Ethnicity-based research and politics: snapshots from the United States and New Zealand
On this Page:
Abstract
Some developments in poverty and ethnic inequality have followed similar trajectories in New Zealand and the United States, as have the political responses to such developments. This paper examines episodes in both countries in which political backlash against controversial policy analysis threatened to block worthy research and debate. Drawing on recent controversies over Simon Chapple’s paper on Maori socio-economic disparities, this article calls for a political climate that will support vigorous research into the full range of factors needed to explain advantage and disadvantage.
Introduction
Simon Chapple was a senior analyst in the New Zealand Department of Labour in 2000 when he wrote “Māori Socio-Economic Disparity”, a wide-ranging paper that marshals evidence intended to show, among other things, that disadvantage in New Zealand is more closely tied to age, marital status, education, skills, and geographic location than it is to ethnicity, broadly conceived. Although the paper did not issue clear recommendations for policy, Chapple’s conclusions pointed toward a gap-closing policy that would target pockets of disadvantage defined geographically and perhaps by sub-cultural features, rather than by targeting services to Māori as Māori.
As is well known, leaders of the ACT party used the Chapple paper to fuel a backlash against the Government’s “closing the gaps” policy in late 1999 and early 2000. The attacks by MPs Muriel Newman and Richard Prebble were reinforced by favourable commentary on Chapple’s analysis from less extreme quarters, such as Simon Upton’s web-based opinion column (Newman 2000, Prebble 2000, Upton 2000). Ultimately, the momentum of the Chapple-induced debates contributed to the Government’s decisions to remove the phrase “closing the gaps” from its vocabulary, to re-label the Government’s policy “social equity,” and to commit to fighting economic disadvantage rather than ethnic disparities (Young 2000a). This wave of backlash then produced a counter-backlash in which supporters of the Government’s original policy sought to discredit Chapple’s reasoning , and, ultimately, Chapple himself (Young 2000b).
This article draws on a somewhat similar episode in the history of American policy analysis in order to (1) illustrate the dangers of silencing debate about racial and ethnic disparities, and (2) discuss elements of an agenda for furthering our understanding of the Māori/non-Māori gap in labour market and other outcomes.
Backlash and the costs of silence - an American instance
Observers of the American scene have witnessed what can happen when a policy analyst foresees a social problem but political backlash silences debate and blocks action.
Daniel Patrick Moynihan, now retired from the United States Senate, was a senior official in President Lyndon Johnson’s Labor Department in 1965 when he wrote The Negro Family: A Case for National Action. That report warned the country that the achievements of the civil rights era could be endangered by the growing trend toward breakdown of the African-American family. He hypothesised from available evidence that the social disorganisation suffered by blacks who migrated to the cities after World War II had grown so deep that it could not respond to improvements in economic indicators, even to the achievement of full employment. According to his thesis, deprivation and disorganisation had formed their own vicious circle.
At the centre of that vicious circle sat the demographic evidence about family structure. Although Moynihan was not the first analyst to describe the breakdown of the black family in terms of divorce, out-of-wedlock births, sole-parent families headed by women, and welfare dependency (see, for example, Frazier 1939), he provided new statistical evidence and a model to explain it in his 1965 report. Most importantly, he caught the attention of President Johnson’s White House with his warning that black patterns of family breakdown and welfare receipt endangered the achievements of the civil rights movement. In Moynihan’s own words (1986:27):
My hypothesis was that a group in which a very large proportion of children are raised in the generalized disorder of welfare dependency will have a disproportionate number of persons not equal to their opportunities. In consequence, there would not be equal results.
President Johnson took up the message in his 1965 commencement address at Howard University, pledging to the graduates a renewed assault on “the next and more profound stage of the battle for civil rights. We seek not just freedom but opportunity – not just legal equity but human ability – not just equality as a right and a theory, but equality as a fact and as a result” (quoted in Moynihan 1986:33). Black civil rights leaders applauded the president’s speech. The White House began planning for a national conference to pursue the president’s pledge to close outcome gaps.
But by the time the White House conference convened in 1966 under the banner “To Fulfill These Rights”, the subject of the black family had become all but taboo. In the year between President Johnson’s speech and the conference, many influential civil rights leaders had labeled Moynihan’s report a subtle form of racism because of its unflattering portrayal of the black family (Wilson 1987). The renunciation of the Moynihan report by black leaders was so effective that it led to an unofficial ban on political discussions of black family dynamics and social structure. Social scientists, policy analysts and political figures of all races cut a wide path around the subject for 10 years after publication of what has come to be known as the Moynihan Report.
