Infinite-horizon policy-gradient estimation

Gradient-based approaches to direct policy search in reinforcement learning have received much recent attention as a means to solve problems of partial observability and to avoid some of the problems associated with policy degradation in value-function methods. In this paper we introduce GPOMDP, a simulation-based algorithm for generating a biased estimate of the gradient of the average reward in Partially Observable Markov Decision Processes (POMDPs) controlled by parameterized stochastic policies. A similar algorithm was proposed by Kimura, Yamamura, and Kobayashi (1995). The algorithm's chief advantages are that it requires storage of only twice the number of policy parameters, uses one free parameter β ∈ [0,1) (which has a natural interpretation in terms of bias-variance trade-off), and requires no knowledge of the underlying state. We prove convergence of GPOMDP, and show how the correct choice of the parameter β is related to the mixing time of the controlled POMDP. We briefly describe extensions of GPOMDP to controlled Markov chains, continuous state, observation and control spaces, multiple-agents, higher-order derivatives, and a version for training stochastic policies with internal states. In a companion paper (Baxter, Bartlett, & Weaver, 2001) we show how the gradient estimates generated by GPOMDP can be used in both a traditional stochastic gradient algorithm and a conjugate-gradient procedure to find local optima of the average reward. ©2001 AI Access Foundation and Morgan Kaufmann Publishers. All rights reserved.

Distribution trade sector output and productivity performance : a case study of Singapore and Hong Kong 2001-2008

This paper employs the industry of origin approach to compare value added and productivity of Singapore and Hong Kong's Distribution Trade Sector for the period 2001-2008. The direct comparison between these two economies was motivated by the statements of the Singapore government: Its services sector, especially in Retail Trade, lags behind Hong Kong's productivity levels. The results show that since 2005, Singapore's Distribution performance in terms of labour productivity was below Hong Kong's level, which was largely due to poor performance in its Retail Trade sector arising from an influx of foreign workers. Results from total factor productivity (TFP) between these two economies also suggest that Hong Kong's better performance (since 2005) was largely due to its ability to employ more educated and trained workers with limited use of capital. The results suggest that polices that worked in Hong Kong may not work for Singapore because its population is more diverse which poses a challenge to policy-makers in raising its productivity level.

To bound or not to bound: preliminary Theory of Time Bounds Specification in strategic alliances

This article augments Resource Dependence Theory with Real Options reasoning in order to explain time bounds specification in strategic alliances. Whereas prior work has found about a 50/50 split between alliances that are time bound and those that are open-ended, their substantive differences and antecedents are ill understood. To address this, we suggest that the two alliance modes present different real options trade-offs in adaptation to environmental uncertainty: ceteris paribus, time-bound alliances are likely to provide abandonment options over open-ended alliances, but require additional investments to extend the alliance when this turns out to be desirable after formation. Open-ended alliances are likely to provide growth options over open-ended alliances, but they demand additional effort to abandon the alliance if post-formation circumstances so desire. Therefore, we expect time bounds specification to be a function of environmental uncertainty: organizations in more uncertain environments will be relatively more likely to place time bounds on their strategic alliances. Longitudinal archival and survey data collected amongst 39 industry clusters provides empirical support for our claims, which contribute to the recent renaissance of resource dependence theory by specifying the conditions under which organizations choose different time windows in strategic partnering.

Risky strategic behaviour : unintended consequences of good intentions in the homelessness service system

The delivery of human services occurs through a complex and often volatile system characterised by both competing and cooperating efforts. A recent strategic intention of government has been to integrate disparate service providers and programs into a more effective and efficient system using competitive funding regimes. A program of amalgamation has also been forecast and promoted as a further mechanism by which to link up smaller agencies thus creating economy and efficiency in the scale and scope of their service modes. Despite the current reliance on competitive funding models and amalgamation as the preferred ways forward for the sector little is known about their integrative capacity including their ability to predict outcomes and their consequences : the ‘unknown unknowns’. Drawing on an extensive data set of human services integration initiatives in Queensland, Australia, this paper examines the impact of government policy and service models and the risks arising from the tensions between competition and accountability on the one hand and the established good will and trust on the other. It is argued that unresolved, these tensions can lead to a weakening of the social infrastructure and make the system more vulnerable to inherent systemic risks. The paper finds that government’s efforts to externalise risk to the non-government sector leads to fragmentation of the service system and fractured collaborative capability. These unintended outcomes themselves have the unintended consequence of leaving governments disconnected from the service system and unable to provide the leadership role and direction necessary for sustained integration. Moreover, facilitating such a leadership role is undermined by behaviours that are directly contrary to collective integration models.

