Toggle high contrast

Recession Report November 2009 - international comparisons on the recession

Issue date

Recession Report Number 13: November 2009

Recession Report: Part One

Headline statistics

The latest labour market figures cover the period July to September 2009, and the headlines show the best estimates for key data:

  • The number of people aged over 16 in employment was 28,927,000
    • 15,432,000 men
    • 13,495,000 women
  • The employment rate for people aged over 16 and under 65 (men) or 60 (women) was 72.5%
    • 75.6% for men
    • 69.2% for women
  • The number of people aged over 16 who were unemployed was 2,461,000
    • 1,524,000 men
    • 937,000 women
  • The unemployment rate for people aged over 16 was 7.8%
    • 9.0% for men
    • 6.5% for women
  • 7,997,000 people aged over 16 and under 65/60 were 'economically inactive
    • 3,332 men
    • 4,665 women
  • The economic inactivity rate for people aged over 16 and under 65 was 21.1%
    • 16.8% for men
    • 25.7% for women
  • There were 2,244,000 economically inactive people who wanted a job
    • 976,000 men
    • 1,268,000 women
  • The latest claimant count figures for unemployment (based on the numbers claiming Jobseeker's Allowance) cover October. This figure stood at 1,639,500
    • 1,194,200 men
    • 445,300 women

Revisions of the statistics

Sometimes people are confused because the comparisons in the media's discussions of this month's figures don't seem to fit in with what they remember about last month's figures. There are two reasons for this. The first is that these figures are estimates and usually have to be revised later on. For instance, when the April-June figures were first released, in August, they showed:

Employment

28,933

Unemployment

2,435

Inactivity

7,955

The November release showed the following figures for April-June:

Employment

28,921

Unemployment

2,432

Inactivity

7,956

Occasionally you hear dark mutterings about how after revisions the labour market figures always look worse than they did at first, suggesting that they have been engineered to give a falsely sunny impression. In fact, most revisions are very small (a Government that wanted to bamboozle the public really would be unlikely to bother with such minor changes) and there isn't a consistent trend. The charts below shows the effect of revisions of the employment and unemployment rates - the horizontal axes show the month when the figures came out and the vertical axes shows how much they had been increased or decreased 12 months later.

Effect of employment rate revisions
graph


Effect of unemployment rate revisions

graph

Making comparisons

Probably a more important reason for confusion is the way in which the results are presented. The Labour market figures are published in a 'Statistical Bulletin'; on the front page the latest three-month period is compared with the three-month period that preceded it. For instance, the front page of the latest report compares the July to September figures with the figures for April to June, but last month's figures weren't for April to June, they were for June to August.

Both are useful, in different ways. Thus, for instance, if we compare the figures for July-September with those for April-June, the employment and unemployment levels are both higher. A much more interesting comparison, however, is the difference between the change in the figures for men and women:

Changes in employment and unemployment levels (000s) and rates (%) between Apr-Jun and Jul-Sep

Jul-Sep

Apr-Jun

Difference

Employment

28,927

28,921

+ 6

M

15,432

15,482

- 50

F

13,495

13,439

+56

Employment rate

72.5

72.7

- 0.1

M

75.6

76.0

- 0.4

F

69.2

69.0

+ 0.2

Unemployment

2,461

2,432

+ 30

M

1,524

1,490

+34

F

937

942

- 4

Unemployment rate

7.8

7.8

+ 0.1

M

9.0

8.8

+ 0.2

F

6.5

6.5

- 0.1

If we compare the figures that were released this month with those that were released last month the scale of the changes is, of course smaller, but it rather confirms the picture of a labour market in which overall performance is rather better for women than for men:

Changes in employment and unemployment levels (000s) and rates (%) between Jun-Aug and Jul-Sep

Jul-Sep

Jun-Aug

Difference

Employment

28,927

28,952

- 25

M

15,432

15,467

- 35

F

13,495

13,486

+ 9

Employment rate

72.5

72.6

- 0.1

M

75.6

75.7

- 0.1

F

69.2

69.1

+ 0.1

Unemployment

2,461

2,469

- 8

M

1,524

1,534

- 10

F

937

935

+ 2

Unemployment rate

7.8

7.9

- 0.1

M

9.0

9.0

-

F

6.5

6.5

-

Young people still faring badly

Although the overall figures show either a pause or the possible start of a recovery, there are several factors that show just how fragile this is. Young people are still losing jobs even though other age groups are doing better. If we look at change between the June-August quarter and the July-September quarter, the small overall reduction in employment is the result of a 42,000 reduction in the employment level for under-25s and a 34,000 increase for over-25s:

