Abstract
Projections or forecasts of future population, and the age and sex structure of the population, are key inputs to decision-making. However, these projections have an associated uncertainty that is often underappreciated by decision-makers. Moreover, a decision-maker faced with multiple population projections has no clear basis on which to decide between alternative projections. In this chapter, we outline the sources of uncertainty in population projections. We then describe the development of population projection methods in New Zealand, where the representation of uncertainty has become central to official projections produced by Statistics New Zealand, as well as alternative projections produced by the National Institute of Demographic and Economic Analysis. Finally, we outline a model-averaging approach that can be used by decision-makers to combine the information from multiple independent population projections. This may provide a more accurate approach to the use of population projections in decision-making, without requiring substantial modelling capability. This approach is illustrated with the example of subnational areas for the Waikato Region of New Zealand.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
The distinction between a projection and a forecast may appear to be semantic, but the distinction is fundamental to how end-users should interpret and use these information tools. A projection is a realization of a model based on a particular (and known) combination of assumptions in terms of the model parameters. Thus, the projection is fully determined by the choice of parameters (which might be chosen empirically or stochastically, as we discuss later in the chapter). Moreover, a projection is based on current policy settings and does not typically try to anticipate future policy changes, some of which may be in response to projected population changes themselves. In contrast, a forecast is a prediction of a likely future. It is the preference for a particular set of parameter values that turns a projection into a forecast. For further discussion of this point, see Dion and Galbraith (2015). We prefer to use the term ‘projection’, which we maintain throughout the chapter, although we recognize that many end-users will interpret projections as forecasts.
- 2.
A deterministic projection typically combines a given set of input parameters, compared with a stochastic projection which varies randomly according to the probability distributions of those input parameters.
- 3.
For simplicity, we refer to Stats NZ, although before 1995 the organization was known as the Department of Statistics.
- 4.
New Zealand is currently one of only a few countries, along with the Netherlands, where the national statistics agency produces official population projections using a stochastic approach.
- 5.
For simplicity, we set aside the case of small area projections, such as for suburbs or sub-county areas, and household projections, where in both cases an even wider variety of projections methods are arising (e.g. see Cameron and Cochrane 2017).
- 6.
Conversely, in the hands of a savvy operator, a naïve model and a simple assumption about a change in the growth rate might be used to justify almost any policy or plan.
- 7.
For example, Cameron et al. (2007) used additional economic development as a driver of migration, in their subnational population projections for the Hamilton sub-region of New Zealand.
- 8.
Although end-users may disagree with the assumptions or the effect of drivers used in these models.
- 9.
Local council input has been a feature of Stats NZ’s small area (‘suburb’) population projections since they began in the 1990s. Local planners are well placed to inform assumptions about the timing and place of local developments such as new subdivisions and major building projects (e.g. retirement homes, prisons), and about zoning changes that can impact housing and population density. Their input makes the resulting projections more attuned to local plans, if not attuned to local aspirations.
- 10.
Although, as it turns out, those beliefs were often aspirational rather than realistic. See later in this section.
- 11.
- 12.
More accurately, while stochastic population projections could be structured to incorporate some forms of model uncertainty, to date we are unaware of any systematic attempt to incorporate model uncertainty into stochastic population projections. For a notable exception that employs Bayesian model averaging, albeit based on relatively naïve time series models of total population, see Abel et al. (2013).
- 13.
Territorial authorities are the smallest administrative unit in New Zealand. They do not necessarily share contiguous boundaries with regions. The Waikato region is comprised of all, or part, of 11 territorial authorities. However, the proportion of Rotorua District that lies within the Waikato Region is small, so it is excluded from consideration here. Taupo and Waitomo Districts also contain parts that lie outside the Waikato Region, but the populations of those areas are small. However, it should be noted that the NIDEA projections only project the parts of each territorial authority that lie within the Waikato Region and so will be underestimates relative to the total population for Taupo and Waitomo Districts.
- 14.
