Regional economic indicators

Transcription

1 Economc & Labour Market Revew Vol 4 No 2 February 2010 ARTICLE Sebnem Oguz and Jonathan Knght Regonal economc ndcators wth a focus on dfferences n sub-regonal economc performances SUMMARY Ths quarter, the regonal economc ndcators artcle focuses on explanng the dfferences n sub-regonal Gross Value Added (GVA) per head between and. Ths tme seres analyss splts the dfferences nto f ve explanatory factors: productvty (def ned as GVA per hour worked), hours worked per job, employment rate, commutng rate and actvty rate. The regular part of the artcle then gves an overvew of the economc actvty of UK regons n terms of ther GVA, GVA per head and labour productvty. Ths s followed by a presentaton of headlne ndcators of regonal welfare, other drvers of regonal productvty and regonal labour market statstcs. The ndcators cover the nne Government Off ce Regons of and the devolved admnstratons of ern, Scotland and Wales. These 12 areas comprse level 1 of the European Nomenclature of Unts for Terrtoral Statstcs (NUTS level 1) for the UK. The term regon s used to descrbe ths level of geography for convenence n the rest of ths artcle. Focus on dfferences n subregonal economc performance Regonal economc ndcators hghlght the economc performance of the nne Englsh regons and the three devolved admnstratons, collectvely known as the NUTS 1 regons of the Unted Kngdom. Ths secton of the artcle takes ths analyss to a lower geographcal level by examnng performance wthn regons and comparng these wth each other. It wll evaluate the performance of the 37 NUTS 2 sub-regons and the 133 NUTS 3 areas wthn these regons n terms of ther Gross Value Added (GVA) per head. Lookng at the NUTS 3 level shows whch areas and factors are contrbutng to the growth or declne of economc performance of the larger NUTS 2 sub-regons. The analyss s done by applyng a methodology developed by the Organsaton for Economc Co-operaton and Development (OECD); (see Techncal Note A). Usng ths methodology, workplace GVA per head can be splt nto four explanatory components: average labour productvty employment rate commutng rate actvty rate Average labour productvty s defned as ether GVA per job or GVA per hour worked; employment rate s defned as workplace based employment as a proporton of workforce (a hgh employment rate suggests relatvely low unemployment); commutng rate s defned as the workplace based labour force as a proporton of resdence based labour force; and actvty rate s defned as the proporton of the populaton that s partcpatng n the labour force. Ths breakdown dentfes key factors n the economy that explan regonal dfferences n GVA per head from the UK average. Each component s nfluenced by regonal factors that affect ther contrbuton to the regonal dvergences from the UK average. These regonal characterstcs may be natural endowments (such as geographcal locaton or natural resources that cannot be changed except n the long run) or untapped resources (such as sklls or transport nfrastructure). Usng these defntons to dentfy reasons for dfferences n regonal economc performance wll hghlght specfc ssues that should be addressed by polces n each regon. The May 2008 verson of the Regonal Economc Indcators artcle examned the four component breakdown at NUTS 2 and NUTS 3 level, where average labour productvty was defned as GVA per flled job, from 2001 to. In ths artcle average labour productvty (GVA per job) s further separated nto two elements: GVA per hour worked hours worked per job As a result, a fve component breakdown of workplace GVA per head ncorporates the 31

