Mining closures, gender, and employment reallocations: the case of UK coal mines

Size: px

Start display at page:

Download "Mining closures, gender, and employment reallocations: the case of UK coal mines"

John Kelley
6 years ago
Views:

1 Mining closures, gender, and employment reallocations: the case of UK coal mines Fernando Aragon (SFU), Juan Pablo Rud (Royal Holloway) and Gerhard Toews (Oxcarre) November 24, 2016

2 Collapse In December 2015 the last deep coal mine in Britain closed!

3 Why should we care?

4 Collapse Fastest and most acrimonious deindustrialisation process in the developed world (Beatty and Fothergill, 1996; Glyn and Machin, 1997; Foden et al., 2014). Employment fell from almost 240,000 workers in 1981 to around 60,000 in 1991.

5 Collapse

6 What happens to local economies once the resource is exhausted?

7 Experiment Ideal Experiment: Take several resource rich (local) economies and then randomly destroy the extractive industry in some. Natural Experiment: In 1984 Margaret Thatcher famously declared war to the labour unions and within 10 years nearly completely destroyed the coal industry.

8 Questions What happened to laid-off miners in the long run? What happened to productivity in other sectors? What is the effect on the reallocation of labour?

9 Questions What happened to laid-off miners in the long run? What happened to productivity in other sectors? What was the effect on the reallocation of labour?

10 Main story The lay off of miners in a male dominated industry 1 results in a persistent substitution effect: crowding out of females in other industries. This is important as there is a documented link between women s labour opportunities and......bargaining power within the family;...children s well being;...political influence. 1 Example: Females represent only 4.4% of the labour force in the extractive industries in India.

11 Results For the average mining district, mine closures reduced the share of women in manufacturing 2 by 3 percentage points. This reduction represents around 10% of the initial values in Similar effect in services albeit not as strong.

12 Background Coal played a key role in UK s industrial revolution and subsequent economic growth (Fernihough and O Rourke, 2014). After WWII the coal industry started a long decline due to increased availability of cheaper substitutes. The increase in oil prices in the early 1970s slowed down the decline in production and employment.

13 Strike in 1984 In 1984 the UK government announced the closure of 20 pits and information of the closure of 70 additional pits was leaked to the public. The National Union of Mineworkers called for a general strike. Thatcher declares war: We had to fight the enemy without in the Falklands. We always have to be aware of the enemy within, which is much more difficult to fight and more dangerous to liberty. 30% of pits closed by 1986 and 90% by 1994.

14 Collapse

15 Predictions We use the model to derive 4 intuitive hypotheses: 1 Decrease in population size and employment; 2 Re-allocation of workers from mining to manufacturing (Corden and Neary, 1982); 3 Substitution of male for female workers in manufacturing; 4 Decrease in wages of male and female workers.

16 What do we do?

17 Data Mining Data: Collect data on the location of mines and the number of employees. Population and Employment Data (by gender and industries): UK population Census aggregated to 339 districts. Employ. and Wage Data (by gender, industries and occupation): New Earning Survey aggregated to 50 counties.

18 Identification We exploit the dramatic closure of coal mines during the 1980s as a quasi natural experiment. Empirical strategy is essentially a Diff-in-Diff: compare mining districts and neighbouring districts before and after the closure.

19 Location of mines

20 Treatment and Control in 1981 Mining Non-mining p-value Mean S.D. Mean S.D. (1)=(3) (1) (2) (3) (4) (5) Nr. active mines Population ( 000s) Participation rate (%) Unemployment rate (%) Nr. of workers ( 000s) Primary Manufacturing Services Number of districts % female workers in: Primary Manufacturing Services Number of regions Note: Column 5 displays the p-value of a mean comparison test of columns 1 and 3.

21 Estimation y it = γnrmineclosures it + α i + δ t + ɛ it OLS with region and year FE; y it is a placeholder for population, unemployment, labour force participation, employment and wages. Nrmineclosures is the cumulative number of mines closed since 1975/1981. Highest possible level of clustering (at least 36).

