The Future Path of The UK Labour Force

Abstract

Introduction

The evolution of the size and composition of the labour force is a key driver of the future productive potential of the economy. Together with population projections they are crucial ingredients in policymakers' decision making on long-term structural issues.

The size and shape of the labour force depends on the composition of the population aged 16+ and age-specific labour market participation rates. The aggregate labour market participation rate in the UK has proved to be relatively stable over the past two decades. However, this hides the considerable historical variation that age-and sex-specific labour market participation rates have shown (Portes, 2012). Over the past three decades the labour market participation rate of men aged 50+ has dropped substantially (Gregg and Wadsworth, 2011). Faggio and Nickell (2003) suggest that some of this drop in labour market participation has been due to men with occupational pensions retiring early, but that a large share of the increase is attributable to a rise in inactivity for health reasons. The participation rate of women aged 50+ has done the exact opposite, and risen dramatically over the past few decades (Gregg and Wadsworth, 2011). The variation of age-specific participation rates occurs for a number of reasons. Some, unsurprisingly, are largely age specific: the decision to participate in full-time education and the decision to retire early to name two. But they are also driven by institutional changes such as active labour market policies of government or changes to the state pension age. They are also determined by wider economic developments. In the recessions of the early 1980s and early 1990s the labour market participation rates of many age groups dropped quite sharply. In contrast, participation rates have remained elevated since 2008. Portes (2012) suggests this is to a large degree due to policymakers avoiding the mistakes of the past and continuing in the direction of active labour market policy that has helped shape the UK labour market since the mid-1980s. Further successful welfare reforms, as well as policy changes designed to facilitate higher employment rates among older workers as healthy life expectancy increases, could lead to future increases in participation rates.

In this note we test the sensitivity of our labour force macroeconomic projections to different variants of future population change. We focus on net migration variants due to the large degree of uncertainty over its contribution to population change. There are recent examples of this uncertainty. ONS (2012), using the Census 2011 results, suggests that the mid-year population estimates for 2011 underestimate the size of the population of England and Wales by 476,000. Around 55 per cent of this under-count is due to an underestimation of net migration over the previous decade; net migration projections based on past trends may therefore be too low. Conversely, the government has committed itself to a migration policy that lowers net migration to the ‘tens of thousands’, the effect of which remains to be seen.

This note is not designed to be an analysis of the impact of migration on the economy; that would require microeconomic analysis of migrants' characteristics and their consequent impact, direct and indirect, on employment, productivity, and wider economic outcomes (see Nathan and Lee, forthcoming). The net migration variant projections published by the ONS are simply a useful tool with which to illustrate the sensitivity of our projections for the labour force and potential GDP to different outturns for population change. We abstract from questions about the difference in productivity levels and participation between migrants and residents. This includes the assumption that age-specific labour market participation rates are exogenous. For an analysis of the economic impact of migration the interested reader should consult Wadsworth (2010).

Figure 1.

Labour market participation rates, per cent, women, 1992–2020

Figure 2.

Labour market participation rates, per cent, men, 1992–2020

The baseline projection

As a baseline for the size and structure of the working age population we use 2010-based ONS principal population projections. Between 2010 and 2020 the total size of population aged 16+ is expected to increase by almost 20 per cent from 50.7 million to 60.3 million people. The labour force is expected to grow at a substantially slower pace, given the aging of the UK population and labour force and the fact that participation rates among older people are significantly lower than average.

Historical age-specific labour force participation rates are derived from the Quarterly Labour Force Survey (QLFS) for 1992–2011.¹ We start by dividing the population aged 16+ into separate groups for both men and women that display similar labour force participation dynamics and/or will be affected by policy changes in a similar fashion. For every group we project the participation rate based on past trends, with some manual corrections to insure consistency; for example, if trend has changed during the past twenty years, only part of the data is taken into account when calculating average growth rates.