Comparing retail union strategy in a neoliberal context : the case of Australia and New Zealand

Retail employees are amongst the most vulnerable workers in the context of neoliberal market economies. In many countries, low paid retail employees comprise around 10 per cent of the workforce (ABS 2011). The retail labour market is typically highly feminised and youthful, with many employees in part time and various forms of precarious employment (Tailby & Pollert 2011). However, the industry and its trade unions have rarely been the focus of academic study (Tilly & Carré 2011). This paper thus aims to analyse and compare trade union strategies in the retail industry in Australia and New Zealand, by utilising findings from a larger comparative study. The respective unions studied are the Shop Distributive and Allied Workers Union (SDA) in Australia and the National Distribution Union (NDU) in New Zealand. Data from interviews with union officials at different levels and from different regional locations in Australia and NZ are analysed. Union policy documents are also utilised to support the empirical data. Key findings from the comparison of retail unions’ strategy in Australia and NZ include: 1) the importance of institutional factors and internal political differences in shaping and constraining union strategies; 2) different emphases on external relationships and variations in partnership approaches; 3) the need to recruit to ‘stand still’ by retail unions in both countries; and, 4) similarities and differences in the unions’ organising approaches. The paper concludes by examining the implications of these findings for retail unions’ strategic choices and their ability to deliver workplace justice for employees.

Introduction: Culture, trade, services

Ben Goldsmith and Christina Spurgeon introduce the special 'Culture, trade, services' issue of Media International Australia.

The Creative Industries, Culture and Policy

The rise of creative industries requires new thinking in communication, media and cultural studies, media and cultural policy, and the arts and information sectors. The Creative Industries, Culture and Policy sets the agenda for these debates, providing a richer understanding of the dynamics of cultural markets, creative labor, finance and risk, and how culture is distributed, marketed and creatively reused through new media technologies. This book: develops a global perspective on the creative industries and creative economy draws insights from media and cultural studies, innovation economics, cultural policy studies, and economic and cultural geography explores what it means for policy-makers when culture and creativity move from the margins to the center of economic dynamics makes extensive use of case studies in ways that are relevant not only to researchers and policy-makers, but also to the generation of students who will increasingly be establishing a ‘portfolio career’ in the creative industries. International in coverage, The Creative Industries traces the historical and contemporary ideas that make the cultural economy more relevant that it has ever been. It is essential reading for students and academics in media, communication and cultural studies. Table of Contents - Introduction - Origins of Creative Industries Policy - International Models of Creative Industries Policy - From Culture Industries to Cultural Economy - Products, Services, Production and Creative Work - Consumption, Markets, Technology and Cultural Trade - Globalization, Cities and Creative Spaces - Creative Industries and Public Policy - Conclusion

Mandated literacy assessment and the reorganisation of teachers’ work : federal policy, local effects

This paper explores how mandated literacy assessment is reorganising teachers’ work in the context of Australia’s National Assessment Program – Literacy and Numeracy (NAPLAN), which was implemented in 2008. Students in Years 3, 5, 7 and 9 are tested annually, with school results publicly available. The wider policy context and the emergence of different forms of interconnected educational work associated with the testing phenomenon are described. Taking an Institutional Ethnography approach, the local effects of the federal policy regime are examined through a case study of one school. What mandated literacy assessment does to educators’ work in a culturally diverse low socioeconomic school community is discussed. Key themes include strategic exclusions of students from the testing process, appropriations and adaptations of literacy theory, work intensification, and ethical mediation of results. Questions concerning equity are raised about the differential effects of policy in different school contexts.