Research Constricts and Policy Shifts
How does a social science and public policy community as diverse and sprawling as the United States maintain a virtual ban on a research topic for more than a decade? Firstly, and most obviously, no one wanted to experience the wrath that had been visited upon Moynihan. The threat of political backlash made it easy to put the problem of race-based social structure in the too-hard basket. Secondly, the experiences of the African-American population bifurcated in the 1970s, and the rapid rise of a successful, highly visible black middle class – the great success story of the post-civil rights era – made the problems of the ghetto easier to forget.
As the within-group class differences for blacks began to overshadow between-group racial differences, sociologist William Julius Wilson (1978) reported “the declining significance of race” as an explanation of American social outcomes. His pronouncement was promptly taken out of context and used to support the false assertion that blacks as a group had reached parity with whites and that full equality of opportunity had been realised at last (Wilson 1987). For those who held this view, a natural corollary followed: that inner-city blacks who did not take advantage of equal opportunity and did not make the step up to the middle class suffered from strictly personal and behavioural failings rather than systemic or structural obstacles (Gilder 1981, Murray 1984, Sowell 1984).
The virtual moratorium on serious study of black social structure that followed the Moynihan controversy left a void into which conservative scholars poured their theories of behavioural poverty (Wilson 1987). Since then, political disputes about socio-economic disparities across races have often taken an overly simple, bimodal form: disadvantage is either the result of external economic forces teamed with racial discrimination, or the result of internal failings – bad personal behaviour reinforced by ghetto values and culture. Although debates across the two schools have shed light on public values and spurred important research into the dynamics of poverty and welfare dependency, neither theory tells the full story.
Toward an integrated explanation of racial disparities in the United States
As political scientist Hugh Heclo (1986) has observed, exploring the interconnections between opposing causal theories is often more productive than trying to determine which single factor dominates the others. Toward that end, Moynihan and Wilson have explained perhaps better than anyone else where the two, seemingly antagonistic, schools of thought on American poverty meet, and how economic forces, social structure, individual behaviour and culture interact to create and reinforce poverty in families, communities and demographic groups. Fortunately, although the 1965 backlash delayed these insights and their acceptance by the policy community, it did not prevent the story from finally being told.
Between 1960 and 1969 the population of black youths in American central cities increased almost 75% (compared to 14% for their white counterparts) and the population of black adults aged 20 to 24 increased at triple the rate of their white counterparts. As Wilson (1978:92) explained at the time, “This population explosion of young blacks has occurred at the very time when structural changes in the economy pose serious problems for unskilled Negroes in and out of the labor force.” The structural changes to which he refers included:
- the movement of blue-collar manufacturing jobs, which had provided low-skilled and semi-skilled workers of all races with stable employment and decent wages, from the central cities where most blacks lived to mostly white suburbs and exurbs;
- technological innovations leading to productivity increases in the remaining goods-producing industries and, inevitably, to lay-offs for low-skilled workers;
- the growth in central cities of service-sector jobs, which brought a bifurcated pattern of labour demand – good jobs at high wages for highly skilled workers (technicians, information managers, real estate salespeople, financial managers, etc.) and unstable jobs at low wages for others (hospital orderlies, hotel and office cleaners, warehouse workers, etc.);
- the combined power of labour unions and large corporate structures to further segment the labour market – protecting senior and more skilled employees while creating barriers to employment for new entrants to the labour market; and
- the rise of affirmative action hiring strategies, particularly in the corporate and government sectors, which enabled the rapid growth in the African-American middle class.
This complex array of demographic and economic factors produced predictable results in the form of very high rates of unemployment for low-skilled and semi-skilled urban blacks in the 1970s (particularly teens) and declining black labour force participation. At the same time, well-educated, higher-skilled blacks found more and more opportunities to use their qualifications in the professions and the corporate and public sectors, where demand for skilled labour has often exceeded supply and managers have embraced affirmative action (at least partly because of the necessity to find more skilled labour). As a result, a class structure arose within the African-American population that had not existed in earlier periods when blacks, due to racial discrimination and a greater abundance of low-skilled and semi-skilled jobs, exhibited a more uniform employment and income profile.
