International Journal of Forecasting

Published by Elsevier
Online ISSN: 0169-2070
Publications
Article
"This paper is concerned with forecasting population, employment, labour force and unemployment in Andalusia [Spain] to 1993. For this purpose, a block-recursive demoeconomic model is presented. The demographic submodel is based on the component-cohort method of forecasting population by age and sex. The economic submodel uses econometric and time series analysis to forecast employment both in the medium and short-term. In the labour market part, the labour force is forecasted taking into account the encouraged-discouraged worker effects of changes in the demand for labour."
 
Article
PIP The accuracy of population forecasts depends in part upon the method chosen for forecasting the vital rates of fertility, mortality, and migration. Methods for handling the stochastic propagation of error calculations in demographic forecasting are hard to do precisely. This paper discusses this obstacle in stochastic cohort-component population forecasts. The uncertainty of forecasts is due to uncertain estimates of the jump-off population and to errors in the forecasts of the vital rates. Empirically based of each source are presented and propagated through a simplified analytical model of population growth that allows assessment of the role of each component in the total error. Numerical estimates based on the errors of an actual vector ARIMA forecast of the US female population. These results broadly agree with those of the analytical model. This work especially uncertainty in the fertility forecasts to be so much higher than that in the other sources that the latter can be ignored in the propagation of error calculations for those cohorts that are born after the jump-off year of the forecast. A methodology is therefore presented which far simplifies the propagation of error calculations. It is noted, however, that the uncertainty of the jump-off population, migration, and mortality in the propagation of error for those alive at the jump-off time of the forecast must still be considered.
 
Article
"A common perception among producers (and users) of population projections is that complex and/or sophisticated techniques produce more accurate forecasts than simple and/or naive techniques. In this paper we test the validity of that perception by evaluating the forecast accuracy and bias of eight commonly used projection techniques drawn from...four categories [trend extrapolation, ratio extrapolation, cohort-component, and structural]. Using data for [U.S.] state population projections from a number of different time periods, we find no evidence that complex and/or sophisticated techniques produce more accurate or less biased forecasts than simple, naive techniques."
 
Article
"This paper describes the methodology used to incorporate AIDS mortality in recently revised World Bank population projections.... The paper first reviews different approaches for projecting AIDS and its demographic consequences. This is followed by a summary of an epidemiological model that simulates the spread of HIV used in this analysis, and a demographic model that translates mortality from AIDS into population outcomes. These models are then used in a set of simulations, from which the effect of current HIV prevalence on projected future mortality is extracted. Finally, the extracted equations linking current HIV prevalence with future mortality indicators are applied to sub-Saharan countries with a measurable level of current HIV prevalence."
 
Article
"This paper reviews the two leading methods used to project the number of AIDS cases: back calculation and extrapolation. These methods are assessed in light of key features of the HIV/AIDS epidemic and of data on the epidemic; they are also assessed in terms of the quality of the projections they yield. Our analysis shows that both methods have tended to overproject, often by sizable amounts, the number of AIDS cases in the United States....A new method for projecting AIDS cases is proposed that exploits knowledge about the process generating AIDS cases and that incorporates readily available information about rates of new HIV infection.... Relative to the method of extrapolation, this method projects 22,000 fewer new AIDS cases for 1995 (a 36% difference). This method also projects that intravenous drug users will replace homosexual/bisexual men as the dominant transmission category for AIDS."
 
Article
With increasing interest in forecast uncertainty, there is an evolving concern with assessing the degree of certainty we can attach to uncertainty itself. This concern is the subject of recent work by Juha Alho. This paper reviews his approach to tackling this problem. Alho focuses on conditional forecasts - conditional on the possible future policies feeding into the forecasts. This approach combines simple statistical methods with expert judgement to generate an overall predictive distribution for the future world population from 1994 to 2030. I examine the approach systematically and draw general conclusions about its efficacy. Suggestions for improvement are made.
 
Article
Over 50% of the American women now in the midst of the childbearing years have never borne a child. These levels of childlessness for women in their 20s are the highest ones in a time series that spans most of the twentieth century. Will postponed parenthood be translated into very high levels of permanent childlessness? Or will these cohorts "catch-up" with a late fertility flurry in their 30s and 40s? The authors examine 3 projection strategies: one using women's stated fertility expectations, a second relying on the patterns of previous cohorts, and a third which posits that current rates will persist into the future. The predictive validity of these different projection strategies are tested with data for the 1980-87 period. They show that the projection based on current period rates performs well. Further, the authors argue that it better captures the first birth process than other models. They forecast levels of 20% childless for cohorts of white women born in the early 1960s. Recent trends for nonwhites are very different from those for whites: levels of 4% are forecast for nonwhite women.
 
