit period from 39 weeks (42 weeks less a waiting period) to 17 weeks; and (3) eliminating eligibility for self-employed fi sh harvesters. In the short run, the federal government would be no worse off because the cost would be the same, and Newfoundland would be no worse off since the revenue transfers would be the same. In the long-run, both parties could be better off bec ause mitigating the perverse incentives of UI would enable the Newfoundland economy to be more self-sustaining and therefore less reliant on UI. The reforms would also be "progressive" in that they involve a redistribution from higher-income fa milies that previously received substantial amounts of UI (especially because of multiple claims of long duration), and towards somewhat lower-income families that were formerly by-passed by UI but that were in economic need.

What is important about the proposed reforms is that they represent an internally driven (rather than externally imposed) set of proposals to move away from the perverse incentives of UI and towards a more viable income maintenance scheme that could faci litate the transition to a more self-sustaining economy. 20 Clearly such proposals merit more consideration as possible solutions to the problem of regionally based transfers under increased economic integration.

LESSONS TO BE LEARNED

This case study of the Newfoundland economy under increased economic interdependence suggests the following lessons to be learned about the regional integration of poor regions dependent upon transfer programs like UI:

· Temporary adjustment assistance programs have ways of becoming permanent income maintenance programs, although they are poorly designed for such purposes.

· Passive income maintenance programs (like unemployment insurance) often work against the reallocation of labour from declining to expanding sectors, and inhibit the constant marginal adjustment that is likely less disruptive than occasional inframargina l adjustments.

· Incredible creative and entrepreneurial activity can become channelled into "working the margins" of maximizing transfers as opposed to direct productive activity.

· There is a tendency for solutions to address the symptoms rather than underlying causes. If people are unemployed give them UI; if the regional unemployment rate is higher, extend the UI benefits; if jobs are not there to enable people to be eligible f or UI, then create jobs; if jobs in the industry disappear, then retrain those who have built up dependency on the industry.

· There is a tendency not to anticipate the behavioral responses of the parties, and then simply to add a new program to deal with the negative side effects of the previous program. This often involves throwing "good money after bad" and a path dependence that creates permanent dependence rather than self-sufficiency.

· Regional economies can become dependent upon external sources of support and financial aid, with the behavioral responses of the actors and institutional features of the local economy geared to maximizing that support rather than being geared to market activity.

· Social and community norms can be important factors in sustaining such artificial regional economies.

· Black markets and the underground economy will flourish in such regions.

· Intergenerational transfers of community norms can be important in sustaining them, but there are also strong internal (and external) pressures to break them, given the cycles of intergenerational dependency that can be fostered.

· As in the general "convergence" literature, there is no guarantee that there will be upward convergence of such regions to the national norm; certain preconditions are required including an outward orientation.

· Increased economic integration and trade liberalization will inhibit regionally based transfer programs like UI, but this can be the impetus for internally driven reforms leading to more viable and self-sustaining regional economies.

REFERENCES AND WORKS CITED

Boadway, R. and A. Green. The Economic Implications of Migration to Newfoundland. Ottawa: Economic Council of Canada, 1981.

Card, D. and W. C. Riddell. A comparative analysis of unemployment in Canada and the United States. In Small Differences that Matter: Labor Markets and Income Maintenance in Canada and the United States. Edited by D. Card and R. Freeman. Chicago: Universi ty of Chicago Press, 1993, pp. 149-189.

Cashin, (Chair). Charting a New Course: Towards the Fishery of the Future. Report of the Task Force on Incomes and Adjustment in the Atlantic Fishery. Ottawa: Ministry of Supply and Services, 1993.

Copithorne, L. Newfoundland Revisited. Ottawa: Economic Council of Canada, 1986.

Corak, M. Repeat users of the unemployment insurance program. Canadian Economic Observer. (January, 1992) 1-25.

Courchene, T. Interprovincial migration and economic adjustment. Candian Journal of Economics. 3 (November 1970) 550-576.

Courchene, T. Migration, Income and Employment. Montreal: Howe Research Institute, 1974.

Courchene, T. Avenues of adjustment: the transfer system and regional disparities. In Canadian Confederation at the Crossroads. Edited by M. Walker. Vancouver: Fraser Institution, 1978.

Economic Recovery Commission. Proposal for a New Income Supplementation Program and Other Reforms to the Income Security System. St. John's: Economic Recovery Commission, Government of Newfoundland and Labrador, 1993.

Feehan, J. Atlantic provinces economic union. Canadian Public Policy. 19 (No. 2, 1993) 134-144.