Thus, the rise of the black middle class occurred just as urban blacks with fewer skills found their opportunities eroded by the shift from a manufacturing to a services economy. Some low-skilled blacks accepted the new jobs and the accompanying decline in standard of living; but others, particularly young black males, pursued one or both of their two remaining options – the benefit check, or illegal income, much of it from the drug markets. Some cycled between casual employment and the benefit.
The changing class structure altered families and neighbourhoods (Wilson 1987). Successful blacks took advantage of housing desegregation and moved to the suburbs, leaving young ghetto dwellers with fewer and fewer role models to emulate (other than wealthy drug bosses). The declining numbers of inner-city black men with decent jobs, which Wilson (1987:83) refers to as the shrinking “male marriageable pool,” wreaked havoc on family formation in central cities. Many young, unmarried women chose to become sole parents because of a dearth of attractive marriage opportunities. These families experienced very high rates of poverty and welfare dependency because of the low earning potential of the female heads of household. Trapped in this cycle of economic stress and family disorganisation, many young people fell prey to domestic violence, drug abuse and crime, and neighbourhood and extended family support structures rapidly eroded. The result is what analysts in the 1980s called the urban underclass phenomenon.
Parallels in New Zealand
A set of demographic and economic trends strikingly similar to those described by William Julius Wilson in the United States of the 1970s converged 10 years later in New Zealand but, as in the United States, policy makers did not heed the warning signs.
As early as 1968, J.R. McCreary had observed that the very large Māori birth cohorts of the 1960s would come of age and begin looking for jobs in the mid-1980s. He warned that if the patterns of rural–urban location and skill distribution of the earlier period persisted, many of these young people would face serious obstacles to labour force entry. With regard to spatial patterns, the Department of Labour produced two maps comparing the geographical distribution of the urban-concentrated New Zealand labour force against the geographical distribution of the rural-concentrated Māori population in 1960. As McCreary (1968:197) put it, “one is the ‘negative’ of the other – where the jobs are, the Māori population are not, and vice versa.” With regard to skill distribution, McCreary (1968:192) anticipated the following problem:
As an expanding population grows up, the adult age groups can become swollen so rapidly that community resources become strained. If the Māori workers are not distributed through the occupational divisions in approximately the same proportions as the non-Māori, and at present they are not, a labour surplus may be created in those divisions now most congenial to Māori work skills.
Although many Māori have enjoyed increased occupational opportunities since that was written, Pool (1991) has described in detail how structural shifts in the economy have undermined these opportunities. According to his analysis, the percentage of Māori workers found in the primary sector dropped from around one-third in the early 1960s to around 15% in the early 1980s. Nonetheless, Māori continued to concentrate in secondary industries such as meat processing and manufacturing, as well as in certain primary industries such as forestry and construction – the very industries that suffered most acutely in the post-1984 period when import liberalisation, price decontrol, industry deregulation, and privatisation unleashed competitive pressures and forced job losses in previously protected and nationalised sectors.
As a result, the employment picture for Māori in the late 1980s featured not only the well-known high levels of official unemployment and low labour force participation rates associated with economic liberalisation, but also the seeds of future social disparities as large numbers of young Māori were denied entrance to the labour market altogether. According to Pool (1991), the younger the cohort, the worse they suffered in the pre-1996 period. New Zealanders born before 1955 experienced the lowest rates of unemployment in the late 1980s and 1990s, while their counterparts born between 1970 and 1975 experienced the highest rates.
Chapple (2000a, 2000b) found large intra-group differences in Māori socio-economic outcomes. This suggests that growing opportunities for skilled Māori in recent decades may have contributed to a bifurcation of the Māori labour market similar to that experienced earlier by African-Americans.
Clearly, the dominant cultural minorities in both the United States and New Zealand experienced similar demographic and economic stresses at different periods in their development; these stresses may go some distance toward explaining current disparities in socio-economic outcomes. The mechanisms of black family and neighbourhood breakdown in American cities, which Wilson describes, may or may not have obvious New Zealand counterparts. The unique history, culture and social structure of Māori whanau and communities have, of course, given rise to distinctive patterns of social success and failure. Further integration of existing research findings as well as new research are needed to explore fully the similarities and differences.