This normal distribution assumes a mean equal to the medium projection and a standard deviation equal to half the difference between the low and high projections.
References
Abel GJ, Bijak J, Forster JJ et al (2013) Integrating uncertainty in time series population forecasts: an illustration using a simple projection model. Demogr Res 29:1187–1226
Alho JM (1990) Stochastic methods in population forecasting. Int J Forecast 6:521–530
Alho J (1997) Scenarios, uncertainty and conditional forecasts of the world population. J R Stat Soc Ser A Stat Soc 160:71–85
Alho J (2005) Remarks on the use of probabilities in demography and forecasting. In: Keilman N (ed) Perspectives on mortality forecasting: probabilistic models. Swedish Social Insurance Agency, Stockholm, pp 27–38
Alho JM, Spencer BD (1985) Uncertain population forecasting. J Am Stat Assoc 80:306–314
Alho JM, Spencer BD (1997) The practical specification of the expected error of population forecasts. J Off Stat 13:203–225
Alho J, Cruijsen H, Keilman N (2008) Empirically based specification of forecast uncertainty. In: Alho J, Hougaard Jensen S, Lassila J (eds) Uncertain demographics and fiscal sustainability. Cambridge University Press, Cambridge, pp 34–54
Auerbach AJ, Lee RD (eds) (2001) Demographic change and fiscal policy. Cambridge University Press, Cambridge
Bedford R (2005) Demographic forecast, 2051: Movement and change in population and households in the Bay of Plenty. Research report commissioned by Bay of Plenty Regional Council. Population Studies Centre, University of Waikato, Hamilton
Bedford RD, Dharmalingam A (2002) Migration and population forecasts for the Western Bay of Plenty, 2001-2051: Implications for the Labour Force. Paper presented at the 16th conference on labour, employment and work, Wellington, 21–22 Nov 2002
Benenson I (1998) Multi-agent simulations of residential dynamics in the city. Comput Environ Urban Syst 22:25–42
Bijak J (2010) Forecasting international migration in Europe: a Bayesian view. Springer, Dordrecht
Billari F, Graziani R, Melilli E (2014) Stochastic population forecasting based on combinations of expert evaluations within the Bayesian paradigm. Demography 51:1933–1954
Booth H (2006) Demographic forecasting: 1980 to 2005 in review. Int J Forecast 22:547–581
Bryant J (2003) Can population projections be used for sensitivity tests on policy models?. New Zealand Treasury working paper 03/07. New Zealand Treasury, Wellington
Bryant J (2005) What can stochastic population projections contribute to policy analysis? N Z Popul Rev 31:1–11
Cameron MP, Cochrane W (2014) Population, household, and labour force projections for the Waikato region, 2013–2063. Research report commissioned by Waikato Regional Council, University of Waikato, Hamilton
Cameron MP, Cochrane W (2017) Using land-use modelling to statistically downscale population projections to small areas. Australas J Reg Stud 23:195–216
Cameron MP, Poot J (2011) Lessons from stochastic small-area population projections: The case of Waikato subregions in New Zealand. J Popul Res (Canberra) 28:245–265
Cameron MP, Poot J (2014a) Developing a systems-based multi-region stochastic population projections model for New Zealand. Paper presented at the 61st Annual North American Meetings of the Regional Science Association International, Washington, D.C., 12–15 Nov 2014
Cameron MP, Poot J (2014b) Projecting future inter-regional migration using age-gender-specific gravity models—application to New Zealand. Paper presented at the 16th conference on labour, employment and work, Wellington, 27–28 Nov 2014
Cameron MP, Cochrane W, Poot J (2007) End-user informed demographic projections for Hamilton up to 2041. Research report commissioned by Hamilton City Council, Population Studies Centre Discussion Paper No. 66, Population Studies Centre, University of Waikato, Hamilton