One big policy change that will affect the aggregate labour force participation rate is the increase in the state pension age. This policy change is executed in two stages — equalisation of state pension age for women and men at 65 years by December 2018 and further increases for both sexes after that. During our projection period the state pension age will reach 66 years for both sexes by October 2020.²

Combining our projections of age-specific labour force participation rates and 2010-based ONS principle population projections, we arrive at our projection for the UK labour force. The labour force is projected to increase by 7 per cent between 2010 and 2020, from 31.7 million to 33.8 million people.

Sensitivity analysis

Among the components of demographic change, migration has the highest uncertainty. It also has the potential to have a short-term effect on the size and composition of the labour force, while fertility and mortality have a medium- to long-term effect. 2010-based ONS principal population projections have a long-term (starting from 2017)³ assumption of net migration increasing by 200 thousands per year. Prior to this the net migration assumption varies between 209 and 240 thousands. The assumption for 2011 was for net migration of 222 thousand. Current estimates suggest the outturn was 216 thousand.

We illustrate the sensitivity of our labour force projections by substituting each of the low and high net migration variants for the principal population projections published by the ONS (see figures 3 and 4). The low migration variant assumes that long-term net migration will reach 140 thousands per year, which is more in line with (although still falls short of) the government's current policy of reducing net migration to the ‘tens of thousands’. In the case of the low net migration assumption, the labour force in 2020 will be 33.4 million, a reduction of 0.4 million compared with our baseline projection. In the high net migration variant, which assumes that long-term net migration will be 260 thousands per year, the size of the labour force in 2020 will be 34.2 million, 0.4 million higher than the baseline projection.

Figure 3.

Net migration variants (difference from principal), women, thousands 2012–20

Figure 4.

Net migration variants (difference from principal), men, thousands 2012–20

Table 1 presents the results for the low net migration variant, while table 2 presents the simulation results for the high net migration variant. In each of the scenarios we assume that financial markets are forward looking, monetary policy is endogenous over the whole sample period while fiscal policy is assumed to be exogenous out to 2020. Allowing an endogous fiscal policy response within these simulations would allow household direct taxes to adjust so that the budget base remains on baseline. This would result in tighter/looser fiscal policy instruments in the low/high migration scenario, affecting demand within the economy. The low/high variant suggests that the working age population decreases/increases by around 1.3 per cent by 2020. Total population decreases/increases by around 0.8 per cent. Net migration changes are concentrated on those of working age, and in particular those of prime working age. Since we assume the labour market participation rates of migrants and the resident population are equal, the percentage change in the labour force is equal to the percentage change in working age population. Changes in the population above working age have little impact in these scenarios, aside from a modest increase in the labour force.

Table 1.

Low net migration variant

	Working age population		Total population		Labour force		Employment		Potential Output
	Abs. diff. ‘000s	% diff.	Abs. diff. ‘000s	% diff.	Abs. diff. ‘000s	% diff.	Abs. diff. ‘000s	% diff.	Abs. diff. £ bns, 2009 prices	% diff.
2012	−82	−0.2	−84	−0.1	−65	−0.2	−26	−0.1	−1	−0.1
2013	−131	−0.3	−135	−0.2	−106	−0.3	−81	−0.3	−3	−0.2
2014	−194	−0.5	−196	−0.3	−155	−0.5	−126	−0.4	−6	−0.3
2015	−243	−0.6	−250	−0.4	−196	−0.6	−167	−0.6	−8	−0.5
2016	−301	−0.8	−310	−0.5	−243	−0.7	−210	−0.7	−10	−0.6
2017	−358	−0.9	−369	−0.6	−290	−0.9	−254	−0.8	−12	−0.7
2018	−409	−1.0	−420	−0.6	−332	−1.0	−297	−1.0	−15	−0.8
2019	−461	−1.1	−478	−0.7	−376	−1.1	−339	−1.1	−17	−0.9
2020	−518	−1.3	−535	−0.8	−422	−1.3	−381	−1.2	−20	−1.1

Table 2.