Article
"The use of the Box-Jenkins approach for forecasting the population of the United States up to the year 2080 is discussed. It is shown that the Box-Jenkins approach is equivalent to a simple trend model when making long-range predictions for the United States. An investigation of forecasting accuracy indicates that the Box-Jenkins method produces population forecasts that are at least as reliable as those done with more traditional demographic methods."
 
Article
"This paper takes [U.K.] General Household Survey (GHS) data at the micro level and ages these households by simulation to the year 2001. Differing scenarios are considered in order to accommodate high and low variants of each household type in the British household distribution."
 
Article
"This article is a review of--and response to--a special issue of Mathematical Population Studies that focused on the relative performance of simpler vs. more complex population projection models. I do not attempt to summarize or comment on each of the articles in the special issue, but rather present an additional perspective on several points: definitions of simplicity and complexity, empirical evidence regarding population forecast accuracy, the costs and benefits of disaggregation, the potential benefits of combining forecasts, criteria for evaluating projection models, and issues of economic efficiency in the production of population projections."
 
Article
PIP Problems concerning the development of long-term forecasts of age variations in the labor supply are discussed. The author examines problems related to using the life cycle theory of labor supply to analyze such variations. An alternative theory, which suggests that the labor supply of younger and older groups is influenced by subsidies from prime-aged groups, is proposed and tested using U.S. data for the last 60 years.
 
Article
"This paper describes a particular approach to stochastic population forecasting, which is implemented for the U.S.A. through 2065. Statistical time series methods are combined with demographic models to produce plausible long run forecasts of vital rates, with probability distributions. The resulting mortality forecasts imply gains in future life expectancy that are roughly twice as large as those forecast by the Office of the Social Security Actuary.... Resulting stochastic forecasts of the elderly population, elderly dependency ratios, and payroll tax rates for health, education and pensions are presented."
 
Article
Changes in the population age structure are known to influence the total income per person, but little is known about whether the changes are equally shared across the population or are concentrated on particular age groups and/or birth cohorts. The answer to this question has potentially important implications for income inequality, for human capital investment, and for fertility decision-making. We propose a new model of intergenerational transfers which distinguishes between the effects of changes in population structure and the effects of changes in family age structure. Using age-specific data from annual income and expenditure surveys of Taiwan between 1978 and 1998, we show that changes in age structure have had a very favorable effect on Taiwan's income growth. The gains are not equally shared by all age groups, however. Children and young adults have benefited the most, while the elderly have benefited the least. The population and family age structures have independent effects on per capita income; the effect of the population age structure is most important. Generational differences in per capita income are closely related to intergenerational differences in earnings, suggesting only a weak form of altruism. Finally, we predict that, on average, population aging will adversely influence per capita income growth in Taiwan in the coming decades.
 
Article
This paper examines forecasted differentials in age-sex-specific mortality in the United States, 1990-2065. A non-linear model, m(x,t) = exp(a(x) + b(x)k(t) + e(x,t)), is fitted for each sex to a matrix of age-specific US death rates, 1933-1988, using SVD to derive a single time-varying index of mortality, k(t). Box-Jenkins techniques are used to estimate and forecast k(t). These forecasts are used to generate age-specific mortality rates and life expectancies to 2065. Independent forecasts of male and female e0's are 82.0 and 90.4, respectively, for 2065, a difference of 8.4 years. These forecasts are substantially higher with narrower confidence intervals than those prepared regularly by the Actuary of the Social Security Administration [Wade (1989)]. These k(t) generated forecasts of e0 appear more plausible than direct forecasts of e0 Life expectancies derived from jointly estimated and forecasted k(t) are competitive with the independent sex forecasts, but have some problems. Joint forecasts of k(t) are juxtaposed to co-integration speculatively as a direction for future research into linkages between male and female mortality.
 
Article
"This paper proposes the use of a log-linear model to obtain long-range micro-level population projections from the aggregative projections. The proposed model is tested using U.S. data. The total population of households is decomposed into 120 subgroups on the basis of the household's region of location, age, size and the female's employment status. The estimated number of households from the log-linear technique is compared against the random-walk and time-trend methods. A comparison of the size of the observed and estimated subgroups for the years 1960 and 1980 suggests that the log-linear technique is superior to the other two methods. Five different measures of errors are used to evaluate the projections."
 
Article
"This paper presents a stochastic version of the demographic cohort-component method of forecasting future population. In this model the sizes of future age-sex groups are non-linear functions of random future vital rates. An approximation to their joint distribution can be obtained using linear approximations or simulation. A stochastic formulation points to the need for new empirical work on both the autocorrelations and the cross-correlations of the vital rates. Problems of forecasting declining mortality and fluctuating fertility are contrasted. A volatility measure for fertility is presented. The model can be used to calculate approximate prediction intervals for births using data from deterministic cohort-component forecasts. The paper compares the use of expert opinion in mortality forecasting with simple extrapolation techniques to see how useful each approach has been in the past. Data from the United States suggest that expert opinion may have caused systematic bias in the forecasts."
 