Cameron MP, Cochrane W, Poot J (2008a) Population Projections until 2061 for Futureproof—the Hamilton sub-regional growth strategy. Research report commissioned by Hamilton City Council, Waikato District Council, and Waipa District Council, Population Studies Centre, University of Waikato, Hamilton
Cameron MP, Cochrane W, Poot J (2008b) Thames-Coromandel District Council: population, household, and dwelling projections. Research report commissioned by Thames-Coromandel District Council, Population Studies Centre, University of Waikato, Hamilton
Cameron MP, Jackson NO, Cochrane W (2014) Baseline and stochastic population projections for the territorial authorities of the Waikato Region for the period 2013–2063. Research report commissioned by Waikato Shared Services. National Institute for Demographic and Economic Analysis, University of Waikato, Hamilton
Claeskens G, Hjort NL (2008) Model selection and model averaging. Cambridge University Press, Cambridge
Dias L, Morton A, Quigley J (eds) (2018) Elicitation: the science and art of structuring judgment. Springer, New York
Dion P, Galbraith N (2015) Back to the future: a review of forty years of population projections at Statistics Canada. Can Stud Popul 42:102–116
Dunstan K, Ball C (2016) Demographic projections: user and producer experiences of adopting a stochastic approach. J Off Stat 32:947–962
Fontaine CM, Rounsevell MDA (2009) An agent-based approach to model future residential pressure on a regional landscape. Landsc Ecol 24:1237–1254
Garthwaite PH, Kadane JB, O’Hagan A (2005) Statistical methods for eliciting probability distributions. Technical paper 1–2005. Carnegie Mellon University, Pittsburgh
Keilman N (1990) Uncertainty in national population forecasting: issues, backgrounds, analyses, recommendations. Swets and Zeitlinger, Amsterdam
Keilman N (1991) Uncertainty in national population forecasting: issues, backgrounds, analyses, recommendations. Int J Forecast 7:392–393
Keilman N (1997) Ex-post errors in official population forecasts in industrialized countries. J Off Stat 13:245–277
Knutti R, Furrer R, Tebaldi C et al (2010) Challenges in combining projections from multiple climate models. J Climate 23:2739–2758
Lee R, Carter L (1992) Modeling and forecasting US mortality. J Am Stat Assoc 87:659–671
Lutz W (2009) Toward a systematic, argument-based approach to defining assumptions for population projections. Interim report IR-09-037, International Institute for Applied Systems Analysis, Vienna
Lutz W, Scherbov S (1998) An expert-based framework for probabilistic national population projections: The example of Austria. Eur J Popul 14:1–17
Lutz W, Butz W, Samir KC (eds) (2014) World J Am Stat Assoc in the 21st century. Oxford University Press, Oxford
O'Donnell J, Raymer J (2015) Identification change and its effect on projections of the Aboriginal and Torres Strait Islander population in Australia. J Popul Res (Canberra) 32:297–319
O'Hagan A, Buck CE, Daneshkhah A et al (2006) Uncertain judgements: eliciting experts’ probabilities. Wiley, Chichester
Pflaumer P (1988) Confidence intervals for population projections based on Monte Carlo methods. Int J Forecast 4:135–142
Poot J, Alimi O, Cameron MP et al (2016) The gravity model of migration: the successful comeback of an ageing superstar in regional science. Investig Reg J Reg Res 36:63–86
Raymer J, Abel GJ, Rogers A (2012) Does specification matter? Experiments with simple multiregional probabilistic population projections. Environ Plan A 44:2664–2686
Royal Society of New Zealand (2014) Our futures: te pae tawhiti. Royal Society of New Zealand, Wellington
Rutledge DT, Cameron M, Elliott S et al (2008) Choosing regional futures: challenges and choices in building integrated models to support long-term regional planning in New Zealand. Reg Sci Policy Pract 1:85–108
Rutledge D, Cameron M, Elliott S et al (2010) WISE—waikato integrated scenario explorer, technical specifications version 1.1. Research report commissioned by Environment Waikato, Landcare Research, Hamilton