Change in employment levels by age group

All aged 16

& over

16 - 59/64

16 - 17

18 - 24

Jun-Aug

28,935

27,532

412

3,438

Jul-Sep

28,927

27,524

410

3,398

Change

-8

-8

-2

-40



25 - 34

35 - 49

50 -

Pension age

Over

Pension age

Jun-Aug

6,195

10,882

6,604

1,403

Jul-Sep

6,210

10,898

6,607

1,403

Change

15

16

3

0

If we look at what has happened over the last 12 months to under-25s and compare that with the picture for 'prime age' workers, we can see no sign of a recovery in young people's employment levels:

Employment levels by age

graph

Long-term unemployment

In addition, long-term unemployment is still rising. Some early comments on the labour market statistics reported a fall in long-term unemployment, but it was short-term unemployment that fell. The number of people unemployed over 6 months is still rising:

Unemployment by Duration

All

Up to 6 months

6 - 12

months

12-24

months

Over

24 months

Jun-Aug

2,465,000

1351,000

511,000

355,000

248,000

Jul-Sep

2,461,000

1307,000

537,000

383,000

235,000

Change

-4,000

-44,000

26,000

28,000

-13,000

It is too early to claim that this is now a trend, but another month or two of similar results will definitely establish one. In the first chart below, we compare short-term unemployment and total unemployment; unemployment of less than 6 months' duration stopped rising in the Spring. In the second the comparison is with the number of people unemployed over 6 months, by contrast, this increase does not yet show any sign of easing off.

Total and short-term unemployment (000s)

graph
Total and long-term unemployment (000s)

graph

More than a quarter of unemployed people (25.1%) have now been unemployed for over 12 months.

Involuntary part-time and temporary work

Another sign of weakness is the large number of people who are working in part-time or temporary jobs even though their preference would be for full time or permanent work. The number of temporary workers and the proportion of workers who are temporary have both been rising over the past 12 months. And there has been a long-term trend for the number of part-time workers and the proportion of workers who are part-time to increase; neither trend has shown any sign of abating during the recession.

Temporary employment (000s)

graph
Part-time employment (000s)

graph

The growth of these forms of employment has kept employment levels higher than they would have been otherwise, but the price has been the growth of involuntary part-time and temporary work - rising numbers of workers only in these jobs because they could not find anything else:

Involuntary part-time and temporary work (%)

graph

Involuntary inactivity

A final indicator of labour market fragility has been the recent trend for an increase in the number of 'economically inactive' people who say they want a job. 'Economically inactive' people are those who are not in employment but cannot be counted as unemployed. Most have very good reasons - they are students, or retired, or are sick/disabled or caring for friends or family - but there are always some who, when asked by the Labour Force Survey, say that they would like a job. In recent years the number of people in this position has been quite stable, at around two million, but, since the start of this year, there has been a steady increase:

Inactive by wants a job (000s)

graph
Pay

Many commentators have claimed to be surprised that pay rises are low at present, but one difference from the recessions of the 70s, 80s and 90s is that inflation has been very low in the past 12 months. Frequently, it has been negative, so it is not surprising that, in most sectors, increases in average earnings have also been low:

AEI (including bonuses), change on year (%), 3-month average, Sept 09

Whole economy

1.2

Private sector

0.8

Public sector

2.8

Manufacturing

1.5

Services

1.2

Private services

0.7

AEI (excluding bonuses), change on year (%), 3-month average, Sept 09

Whole economy

1.8

Private sector

1.5

Public sector

2.8

Manufacturing

1.4

Services

2.0

Private services

2.6

And the trend has been for increases to get smaller:

Increases in the Average Earnings Index (AEI)
graph

But a key point to notice about all these figures is that, except for a blip in the figure including bonuses earlier this year, Average Earnings Index changes have been steadily positive. There have been freezes and even some cuts, but changes overall have been positive - sometimes quite strongly so when negative inflation is taken into account. The talk about widespread pay cuts has been proven false in 2009.