Article
"The properties and uses of stochastic forecasts are discussed here. For linear stochastic projections, we show how the computation of forecast moments and the statistical distribution of forecasts depend on the multiplicative and autoregressive structure of the dynamics. Both scalar and vector projection methods are discussed, and their similarities are explored. Next we discuss the uses of stochastic forecasts, arguing that it is important to relate forecasts to the specific decision-making criteria of particular forecast users. The example of [the U.S. system of] Social Security is used to show how a dynamic programming approach may be used to explore alternative decisions in a probabilistic context."
 
Article
This paper proposes a new model for population projections. This model projects an initial population under conditions of fertility, mortality, and international migration (like standard cohort-component models), but considers the population arrayed by generation. The model incorporates 4 generations: a foreign-born first generation (the immigrants), a second generation (sons and daughters of immigrants), a third generation (grandsons and granddaughters of immigrants), and fourth-and-higher generations. The model requires fertility, mortality, and migration equations by generation, which take a somewhat different form than in conventional cohort-component population projection. Consideration of the model also makes apparent that assignment of births to generations may not follow a simple form: the paper presents a method for including the empirical description of intergenerational births within the generational framework. As an example, the authors examine the next century of population growth for the Asian, Black, Hispanic, and White non-Hispanic populations in the US, comparing their growth rates and their composition within the total US population. With annual net immigration of 950,000, the total US population of 249 million in 1990 will top 400 million in 2070 and reach about 432 million in 2090. Thus, the level of immigration and emigration assumed in these projections suggests considerable population growth for the next hundred years. The racial/ethnic composition of the US will shift markedly during the next century, as described in the paper.
 
Article
"The purpose of this paper is to extend the headship rate method for projecting households to encompass both sexes. Four models are considered that explicitly incorporate the impact of changes in the number of men and women on the number and joint age distribution of husband-wife households. The models are applied to the Philippines using data from the 1988 National Demographic Survey to project households to 2010. The models are also evaluated by 'backcasting' and comparing the results with special tabulations from the 1970 and 1980 censuses and the 1975 National Demographic Survey."
 
Article
Policymakers need to know whether prediction is possible and, if so, whether any proposed forecasting method will provide forecasts that are substantially more accurate than those from the relevant benchmark method. An inspection of global temperature data suggests that temperature is subject to irregular variations on all relevant time scales, and that variations during the late 1900s were not unusual. In such a situation, a "no change" extrapolation is an appropriate benchmark forecasting method. We used the UK Met Office Hadley Centre's annual average thermometer data from 1850 through 2007 to examine the performance of the benchmark method. The accuracy of forecasts from the benchmark is such that even perfect forecasts would be unlikely to help policymakers. For example, mean absolute errors for the 20- and 50-year horizons were 0.18 � oC and 0.24 � oC respectively. We nevertheless demonstrate the use of benchmarking with the example of the Intergovernmental Panel on Climate Change's 1992 linear projection of long-term warming at a rate of 0.03 � oC per year. The small sample of errors from ex ante projections at 0.03 � oC per year for 1992 through 2008 was practically indistinguishable from the benchmark errors. Validation for long-term forecasting, however, requires a much longer horizon. Again using the IPCC warming rate for our demonstration, we projected the rate successively over a period analogous to that envisaged in their scenario of exponential CO2 growth--the years 1851 to 1975. The errors from the projections were more than seven times greater than the errors from the benchmark method. Relative errors were larger for longer forecast horizons. Our validation exercise illustrates the importance of determining whether it is possible to obtain forecasts that are more useful than those from a simple benchmark before making expensive policy decisions.
 
Article
In this paper we explore the hierarchical nature of tourism demand time series and produce short-term forecasts for Australian domestic tourism. The data and forecasts are organized in a hierarchy based on disaggregating the data according to geographical regions and purposes of travel. We consider five approaches to hierarchical forecasting: two variations of the top-down approach, the bottom-up method, a newly proposed top-down approach where top-level forecasts are disaggregated according to the forecasted proportions of lower level series, and a recently proposed optimal combination approach. Our forecast performance evaluation shows that the top-down approach based on forecast proportions and the optimal combination method perform best for the tourism hierarchies we consider. By applying these methods, we produce detailed forecasts of the Australian domestic tourism market.
 
Top-cited authors
Rob J Hyndman
  • Monash University (Australia)
Anne B. Koehler
Spyros Makridakis
  • University of Nicosia
Eddy Patuwo
  • Kent State University
Peter G. Zhang
  • Georgia State University