A focus for future research and debate
I believe that it is too early to say whether or not the controversy surrounding the Chapple paper will lead to the kind of moratorium on research and debate that followed the Moynihan report in the United States. There are, however, at least three good reasons to hope that it does not. Firstly, his analysis identifies subgroups of the population – namely, “sole Māori” with low education levels living in rural areas – whose dramatically poorer labour market outcomes deserve special attention in any “gaps” debate. Secondly, it is worthwhile to enter into a vigorous debate on the meaning of ethnicity for public policy and for the nation as a whole. Finally, there needs to be room for debating the rationale for government policy in general.
Labour Market Outcomes for Subgroups of Māori
Although Chapple’s results have been questioned on methodological grounds (Alexander and Williams 2001), his exhortation to look at the shape of the distributions that lay below New Zealand’s average gap figures deserves serious attention. Chapple’s discussion focuses on the need to understand the second moment of the distribution, by which he means the variance or spread among rich, middle-class and poor individuals within the Māori group. This contrasts with many researchers’ focus on the first moment of the distribution, or the mean difference between Māori as a group and non-Māori as a group. The results of his analysis further point to the need for better understanding of both variance among Māori and variance between Māori and non-Māori at the bottom of the socio-economic ladder. In particular, his regression analyses point to a possible pattern of entrenched disadvantage for certain subgroups of Māori that resembles the patterns displayed by the African-American underclass.
The clues emerge as follows. In most of Chapple’s (2000a) models of socio-economic outcomes, the independent variables representing ethnicity (sole Māori, mixed Māori, or non-Māori) lose their statistical significance when other explanatory variables are added – namely age, marital status, education and literacy. Chapple takes this as evidence that ethnicity does not make a discernible contribution to the gaps, independent of its association with demographic and productivity-related characteristics. However, as Alexander and Williams (2001) have observed, Chapple’s technique may be misleading because it does not take account of the strong correlations that are likely to occur between and among the explanatory variables themselves – age and ethnicity, for example, due to the different age pyramids of various ethnic groups, or education and ethnicity, due to lower educational attainment by Māori. In the presence of this suspected multicollinearity, one cannot separate out the independent contributions of the ethnicity variables from the contributions of other variables in the model, and the values for the regression coefficients cannot be trusted. Unfortunately, we do not know the degree of multicollinearity present.
Despite this potential problem, two results stand out. First, Māori/non-Māori gaps in employment status appear more robust than the gaps in earnings. With earnings disparity as the dependent variable, Chapple’s “sole Māori” ethnicity variable remains significant through only one (for females) and two (for males) out of five iterations of the model. (Each iteration adds another independent variable.) With employment rate disparity as the dependent variable instead, the “sole Māori” ethnicity variable remains significant through three (for males) and four (for females) out of four iterations. In other words, even controlling for differences in demographic and productivity-related features, and even in the presence of possibly high levels of multicollinearity, Chapple finds that “sole Māori” women experience discernibly lower rates of employment than non-Māori women. For men, the differences are almost discernible.
Second, as Chapple notes, both income gaps and employment rate gaps prove particularly robust among defined subgroups of the population; namely, subjects in the lowest educational category (some secondary education but not completed) and subjects in rural areas. For both of these subgroups, the “sole Māori” variable had statistically significant explanatory power even when the full set of potentially multicollinear controls was included (see Chapple 2000a, Tables 8 and 9).
Regardless of whether the actual values of the regression coefficients in these models are or are not reliable, the persistent power of the “sole Māori” factor suggests that the strongest interactions between ethnicity and labour market outcomes may occur (1) at the entry gate to the labour market, rather than within its halls, and (2) among a particular subgroup of Māori – those with sole Māori ancestry and less-than-complete secondary education living in the provinces.
If further research confirms these suggestions, it would help to explain other recent findings on labour market discrimination. Most notably, Alexander et al. (2001) estimated wage regressions using a two-stage procedure that models both employment status (employed or unemployed) and, among the employed, amount of actual hourly wages, while controlling for characteristics such as age and education that affect an individual’s value to potential employers (his or her productivity). When the first stage of this procedure is omitted, the sample includes only employed individuals with non-zero wages, and the “sole Māori” ethnicity variable achieves statistical significance in only one of the three years studied (1997 but not 1998 or 1999). In this version of the model, ethnicity appears not to affect actual hourly wage rates among the employed. However, when the two-stage procedure is used, the sample effectively includes both employed and unemployed individuals and the “sole Māori” variable achieves statistical significance in all three years. Among individuals of presumably equal productivity (as measured by the proxies listed above), including both employed and unemployed, Māori received actual hourly earnings between 8.7% and 13.1% lower than those of non-Māori in the period 1997–99.