Sceats J (1999) Cohort patterns and trends in the timing and spacing of births. In: Pool I, Johnstone K (eds) The lifecourses of New Zealand women: fertility, family formation and structure, fertility regulation, education, work and economic wellbeing. University of Waikato, Hamilton, Population Studies Centre, pp 41–44
Siegel JS (2002) Applied demography: applications to business, government, law and public policy. Academic Press, San Diego
Statistics Canada (2019) Using expert elicitation to build long-term projection assumptions. Paper presented at the Eurostat-UNECE work session on demographic projections, Belgrade, 25–27 Nov 2019
Statistics NZ (2016) How accurate are population estimates and projections? An evaluation of Statistics New Zealand population estimates and projections, 1996–2013. Statistics NZ. http://archive.stats.govt.nz/browse_for_stats/population/estimates_and_projections/how-accurate-pop-estimates-projns-1996-2013.aspx. Accessed 21 Jan 2020
Statistics NZ (2017a) National ethnic population projections: 2013(base)–2038 (update). Statistics NZ. https://www.stats.govt.nz/information-releases/national-ethnic-population-projections-2013base2038-update. Accessed 21 Jan 2020
Statistics NZ (2017b) National labour force projections: 2017(base)–2068. Statistics NZ. https://www.stats.govt.nz/information-releases/national-labour-force-projections-2017base2068. Accessed 21 Jan 2020
Stoto M (1983) Accuracy of population projections. J Am Stat Assoc 78:13–20
Tebaldi C, Knutti R (2007) The use of the multi-model ensemble in probabilistic climate projections. Philos Trans R Soc A 365:2053–2075
Tuljapurkar S (1992) Stochastic population forecasts and their uses. Int J Forecast 8:385–391
Tuljapurkar S, Lee RD, Li Q (2004) Random scenario forecasts versus stochastic forecasts. Int Stat Rev 72:185–199
van der Gaag N, van Wissen L, Rees P et al (2003) Study of past and future interregional migration trends and patterns within European Union countries: in search for a generally applicable explanatory model. Netherlands Interdisciplinary Demographic Institute, The Hague
Willekens FJ (1983) Multiregional population analysis for urban and regional planning. In: Batty M, Hutchinson M (eds) System analysis in urban policy making and planning. Plenum Press, New York
Woods C, Dunstan K (2014) Forecasting mortality in New Zealand: A new approach for population projections using a coherent functional demographic model. Statistics New Zealand working paper no. 14–01, Statistics NZ, Wellington
Acknowledgements
This research is supported by the New Zealand Ministry of Business Innovation and Employment (MBIE)-funded projects UOWX1404 Capturing the Diversity Dividend of Aotearoa New Zealand (CADDANZ) and CONT-29661-HASTR-MAU Nga Tangata Oho Mairangi. We thank Jacques Poot for his leadership and vision in the development of subnational stochastic population projections models in New Zealand and Ian Pool, John Bryant, Niko Keilman, for comments and suggestions on the development of these methods at the National Institute of Demographic and Economic Analysis (NIDEA) at the University of Waikato. The opinions, findings, recommendations, and conclusions expressed in this paper are those of the authors. They do not necessarily represent those of Stats NZ or NIDEA, who take no responsibility for any omissions or errors in the information contained here.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this chapter
Cite this chapter
Cameron, M.P., Dunstan, K., Cook, L. (2021). The Development of Uncertainty in National and Subnational Population Projections: A New Zealand Perspective. In: Cochrane, W., Cameron, M.P., Alimi, O. (eds) Labor Markets, Migration, and Mobility. New Frontiers in Regional Science: Asian Perspectives, vol 45. Springer, Singapore. https://doi.org/10.1007/978-981-15-9275-1_9
Download citation
DOI: https://doi.org/10.1007/978-981-15-9275-1_9
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-15-9274-4
Online ISBN: 978-981-15-9275-1
eBook Packages: Economics and FinanceEconomics and Finance (R0)