Recession Report: Part Two

International comparisons

In this second section of the Recession Report we consider how the UK labour market compares to others across Europe and the developed world. We consider trends in employment, unemployment, and economic inactivity as well as variation in the measures that different governments have taken to tackle the recession.

Employment

Data for the Q2 2009 shows that the EU 27 average employment rate is currently 64.8%.[i] While the UK employment rate (69.6%) remains higher than many other European countries, the UK's rate of annual employment rate change is slightly above the EU average (a drop of 2 percentage points on the year compared to an EU average fall of 1.9 percentage points). The UK's drop in GDP has also been greater than the EU and eurozone averages.[ii]

Relative to the fall in output, employment has fallen less in the UK than on average across the EU - the EU 27 mean is for a 1 point fall in GDP to lead to a 0.39 point drop in employment, but in the UK the ratio is 1 to 0.36. However, there are many other countries where large output falls have led to employment outcomes that are better than the UK. For example, in Germany GDP falls of 5.9 points only led to a fall of 0.1 point in the employment rate and the Netherlands has seen a 5.2 point drop in GDP but only a 0.8 point drop in employment.

Growth (%) of real GDP and employment across the euro area and UK

Country

GDP Q2 2008 - Q2 2009

Employment Q2 2008 - Q2 2009

Slovenia

-9

-1.6

Finland

-8.9

-3

Ireland

-7.3

-8.3

Italy

-6

-0.9

Germany

-5.9

-0.1

United Kingdom

-5.5

-2

Slovakia

-5.4

-1.3

Luxembourg

-5.3

1.3

Netherlands

-5.2

-0.8

EU27

-4.9

-1.9

Euro area

-4.8

-1.8

Austria

-4.5

-1.1

Spain

-4.2

-7.1

Portugal

-3.7

-2.7

Belgium

-3.7

-0.7

Malta

-3

-0.8

France

-2.8

-1.2

Cyprus

-0.7

-0.5

Greece

-0.3

-1

Reductions in hours

Across the EU one of the reasons that falls in employment have been less than contractions in economic activity is the ability of employers to reduce the volume of hours their staff work and to increase the use of part-time employment. Recent analysis from Eurostat[iii] shows that this has been the case in the EU27 and in the euro area between the second quarters of 2008 and 2009. In the year up to the second quarter of 2009, the average number of actual hours worked per week by persons in full-time employment fell by 0.7 hours (from 41.0 hours per week to 40.3) in the EU27 and by 0.8 hours (from 40.8 to 40.0) in the euro area, while between the second quarters of 2007 and 2008 there had been a rise by 0.3 hours in both zones.

The increase in part-time work in the UK has been above the EU27 average, while the reduction in hours has been below average. This could be a consequence of the lack of a short-time working scheme in the UK, and the proliferation of such schemes across many other EU countries (as discussed below). It could also reflect greater initial existence of part-time work in the UK than in other countries.

Percentage points changes in proportions of part-time employment and changes in the number of hours worked on average per week across the euro area and UK, Q2 2008 - Q2 2009

Change in part time work (% of total employment)

Change in hours

Ireland

2.3

-0.7

Slovakia

1.8

-1.4

Slovenia

1.7

-1.1

Austria

1.3

-1.4

Luxembourg

1.2

1

Netherlands

1

-0.3

Spain

0.9

-0.6

United Kingdom

0.8

-0.4

Cyprus

0.7

-0.1

Belgium

0.6

-0.5

Greece

0.5

-0.2

EU27

0.5

-0.7

Finland

0.5

-1.4

Euro area

0.4

-0.8

France

0.1

-0.8

Germany

0.1

-1.3

Malta

-0.3

-0.2

Italy

-0.3

-0.6

Portugal

-0.4

-0.1

Employment and skills

Between the second quarters of 2008 and 2009, employment among those with a low level of education (up to lower secondary education) fell by 4.9% in the EU27 and by 5.4% in the euro area. Among those with a medium level of education (upper secondary and post-secondary non-tertiary education), employment fell by 2.6% in the EU27 and by 1.6% in the euro area. By contrast, employment of those with a high level of education (tertiary education) rose by 3.0% and 2.6% respectively.[iv]

In the UK, the falls in employment for people with low skills have been above the EU27 and euro area averages, as have the increases in employment for those with higher qualifications.