22 Share of female manufacturing workers

23 Aggregate Effect Nr. primary ln(pop.) ln(nr. Particip. Unemploy. workers 000s workers) rate rate (1) (2) (3) (4) (5) A. Men Nr. of mines *** ** *** *** closed since 1981 (0.066) (0.002) (0.003) (0.074) (0.075) B. Women Nr. of mines * * ** closed since 1981 (0.015) (0.002) (0.003) (0.079) (0.047) Notes: Robust standard errors in parentheses. Standard errors are clustered at county level. * denotes significant at 10%, ** significant at 5% and *** significant at 1%. All regressions are estimated using OLS, and include district and year fixed effects. Sample includes districts within 30 miles of a mine. Panel A reports estimates using outcomes for men, while Panel B uses outcomes for women. Primary sector includes mining plus agriculture, forestry, fishing, energy and water supply. Number of primary workers is measured in thousands. Number of observations = 696, number of districts=174.

24 Substitution A. Manufacturing ln(nr. of workers) % female Total Women Men workers (1) (2) (3) (4) Nr. of mines closed 0.011* ** 0.022*** *** since 1981 (0.006) (0.006) (0.007) (0.140) B. Services Nr. of mines closed ** *** since 1981 (0.005) (0.004) (0.005) (0.064) Notes: Robust standard errors in parentheses. Standard errors are clustered at county level. * denotes significant at 10%, ** significant at 5% and *** significant at 1%. All regressions are estimated using OLS, and include district and year fixed effects. Sample includes districts within 30 miles of a mine active in Number of observations = 696, number of districts=174. Robustness

25 Wages Dep. variable = Ln(wage) Manufacturing workers Service workers Female Male Female Male (1) (2) (3) (4) Nr. of mines ** ** * closed since 1975 (0.0007) (0.0006) (0.0005) (0.0009) Nr. Obs. 212, , , ,805 Notes: Robust standard errors in parentheses. Standard errors are clustered at county level. * denotes significant at 10%, ** significant at 5% and *** significant at 1%. All regressions are estimated using OLS, and include country-by-year and county fixed effects, county-specific trends, and individual controls such as: age and its square, occupation and industry dummies and indicators of full time job, junior position, being a experienced worker, having an additional job, and being under a national or subnational wage agreement. Columns 1 to 2 use a sample of manufacturing workers. In column 3 to 4 we use the sample of service worers (excluding government, health and education and research workers). Sample includes only counties located within 30 miles of a mine for years 1975 to Number of counties = 38.

26 Mining closures, gender, and employment reallocations: the case of UK coal mines Fernando Aragon (SFU), Juan Pablo Rud (Royal Holloway) and Gerhard Toews (Oxcarre) November 24, 2016

27 Assumptions 1 1 factor of production = labour: two types of workers: men and women 2 Small local economy: two tradable goods sectors: mining and manufacturing workers are mobile, but have idiosyncratic preferences over location 3 Gender bias: mining employs only men manufacturing employs both (imperfect substitutes) Substitution 4 Mine closure: negative shock to demand of male workers

28 Wage Figure: Miners S D D Nr.Miners

29 Figure: Manufacturing Wage Wage S S S D D D Nr.Males Nr.Females Back

30 Robustness share of female manufacturing workers (1) (2) (3) (4) (5) (6) Nr. of mines *** *** *** *** *** *** closed since (0.176) (0.104) (0.142) (0.120) (0.159) (0.148) 1981 Robustness All Only mining EU & UK Non-param. IV Conley check: districts districts regional funds trends S.E. Observations 1, R-squared Nr. districts Notes: Robust standard errors in parentheses. Standard errors are clustered at county level. * denotes significant at 10%, ** significant at 5% and *** significant at 1%. All regressions are estimated using OLS, and include district and year fixed effects. Columns 1-2 change the sample definition. The baseline sample refers to districts within 30 miles of a mine. Column 3 includes the log of UK and EU regional funds as proxy for regeneration policies. Column 4 includes region-by-year fixed effects and interaction of year fixed effects with quartiles of distance to London and indicators of above-the-median values in 1981 of population size, manufacturing and service employment, and share of female manufacturing and service workers. In column 5 we instrument the number of closed mines with the price of coal nteracted with the number of active mines on the district in The first stage F-test is 1,722. Column 6 estimates the baseline regression with standard errors corrected for spatial and serial correlation using the procedure described by conley2008. Back