Enabling access to research data in developing countries : designing a policy and practice framework for Malaysia’s Public Research Universities

Members of the World Trade Organisation (WTO) are obliged to implement the Agreement on Trade-related Intellectual Property Rights 1994 (TRIPS) which establishes minimum standards for the protection and enforcement of intellectual property rights. Almost two decades after TRIPS was adopted at the conclusion of the Uruguay Round of trade negotiations, it is widely accepted that intellectual property systems in developing and least-developed countries must be consistent with, and serve, their development needs and objectives. In adopting the Development Agenda in 2007, the World Intellectual Property Organisation (WIPO) emphasised the importance to developing and least-developed countries of being able to obtain access to knowledge and technology and to participate in collaborations and exchanges with research and scientific institutions in other countries. Access to knowledge, information and technology is crucial if creativity and innovation is to be fostered in developing and least-developed countries. It is particularly important that developing and least-developed countries give effect to their TRIPS obligations by implementing intellectual property systems and adopting intellectual property management practices that enable them to benefit from knowledge flows and support their engagement in international research and science collaborations. However, developing and least-developed countries did not participate in the deliberations leading to the adoption in 2004 by Organisation for Economic Co-operation and Development (OECD) member countries of the Ministerial Declaration on Access to Research Data from Public Funding, nor have they formulated policies on access to publicly funded research outputs such as those developed by the National Institutes of Health in the United States, the United Kingdom Research Councils or the Australian National Health and Medical Research Council. These issues are considered from the viewpoint of Malaysia, a developing country whose economy has grown strongly in recent years. Lacking an established policy covering access to the outputs of publicly funded research, data sharing and licensing practices continue to be fragmented. Obtaining access to research data requires arrangements to be negotiated with individual data owners and custodians. Given the potential for restrictions on access to impact negatively on scientific progress and development in Malaysia, measures are required to ensure that access to knowledge and research results is facilitated. This paper proposes a policy framework for Malaysia‘s public research universities that recognises intellectual property rights while enabling the open access to research data that is essential for innovation and development. It also considers how intellectual property rights in research data can be managed in order to give effect to the policy‘s open access objectives.

Waste and resource management practices, legislation and policy encouraging and influencing ‘regeneration reuse’ of property assets

Carbon taxation governance is becoming increasingly popular, further evolving the polluter pays concept already well established in the built environment as a mechanism to controlling and licensing waste generation. This paper presents an explanation of property asset ‘regeneration reuse’ principles following deconstruction, which reduce waste generation associated with the process of demolition, construction and operation. An analysis is made of strategies in Australia and the United Kingdom, comparing jurisdiction targets pertaining to construction and demolition waste that encourage ‘regeneration reuse’. From examination of applicable Australian and United Kingdom legislation, strategic, fiscal and policy that influence on the 'regeneration reuse' of property assets, an evaluation to the variety of issues relevant to waste and resource management practices is reached. The paper concludes that a systematic evaluation framework to selecting building components and structures suitable for reuse after deconstruction must be considered in legislation.

Enterprise policy and the metagovernance of firm capabilities

The underlying logic of enterprise policy is that there are impediments to change in economic systems that can be traced to the path-dependent behaviors of economic actors that prevent them from exploring new knowledge and new ways of doing things. Enterprise policy involves firm-level interventions delivered by distributed networks of business advisors coordinated by knowledge intermediaries. These metagovernance arrangements are able to disrupt the path-dependent behaviors of organizations. The logic and benefits of enterprise policy are explored through reference to public administration, strategic management and evolutionary theory, and three case studies.

Asset allocation policy, returns and expenses of superannuation funds : recent evidence based on default options

We examine the asset allocation, returns, and expenses of superannuation funds whose assets are mainly invested in default investment options between 2004 and 2012. A majority of these funds fail to earn returns commensurate with their strategic asset allocation policy. It appears that much of the variation of returns between the funds might be a result of their engaging in significant active management of assets. Our results indicate that returns from active management are negatively related to expenses. We also find strong evidence of economies of scale existing in these superannuation funds across different size categories.

Part-time working by students : is it a policy issue, and for whom?

This paper uses data from interviews with representatives of national and state organisations that have a policy interest in student-working in Australia. The interviewees included representatives from employer bodies and trade unions as well as government organisations. The data are used to discuss these stakeholders’ perceptions of the main advantages and disadvantages of working by young full-time students and the ways in which organisations in the business and educational sectors have adapted their policies and practices for student-working. The analysis is then used to inform a discussion about whether this is a legitimate area for public policy formulation and if so, what principles might underpin such policy and what some policies might look like.