Employment change by level of education across the euro area and UK, Q2 2008 - Q2 2009[v]

Country

Low

Medium

High

Ireland

-17.4

-9.7

0.7

Slovakia

-14.4

-2

6.5

Slovenia

-12.2

-0.3

3.3

Spain

-11.6

-5.9

-2.6

United Kingdom

-6.9

-4.2

3.3

Austria

-6.8

-0.9

6.1

Finland

-6.4

-1.4

-3.4

Portugal

-6.3

6.2

4.1

Euro area

-5.4

-1.6

2.6

EU27

-4.9

-2.6

3

Italy

-4.6

0.3

0.2

France

-3.1

-2.6

5.1

Germany

-2.3

-1.2

5.9

Greece

-1.7

-1.5

0.2

Netherlands

1.6

-1

1.2

Belgium

2.6

-3.5

1.9

Malta

4.4

-5.9

-3

Cyprus

4.8

-2

-1.3

*Reliable data are not available for Luxembourg

Unemployment

Every country across the EU (except Germany) has seen a rise in its unemployment rate since the start of 2008. The highest rates are in Spain and Ireland, with the EU 27 average (which is skewed by the very high rates in these countries) at 9.1 per cent. The UK unemployment rate remains in the bottom half of the distribution, although the rate at which unemployment has increased has been faster than the majority of other EU countries. Countries with stronger labour market regulation (for example France, Germany, Belgium and Italy) have all done rather better than the UK and USA in terms of increase in unemployment.

Unemployment rates and annual change in unemployment rates across the euro area, UK, USA and Japan Jan 2008 - August 2009

Country

August 09 rate

Change Jan 08 - August 09

Spain

18.8

9.8

Ireland

12.5

7.9

Slovakia

11.7

1.3

France

9.8

2.1

United States

9.7

4.8

Euro area

9.6

2.4

Greece*

9.2

1.4

EU27

9.1

2.3

Portugal

9.1

1.4

Finland

8.5

2.1

Belgium

7.9

0.8

United Kingdom

7.8

2.7

Germany

7.6

-0.2

Italy*

7.4

0.8

Malta

7.2

1.2

Luxembourg

6.5

2.2

Slovenia

5.9

1.1

Cyprus

5.6

1.8

Japan

5.5

1.7

Austria

4.7

0.5

Netherlands

3.5

0.7

  • *rates and change from June 2009 (latest period from which data are available).

Economic inactivity

Economic inactivity levels in the UK are far lower than in many other EU countries. In common with many other countries, the UK has also seen a relatively small increase in inactivity levels since the recession began.

Economic inactivity levels (15-59 year olds) across the euro area and UK

Country

2008 Q2

2009 Q2

Change

Malta

35.7

35.1

-0.6

Italy

32.5

33.4

0.9

Belgium

29.1

29.0

-0.1

Greece

29.4

28.7

-0.7

Slovakia

28.0

28.1

0.1

Ireland

26.3

27.3

1.0

Luxembourg

28.6

27.3

-1.3

EU27

25.6

25.3

-0.3

Euro area

24.7

24.6

-0.1

Slovenia

24.5

24.3

-0.2

Spain

24.5

24.0

-0.5

France

25.1

23.9

-1.2

Cyprus

23.9

23.7

-0.2

Portugal

22.9

23.3

0.4

United Kingdom

21.6

21.7

0.1

Germany

20.4

20.2

-0.2

Austria

20.5

20.1

-0.4

Finland

17.9

18.4

0.5

Netherlands

16.2

16.0

-0.2

The OECD's most recent Employment Outlook[vi] included unemployment forecasts until 2011. Unemployment in the UK is projected to peak at 9.6 per cent in 2011. While unemployment is forecast to fall earlier in many other countries, the UK's overall rate is projected to be lower than the euro area average.

Annual unemployment rates (2005 - 2008) and OECD projections for annual unemployment rates (2009 - 2011)

graph

Youth unemployment

Since 2008, there has been significantly variation in the speed at which unemployment among young people has increased. The EU 27 average increase has been 5.4 percentage points, with the average youth unemployment rate now 20.1 per cent. In the UK, the speed of increase has been above average, but the rate remains slightly below that of the EU27.