High net migration variant

	Working age population		Total population		Labour force		Employment		Potential Output
	Abs. diff. ‘000s	% diff.	Abs. diff. ‘000s	% diff.	Abs. diff. ‘000s	% diff.	Abs. diff. ‘000s	% diff.	Abs. diff. £ bns, 2009 prices	% diff.
2012	83	0.2	87	0.1	66	0.2	25	0.1	1	0.1
2013	142	0.4	144	0.2	112	0.3	83	0.3	3	0.2
2014	190	0.5	196	0.3	152	0.5	127	0.4	6	0.3
2015	248	0.6	256	0.4	199	0.6	168	0.6	8	0.5
2016	300	0.7	308	0.5	241	0.7	210	0.7	10	0.6
2017	355	0.9	365	0.6	287	0.9	252	0.8	12	0.7
2018	414	1.0	425	0.6	334	1.0	296	1.0	14	0.8
2019	465	1.1	477	0.7	376	1.1	339	1.1	17	0.9
2020	518	1.3	535	0.8	421	1.3	381	1.2	20	1.0

We assume wages are set with respect to anchored inflation expectations in both scenarios and that migrants are perfect substitutes for native employees with equivalent productivity levels. In the simulations the change in the labour supply is not immediately matched by demand; the unemployment rate falls/rises temporarily. This puts upward/downward pressure on real wages leading to a fall/rise in the demand for labour allowing the labour market to adjustment to its new equilibrium.

The migration variants are assumed to be unanticipated by business and policymakers. The level of the capital stock in the economy therefore is inconsistent with the equilibrium capital-output ratio. Business investment, in particular, adjusts to ensure the capital stock reaches this new equilibrium. This accelerator effect is one of the key drivers of the change in GDP in the scenarios. Given that we implicitly assume migrants are as productive as natives, the effect on capacity output in the long run is close to the overall effect on employment. However, in the short term there is a positive/negative effect on productivity as the gradual adjustment of the capital stock means there is some temporary capital deepening/shallowing.

Overall, the effect of these scenarios is to decrease/increase the long-run rate of growth by 0.1 percentage point per annum. These scenarios illustrate that changes to the population, resulting from migration policy changes or other factors, will affect the long-run potential of the economy through their direct effect on the size of the labour force. However, it is important to note that in practice the actual impact of changes to migration flows — resulting from policy or other factors — will also depend on the characteristics of migrants themselves, which we do not attempt to model here.

Footnotes

1

The Spring quarter in each year is used.

2

See Barrell et al., 2011, for an analysis of the macroeconomic effects of extending working lives.

3

For more details see .

References

Barrell

, Kirby

and Orazgani

(2011), ‘The macroeconomic impact from extending working lives,’ Department for Work and Pensions Working Paper No. 95.

Department for Work and Pensions (2011), Pension Act 2011, available at http://dwp.gov.uk/policy/pensions-reform/pensions-bill-2011/ accessed on October 22, 2012.

Faggio

and Nickel

(2003), ‘The rise in inactivity among adult men’, in Dickens

, Gregg

and Wadsworth

(eds), The Labour Market under New Labour: The State of Working Britain, Palgrave MacMillan, pp. 40–52.

Gregg

and Wadsworth

(2011), ‘Unemployment and inactivity’, in Gregg

and Wadsworth

(eds), The Labour Market in Winter: The State of Working Britain, Oxford University Press, pp. 22–38.

Nathan

and Lee

(forthcoming), ‘Cultural diversity, innovation and entrepreneurship: firm-level evidence from London’, Economic Geography.

ONS (2006), Labour Force Projections 2006–2020, Lodon, ONS.

ONS (2011), ‘Migration assumptions: 2010-based national population projections’, available at http://www.ons.gov.uk/ons/dcp171776237749.pdf, accessed on October 23, 2012.

ONS (2012), ‘Explaining the difference between the 201 1 census estimates and the rolled-forward population estimates’, available at: http://www.ons.gov.uk/ons/guide-method/census/2011/census-data/2011-census-data/2011-first-release/first-release–quality-assurance-and-methodology-papers/differences-between-2011-census-est-and-rolled-forward-pop-est.pdf.

Portes

(2012), ‘Welfare and work: continuity and change’, National Institute Economic Review, 221, pp. F4–9.

10.

Wadsworth

(2010), ‘The UK labour market and immigration’, National Institute Economic Review, 213, pp. R35–42.