Gauthier, D. Some Aspects of Internal Migration: Newfoundland's Case. Ottawa: Economic Council of Canada, 1981.

Grant, J. and J. Vanderkamp. The Economic Causes and Effects of Migration. Ottawa: Economic Council of Canada, 1976.

Gunderson, M. Regional dimensions of the impact of free trade on labour. Canadian Journal of Regional Science. 13 (Summer/Autumn, 1990) 247-258.

Gunderson, M. Regional productivity, wage and income convergence in Canada. Stanford University, North America Forum, mimeo, 1995.

Gunderson, M. and C. Riddell. Labour Market Economics: Theory, Evidence and Policy in Canada. 3rd Edition. Toronto: McGraw-Hill, 1993.

Hill, R. The Meaning of Work and the Nature of Unemployment in the Newfoundland Context. St. John's: Community Services Council, 1983.

House, D. (Chair). Building on Our Strengths. St. John's: Report of the Royal Commission on Employment and Unemployment, 1986.

Kaliski. S. Unemployment and unemployment insurance. Canadian Journal of Economics. 9 (November, 1976) 705-712.

Laber, G. and R. Chase. Interprovincial migration in Canada as a human capital decision. Journal of Political Economy. 79 (July/August, 1971) 795-804.

Maki, D. Unemployment benefits and the duration of claims in Canada. Applied Economics. 9 (1977) 227-236.

Mansell, R. and L. Copithorne. Canadian regional economic disparities. In Disparities and Interregional Adjustment. Edited by K. Norrie. Toronto: University of Toronto Press, 1986, pp. 1-51.

May, D. and A. Hollett. The Causes of Unemployment in Newfoundland. Background Report for the Royal Commission on Employment and Unemployment, Newfoundland and Labrador, 1986.

May, D. and A. Hollett. The Rock as a Hard Place: Atlantic Canada and the UI Trap. Toronto: CD Howe Institute, 1995.

Milne, W. and M. Tucker. Income convergence across Canadian provinces: does growth theory help explain the process? Atlantic Canada Economics Association . 21 (1992) 170-82.

Power, E. (Chair). Issues in Income Security Reform: Newfoundland and Labrador. Report of the Task Force on Income Security, 1988.

Richling, B. You'd never starve here: return migration to rural Newfoundland. Canadian Review of Sociology and Anthropology. 22 (No. 2, 1985) 236-249.

Riddell, W. C. Unanticipated Inflation and Unemployment in Canada, Ontario and Newfoundland. Ottawa: Economic Council of Canada, 1979.

Swan, N. and P. Kovacs. Empirical Tests on Newfoundland Data of a Theory of Regional Disparities. Ottawa: Economic Council of Canada, 1980.

Vanderkamp, J. The efficiency of the interregional adjustment process. In Disparities and Interregional Adjustment. Edited by K. Norrie. Toronto: University of Toronto Press, 1986, pp. 53-108.

Vankerkamp, J. Migration flows, their determinants and the effects of return migration. Journal of Political Economy. 79 (September/October, 1971) 1012-1031.

Vanderkamp, J. The role of population size in migration studies. Canadian Journal of Economics. 9 (August, 1976) 508-516.

Whalley, J. Economic, regional and labour market adjustment implications of Canada-U.S. free trade. In Assessing the Canada-U.S. Free Trade Agreement. Edited by Murray Stone and Frank Stone. Halifax: Institute for Research on Public Policy, 1987.

Wilson, L. Seasonal Unemployment in Newfoundland. Ottawa: Economic Council of Canada, 1982.

Winer, S. and D. Gauthier. Internal Migration and Fiscal Structure. Ottawa: Economic Council of Canada, 1981.

TABLE 1 -- BACKGROUND STATISTICS, NEWFOUNDLAND, 1993

Source: Statistics Canada. Canadian Economic Observer: Historical Statistical Supplement, 1993/94. Ottawa: Supply and Services, 1994, for population, unemployment and earnings figures; Statistics Canada, Report on the Demographic Situation in Canada. C atalogue No. 91-209E. Ottawa: Supply and Services, 1993, for migration figures; Statistics Canada. Labour Force Annual Averages: 1989-1994, Catalogue No. 71-529. Ottawa: Supply and Services, 1995, for industrial distribution.

TABLE 2 -- UNEMPLOYMENT BENEFITS IN A HIGH

VERSUS LOW UNEMPLOYMENT REGION

2 In the late 1980s in Newfoundland, half of the employed workforce (70% in rural regions) had received UI at some point during the year. The dependency starts early, with half of all individuals age 19 being on UI at some time during the year (Economi c Recovery Commission, 1993, p. 4-6).