Unemployment rates among 15-24 year olds UK and EU27 Jan 2008 - Aug 09

graph

  • Educational participation of young people aged
  • 18 is lower in the UK than in most other
  • European countries.[vii] This means that the UK
  • faces a particular challenge with unemployment
  • among very young adults, and makes clear the
  • importance of the current Government focus on
  • increasing educational participation among 16
  • 17 year olds.
  • Rates of educational participation among 18
  • year olds across the EU (2007)

graph

Spain and Ireland have seen increases of over 15 percentage points, and all have youth unemployment rates of over 25 per cent. In contrast, The Netherlands has fared exceptionally well, with the youth unemployment rate now at 6.6 per cent having only increased by 1.5 percentage points since the start of 2008.

Unemployment rates and change (15-24 year olds) Jan 2008 - Aug 2009, euro area, UK and USA[viii]

Country

August 09 rate

Change Jan 08 - August 09

Spain

40.9

20.7

Ireland

26.3

16.9

Slovakia

26.3

7.2

Greece*

25.2

3

Italy*

24.5

4.2

France

24.3

6.4

Finland

22.6

6.4

Luxembourg

21.8

6.6

Belgium

21.2

3.2

EU 27

20.1

5.4

Euro area

20.0

5.5

United Kingdom

19.7

6

Portugal

18.6

2.6

United States

18.2

6.6

Malta

14.9

3

Cyprus

14.1

5

Slovenia

11.7

0.6

Germany

10.7

0.3

Austria

9.9

1.3

Netherlands

6.6

1.5

*rates and change from June 2009 (latest period from which data are available).

Government interventions

Fiscal stimulus

OECD analysis has recently considered the scale of fiscal stimulus packages across the OECD.[ix] The results show that all countries have introduced some degree of stimulus, with the largest packages in the USA, Japan, Korea and Australia. The data show that as a proportion of GDP, the UK package is relatively small. They also demonstrate that the UK is projected to reduce its stimulus at a faster speed than most other OECD countries, and that only the UK and New Zealand are planning to reduce spending in 2010 (with the latter retaining significant tax measures to act as a stimulus). In addition, only two other countries are planning to withdraw combined spending and tax measures faster than the UK, and both have higher automatic stabilisers (Portugal and Norway).

The scale of fiscal stimulus packages across the OCED (as a proportion of 2008 GDP)

graph

Active Labour Market Programmes

The OECD[x] also provides an international overview of labour market programmes. They note that expenditures on labour market programmes vary widely across OECD countries, with spending on active labour market programmes ranging from 0.01 per cent to 1.3 per cent of GDP.

The UK's spending on such programmes is below the OECD average (0.6 per cent of GDP) at 0.3 per cent of GDP. [xi] However, since the start of the downturn the UK has seen significant increases in spending on active labour market programmes. The chart below shows the extent of additional funding that has been made available for labour market programmes across the OECD (it does not include expenditure which increases automatically as unemployment rises).

Discretionary spend on ALMPs across the OECD[xii], average annual expenditure for 2008-10

graph

Short-time working

Two-thirds of OECD countries have reported setting up new short-time working measures, or reinforcing existing measures, in response to the current downturn. They also report on a tendency for these schemes to be combined with access to training. They conclude that 'short-time working schemes are a potentially valuable tool for helping workers faced with substantial earnings losses due to involuntary hours reductions'.[xiii] While noting the potential deadweight and displacement effects these schemes can have, the OECD also recognise that attaching clear and credible time limits to the schemes, and designing interventions to encourage viable firms to self-select into them, can offset these risks.

New research from the European Foundation[xiv] also finds that large numbers of countries have introduced such measures. The study notes that variation in the design and application of the schemes, with some supporting all employees, others focusing only on skilled or full-time workers. There is also variation in the extent of wage compensation that workers receive, which ranges from 50 per cent (Poland and Slovenia) to 80 per cent of the workers' previous wage (Hungary, Italy and Luxembourg). In Wales, workers' receive a subsidy of £50 per day of down-time if they participate in training. Full details of the operation of all the schemes are provided in the full report.[xv]

Replacement rates

The OECD also report on benefit replacement rates (the income that the state provides to unemployed people as a proportion of in-work income) across different countries. Their data show that for prime-age workers entitled to unemployment benefits, net incomes during the first year of unemployment are above 60 per cent in just under half of the countries.[xvi] The UK fares badly in comparison to most other nations, with a net replacement rate of 28 per cent in year 1 of unemployment, the lowest in the EU and the same as the USA. For those experiencing longer periods of unemployment the UK's relative position improves slightly, as in several other countries entitlement to unemployment benefit ends after one year out of work (for example in Italy and the USA). The following table compares the generosity of unemployment benefits across the OECD (rates are taken from 2007).