31 Demographics % female % prime % population with Children % single pop. age pop. tertiary education per indiv. Women Men woman (1) (2) (3) (4) (5) (6) Nr. mines 0.041*** *** ** *** 0.229*** closed since 1981 (0.012) (0.058) (0.059) (0.058) (0.002) (0.053) Notes: Robust standard errors in parentheses. Standard errors are clustered at county level. * denotes significant at 10%, ** significant at 5% and *** significant at 1%. All regressions are estimated using OLS, and include district and year fixed effects. Sample is the same as in baseline regression. Columns 1 and 2 use as outcomes the population share of women and prime age individuals (16-44 years old). Columns 3 and 4 use the share of population over 16 years with tertiary education. Column 6 uses the ratio of population age 0 to 15 years to women age

32 Persistence ln(pop.) Particip. Manufacturing rate ln(nr. of workers) % female Women Men workers (1) (2) (3) (4) (5) Mining district *** ** *** year 2001 (0.022) (0.708) (0.057) (0.049) (0.843) Mining district * ** *** *** year 2011 (0.026) (0.538) (0.063) (0.060) (0.862) Observations R-squared Notes: Robust standard errors in parentheses. Standard errors are clustered at county level. * denotes significant at 10%, ** significant at 5% and *** significant at 1%. All regressions are estimated using OLS, and include district and year fixed effects. Sample is the same as in baseline regression. Mining district is an indicator equal to 1 if the district contains at least one mine. "year 2001" is an indicator equal to 1 if year is 2001, likewise for "year 2011".

33 Persistence

34 by Age Dep. variable = ln(number of workers in sector) Women Men (1) (2) (3) (4) (5) (6) A. Manufacturing Nr. of mines *** ** *** 0.024*** 0.028*** closed since 1981 (0.007) (0.006) (0.009) (0.007) (0.007) (0.009) B. Services Nr. of mines ** 0.015** closed since 1981 (0.004) (0.004) (0.005) (0.004) (0.005) (0.006) C. Construction and transport Nr. of mines ** 0.015* closed since 1981 (0.009) (0.008) (0.010) (0.008) (0.007) (0.008) D. Retail, catering and other services Nr. of mines ** 0.015** closed since 1981 (0.004) (0.004) (0.005) (0.005) (0.005) (0.006) Notes: Robust standard errors in parentheses. Standard errors are clustered at county level. * denotes significant at 10%, ** significant at 5% and *** significant at 1%. All regressions are estimated using OLS, and include district and year fixed effects. Sample includes districts within 30 miles of a mine active in Number of observations = 696, number of districts=174.

35 Manual Workers ln(nr. of workers) % non-manual Non-manual Manual workers (1) (2) (3) A. Manufacturing Nr. of mine workers 0.012* 0.010* laid-off since 1981 (0.006) (0.006) (0.001) B. Services Nr. of mine workers *** laid-off since 1981 (0.004) (0.004) (0.001) Notes: Robust standard errors in parentheses. Standard errors are clustered at county level. * denotes significant at 10%, ** significant at 5% and *** significant at 1%. All regressions are estimated using OLS, and include district and year fixed effects as in Table 24. Panels A and B report estimates using outcomes for different sectors. Sample includes districts within 30 miles of a mine active in Number of observations = 696, number of districts = 174.

OxCarre Research Paper 161. Mining Closure, Gender and Employment Reallocations: The case of UK coal mines. Fernando M Aragón.

OxCarre Research Paper 161. Mining Closure, Gender and Employment Reallocations: The case of UK coal mines. Fernando M Aragón. DEPARTMENT OF ECONOMICS OxCarre Oxford Centre for the Analysis of Resource Rich Economies Manor Road Building, Manor Road, Oxford OX1 3UQ Tel: +44(0)1865 281281 Fax: +44(0)1865 271094 oxcarre@economics.ox.ac.uk