3 In 1971, major liberalization of UI occurred in various dimensions: expanded coverage (to approximately 90% of the labour force); an increase in the benefit or income replacement rate (from 43% to 67% of previous earnings); a reduction in the minimum period of previous employment required to qualify for UI; an increase in the benefit period; and the introduction of extended benefits for regions of high unemployment. In 1978, some of this liberalization was reversed (e.g., the benefit rate was reduced from 67% to 60%), although for regions of high unemployment the benefits were expanded and a reduced qualification period was introduced. More stringent requirements have since been introduced and are continually being introduced. In 1994, the benefit rate was reduced to 55% of previous earnings and the qualification period increased to 12 to 20 weeks.

4 The maximum extended benefits are received as long as the unemployment rate is 12% or more. An unemployment rate of 12.0 is 8.0 percentage points above the 4.0 standard, and therefore gives rise to the maximum 32 weeks (i.e., 8 x 4) of additional ben efits since an additional four weeks of benefits are given for every 1.0 percentage point the regional unemployment rate exceeds the 4.0 standard. (Actually, an additional 2 weeks are given for every 0.5 percentage point above the standard; the 4 weeks pe r 1.0 numbers are used here for expositional purposes).

5 They actually receive a total of 46 weeks of UI, 14 from regular UI and the maximum of 32 from regionally extended UI. However, since they are working 14 weeks, they will only receive 38 weeks of UI (i.e., 52 - 14) in the one-year time period used he re.

6 Such projects are discussed in House, 1986, p. 406, and May and Hollett (1995, p. 48).

7 As stated by Power (1988, p. 60) "the practice of worksharing ... has actually become an effective mechanism for coping with a shortage of full year jobs in rural areas and one which evolved in response to very rapid growth of the labour force.&q uot; May and Hollett (1995, p. 81) state "workers accept comunity interests, which dictate that the largest possible number of people in the community who wish to earn a living at the fish plant should be able to do so. The dictates of survival hav e taught fishing families in small rural communities to cooperate."

8 That evidence is summarized in Gunderson and Riddell (1993, p.666-671).

9 Kaliski (1976), Wilson (1981).

10 Courchene (1978), Copithorne (1986), Maki (1977), Vanderkamp (1986), Winer and Gauthier (1982).

11 Courchene (1970, 1974), Grant and Vanderkamp (1976), Laber and Chase (1971), Mansell and Copithorne (1986), and Vanderkamp (1968, 1971, 1976, 1986).

12 Boadway and Green (1981), Richling (1985).

13 Theory and evidence on that convergence is discussed in Gunderson (1995).

14 Boadway and Green (1981), Copithorne (1986) and Swan and Kovaks (1980).

15 This is facilitated by the fact that prior to 1949 (when Newfoundland entered Confederation with Canada), and especially prior to the 1940s, Newfoundland was largely a subsistence, cashless, barter economy where "labour markets as we know them to -day were virtually non-existent and opportunities to work in paid employment were extremely limited" (Power, 1988, p.27).

16 The importance of household production in Newfoundland is amplified in Hill (1983).

17 In spite of its poverty, Newfoundland has the highest rate of home ownership in Canada (May and Hollett (1995, p. 53)), and its housing stock is of reasonably high standards, especially when compared to other poor rural regions of Canada. May and Hol lett (1986, p. 209) indicate that "the majority (76%) of households in urban Newfoundland reside in single detached dwellings. Of the people living in these dwellings 70% have no mortgage. These figures are higher than elsewhere in Canada, where a relatively higher proportion of households reside in rented apartments or have mortgages. For the rural regions of Newfoundland, mortgages are relatively unknown."

18 Relative to the rest of Canada, the Atlantic provinces including Newfoundland are expected to benefit disproportionately slightly more from the Canada-U.S. FTA because of the lower tariffs and because of increased exports of energy, natural resources and raw material. The employment effects are likely to be positive but small and slightly above average for Newfoundland relative to the rest of Canada (Gunderson 1990; Whalley 1987).

19 Arguments have been made, however, that the former protective tariffs favoured manufacturing in provinces like Ontario and cities like Toronto, at the expense of the rest of the country which had to purchase those higher priced good. More extensive f ishing rights would have been a property right that would have favoured the Atlantic provinces.

20 Other reform proposals in that same vein are outlined in May and Hollett (1995, p. 97-110).