Net replacement rates at different points during an unemployment spell across the OECD (rates from 2007)

Country

year 1

year 3

year 5

Five-year average

Norway

72

72

72

72

Belgium

65

63

63

63

Austria

61

58

58

59

Denmark

68

68

9

56

Ireland

50

50

50

50

Portugal

79

56

3

48

Germany

64

42

36

45

France

67

31

31

45

Finland

60

33

33

43

Australia

42

42

42

42

Spain

69

25

13

39

New Zealand

38

38

38

38

Sweden

66

41

8

37

Iceland

57

54

8

36

United Kingdom

28

28

28

28

Netherlands

71

3

3

28

Switzerland

80

0

0

24

Luxembourg

87

8

8

24

Canada

52

14

14

22

Hungary

48

13

13

20

Poland

42

8

8

16

Czech Republic

33

11

11

15

Japan

45

3

3

11

Turkey

46

0

0

9

Slovak Republic

32

3

3

9

Greece

33

1

1

8

Italy

37

0

0

7

Korea

31

0

0

6

United States

28

0

0

6

The OECD also consider how unemployment benefits interact with other types of benefit. When unemployment benefits and cash assistance are considered, the levels of assistance provided in the UK remain relatively low.

Unemployment benefits and cash social assistance across the OECD, average net replacement rates over five-year unemployment spell

graph

However, once Housing Benefit is included the UK's position improves considerably. This suggests that if there is a work disincentive problem in the benefit system it is related to Housing Benefit (where marginal effective tax rates are also high) and not to Jobseekers Allowance.

Unemployment benefits, cash social assistance and cash housing-related benefits across the OECD, average net replacement rates over five-year unemployment spell

graph

Conclusion

In comparison with other countries the UK's performance during the downturn is probably best described as average. In common with many other developed countries, employment has fallen, although not as quickly as GDP, unemployment has increased and the lowest skilled and young people have been hardest hit. The UK's strong pre-recession labour market performance is however still bringing us some advantages - levels of economic inactivity remain low and while youth unemployment rates have significantly increased they are still lower than many other EU countries.

Government investment is playing a vital role in supporting labour markets across Europe - and will continue to do so into 2010. Some countries have performed exceptionally well, for example Germany has seen sharp GDP falls but limited employment reductions and a small reduction in economic inactivity. And The Netherlands has retained its pre-recessionary low unemployment and inactivity rates, despite a GDP falls of over 5 percentage points. There may well therefore be lessons that we can learn from the policy responses in these countries, particularly around their use of employment regulation, the levels of investment they make in Active Labour Market Programmes and the scale of their stimulus packages.


[i] Apart from the UK employment rate, these data are not seasonally adjusted.

[ii] Eurostat (2009) The impact of the crisis on employment Luxembourg: Eurostat.

[iii] This analysis is limited to countries for which spending estimates could be obtained. Denmark and Switzerland are not shown because ALMP expenditure automatically rises with unemployment in these countries, greatly limiting the need for discretionary increases.

[iv] Eurostat news release, 5th November 2009, Fewer Hours Worked and More Part-time Work in the EU27.

[v] Low level of education refers to pre-primary, primary and lower secondary education, medium level to upper secondary and post-secondary non-tertiary education and high level to tertiary education.

[vi] OECD (2009) Economic Outlook No. 86 Paris: OECD.

[vii] As above.

[viii] Data downloaded from Eurostat statistics database, 20th November 2009.

[ix] OECD (2009) Employment Outlook: Tackling the Jobs Crisis Paris: OECD.

[x] Ibid.

[xi] Data downloaded from OECD key statistics, 20th November 2009, www.oecd.org/employment/keystatistics.

[xii] OECD (2009) Ibid.

[xiii] Ibid.

[xiv] European Foundation (2009) Restructuring in Recession Dublin: European Foundation for the Improvement of Living and Working Conditions.

[xv] Ibid. p109

[xvi] OECD (2009) Ibid.