Although the Alexander et al. model cannot tell us precisely what portion of this ethnic earnings disparity is borne by those with zero wages, the differences between the one-stage and two-stage models suggest that Māori labour market disadvantage may concentrate either exclusively, or more heavily, along the employment–unemployment dimension than along the dimension of pay-cheque variation for those already employed. Together with a similar suggestion from Chapple’s analysis, we have at least enough circumstantial evidence to prompt further research into the dynamics of labour market entry as a key explanator of ethnic disparities.
A Dialogue on Ethnicity
A second reason for hoping that debate on the Chapple paper remains vigorous is pent-up demand for dialogue on the broad subject of ethnicity and what it means for national identity and public policy. The current government’s Closing the Gaps strategy never used ethnicity as the sole targeting basis for social welfare services. Nonetheless, the idea of ethnicity-based distribution has not disappeared from the policy lexicon, because politics is largely the process by which we decide who gets what from government, and policy is the articulation of those decisions. The definition of “who” thus becomes central to the social policy debate in a way that transcends particular governments, strategies, or programmes
In his contribution to the “who” debate, Chapple (2000a, 2000b) rehearsed important problems with the official Census approach to ethnic identification, which make counting of Māori difficult, and discussed some of the psychological and sociological issues associated with ethnic self-identification and ethnic mixing. Together, Chapple’s comments on (1) problems distinguishing Māori from non-Māori, and (2) evidence that the gaps among groups of Māori exceed the gaps between Māori and non-Māori appear to pose serious threats to the concept of ethnicity-based distribution of public goods and services.
One generally justifies strict ethnicity-based distribution through one of two broad principles – ascription, the idea that membership in a particular group is sufficient to entitle the member to something (the idea is related to desert and arises in the context of Treaty claims, among others); or group need, the idea that group membership serves as a good proxy for some relevant dimension of deprivation or neediness. Chapple’s analysis appears to defeat both principles – and therefore ethnicity-based distribution – because ascription presumably cannot occur if stable criteria for membership (Census counts of Māori) cannot be delineated, and group need cannot operate as an effective targeting mechanism if the differences within the target group (Māori) swamp the differences between the target group and the culturally dominant group.
However, these arguments against ethnicity-based distribution are incomplete for two reasons. Firstly, they neglect the fact that difficulties in determining group membership are not peculiar to ethnicity. Deciding who will and will not count as “disabled” requires judgement calls and arbitrary lines of demarcation not unlike those associated with ethnicity classification. Likewise, definitions of who are the needy “elderly” may cause fewer problems at a point in time but will surely change as life expectancy increases and quality of life at older ages improves. Secondly, the arguments above construe the concept of distribution too narrowly. Programmes may be designed to ensure appropriate service choices for Māori and other disadvantaged groups, rather than to secure proportional group shares of defined goods. In this policy setting, concepts of access, choice and opportunity become more relevant than the older notions of ascription and group need. Arguments against ascription and group need, in turn, lose their sting.
A climate open to wide-ranging study and debate is the key to addressing these conceptual and methodological issues and to ensuring that social policies are developed deliberatively and democratically.
Clear Rationale for Social Policy
A third and final reason for preferring debate to silence is the need for clear justifications of government policy. The Government’s previous “closing the gaps” policy suffered from a lack of clear and legitimate justification (James 2000), and the success of the newly branded “social equity” policy will depend partly on not repeating that mistake. The argument over which matters more –within-group variation or between-group variation – boils down to an argument over how we define the policy problem – as poverty or inequality. Chapple’s (2000b:115) conclusion that “effective policy to close the gaps needs to focus on those most disadvantaged or at risk” appears to place him in the poverty camp, but as noted earlier in this paper, his regression analyses point to important interactions between poverty and ethnicity-based inequality at the lower end of the income and employment distributions. Most of the between-group variation described by Te Puni Kokiri (2000) and others may, indeed, reside there.
These interaction effects provide a potentially powerful rationale for government gap-closing policies – namely, the efficiency costs and threats to national unity that occur when certain subgroups of Māori sink deeper into underclass status (poverty) and pull farther away from their non-Māori and mixed-Māori counterparts (inequality). It may be that sole-Māori ethnicity can serve as a predictor of deep labour market disadvantages among low-skilled and rural populations, particularly with regard to labour market entry. If so, and if such disadvantages are allowed to persist and to become more concentrated geographically and demographically, as the deepest forms of disadvantage are wont to do, the effects on productivity (human capital) and cohesion (social capital) may be profound indeed.
Conclusion
All of the Chapple-inspired discussion points deserve extensive scholarly study and public debate. Analysts and political leaders must resist the temptation to use the political controversy over “Māori Socio-Economic Disparity” to silence opponents and block discourse.
The African-American experience of the past 30 years illustrates one complex set of processes by which race-based labour market disadvantage develops, replicates itself, and spins off other forms of social disadvantage. Similarities in the demographic trajectories and macroeconomic experiences of Māori and African-Americans (albeit in different decades) suggest that the story of America’s urban black underclass may offer some useful warnings to New Zealand. The challenges for New Zealand social scientists are (1) to resist the forces from both right and left that would silence debate about these matters; (2) to follow up on the intriguing questions raised by the work of Chapple, Alexander, and others about Māori labour market entry, occupational segregation, and within-group versus between-group disparities; and (3) to explain in the greatest possible detail how the mechanisms of advantage and disadvantage operate in New Zealand.
References
Alexander, W.R.J. and J.R. Williams (2001) "A critique of ‘Māori socio-economic disparity’" Economics Discussion Paper 0107, University of Otago, Dunedin.
Alexander, W.R.J., M. Genc, and M. Jaforullah (2001) "Māori labour market disadvantage" Economics Discussion Paper 0109, University of Otago, Dunedin.
Chapple, S. (2000b) "Māori socio-economic disparity" Political Science, 52(2):101-15.
Frazier, E. F. (1939) The Negro Family in the United States, University of Chicago Press.
Gilder, G. (1981) Wealth and Poverty, Basic Books, New York.
Gould, J. (1996) "Socio-economic differences between Māori iwi" Journal of the Polynesian Society, 105(2):165-183.
Heclo, H. (1986) "The political foundations of anti-poverty policy" in S. Danziger and D. Weinberg (eds.) Fighting Poverty: What Works and What Doesn’t, Harvard University Press, Cambridge, pp. 312-340.
McCreary, J.R. (1968) "Population growth and urbanisation" in E. Schwimmer and J. Forster (eds.) The Māori People in the Nineteen Sixties, Blackwood and Janet Paul, Auckland.
Moynihan, D.P. (1986) Family and Nation, The Godkin Lectures, Harvard University, Harcourt Brace Jovanovich, New York.
Moynihan, D.P. (1965) The Negro Family: The Case for Nation Action, Office of Policy Planning and Research, US Department of Labor, Washington, D.C.
Murray, C. (1984) Losing Ground: American Social Policy 1950-1980, Basic Books, New York.
Pool, I. (1991) Te Iwi Māori: A New Zealand Population Past, Present and Projected, Auckland University Press.
Schwimmer, E. and J. Forster (eds.) (1968) The Māori People in the Nineteen Sixties, Blackwood and Janet Paul, Auckland.
Sowell, T. (1984) Civil Rights: Rhetoric or Reality? William Morrow, New York.
Te Puni Kokiri (2000) Progress Toward Closing Social and Economic Gaps Between Māori and Non-Māori: A Report to the Minister of Māori Affairs, Te Puni Kokiri, Wellington.
Te Ropu Rangahau Hauora a Eru Pomare (2000) "Counting for nothing: understanding the issues in monitoring disparities" Social Policy Journal of New Zealand, 14:1-16.
Wilson, J. Q. (1983) Thinking about Crime, 2nd edition, Basic Books, New York.
Wilson, W.J. (1987) The Truly Disadvantaged: The Inner City, the Underclass, and Public Policy, University of Chicago Press.
Wilson, W.J. (1978) The Declining Significance of Race: Blacks and Changing American Institutions, University of Chicago Press.
Newman, M. (2000) "Labour panics over the gaps policy" Scoop Archive, 3 October
Upton, Hon simon (2000) "Upton-on-line: Olympic kitsch and closing the gaps"