2 Regonal economc ndcators Economc & Labour Market Revew Vol 4 No 2 February 2010 preferred productvty ndcator of GVA per hour worked and the effect of hours worked per job. The underlyng data comes from varous sources. Resdence-based employment and unemployment were retreved from the Annual Populaton Survey (APS), whle workplace-based employment data were taken from the Sub-regonal Workforce Jobs seres. The latter s compled by combnng several sources, ncludng both household and busness surveys. Work force jobs seres are annually benchmarked to the Annual Busness Inqury (ABI). Sub-regonal hours worked data were compled by combnng Labour Force Survey LFS), sub-regonal workforce jobs and UK hours worked seres. In order to ensure consstency, the data seres have been constraned to ther regonal totals where necessary. Explanng the dfferences n GVA per head from the UK average n NUTS 2 areas, Fgure 1 shows the fve-component breakdown of GVA per head for all 37 NUTS 2 sub-regons from to (the top bars represent and the bottom bars represent ). Inner London s represented n a separate graph as ts GVA per head exceeds the UK average by far more percentage ponts than other NUTS 2 sub-regons. The UK average GVA per head s represented by the vertcal axs at zero. Those components that contrbute negatvely to the relatve GVA per head of a sub-regon are dsplayed to the left of the vertcal axs whle those factors that ncrease sub-regonal economc performance are shown to ts rght. Ths analyss attempts to hghlght the man trends n these components over the perod, and the data have been smoothed usng a fve-perod movng average (Techncal Note B). Fgure1 shows that the above-average ncreasng GVA per head of London as a whole was manly drven by Inner London s economc performance between and. Ths s largely explaned by an ncreasng above average productvty (GVA per hour worked) performance and strong nward commutng rate. Whle productvty explans roughly a thrd of the area s hgh GVA per head performance compared wth the UK average, the nward commutng rate explans half of ts dfference from the UK average over the perod. In Outer London, however, hgher than average productvty and actvty rates are partly offset by large outward commutng rates, leavng ts relatve standng n terms of GVA per head postve and stable throughout the perod consdered. In the, hgher than average GVA per head s manly drven by the Berkshre, Bucknghamshre and Oxfordshre NUTS 2 sub-regon; all fve components had a postve mpact on the economc performance between and. In contrast, four of the fve factors (excludng actvty rate) mpacted negatvely on the economc performance of Kent, producng a below average GVA per head. In Surrey, and Sussex the negatve mpact of the commutng rate and hours per job components on GVA per head are more than offset by hgh productvty, actvty and employment rates, resultng n above average GVA per head. In Hampshre and Isle of Wght hgher than average employment and actvty rates offset the negatve mpact of outward commutng and productvty over the perod. Ths resulted n a GVA per head equal to the natonal average n and (a slght mprovement from 2 percentage ponts below natonal average n the prevous three years). Apart from London and the, Scotland,, of and were the only regons whch contaned hgh performng NUTS 2 sub-regons n. Wthn these hgh-performng subregons, above average GVA per head n Gloucestershre, Wltshre and Somerset, Bedfordshre and Hertfordshre, and Cheshre were manly drven by hgh productvty (GVA per hour worked). However, n ern and ern Scotland t was manly due to the postve mpact of hgh actvty rates. Inward commutng also contrbuted sgnfcantly to the above average GVA per head n ern Scotland. Most NUTS 2 regons at the lower end of the GVA per head performance scale dsplay a combnaton of low productvty and hgh outward commutng. Ths explans ther relatvely low GVA per head. Addtonally, below average actvty rates contrbuted sgnfcantly to the low performance of ern, Wales and the Valleys, Cornwall and Isles of Sclly, Merseysde and Tees Valley and Durham. The combnaton of outward commutng and low productvty also mpacts negatvely on the GVA per head dvergences n sub-regons where the GVA per head s below the UK average but ranked n the mddle or closer to the UK average. The negatve mpact of commutng on GVA per head explans a larger proporton of the dvergence n regons such as Essex and Derbyshre and Nottnghamshre. In regons such as Yorkshre and Herefordshre, Worcestershre and Warwckshre and Angla below average GVA per head s manly drven by lower productvty performance. Actvty rate had a mxed mpact on GVA per head dvergences. In Lecestershre, Rutland and amptonshre for example, a hgher than average actvty rate almost offset the negatve mpact of a lower than average productvty and outward commutng rate on GVA per head over the perod. In, however, a lower than average actvty rate worsened ts GVA per head performance over the perod. Addtonally, from onwards, productvty had an ncreasngly negatve contrbuton to the GVA per head n the. Explanng the dfferences n GVA per head from the UK average n selected NUTS 3 areas, Specfc NUTS 2 regons have now been selected to demonstrate that varatons n GVA per head wthn NUTS 2 areas can be as large f not greater than the varatons between them. Fgure 2 dsplays the economc performance of three NUTS 2 areas wth a postve dvergence from the UK average GVA per head from to (the top bars represent and the bottom bars represent ). Fgure 2 a dsplays the economc performance of the NUTS 2 sub-regon of Surrey, and Sussex, and the contrasts between the four NUTS 3 areas wthn ths sub-regon. The strong negatve nfluence of Sussex CC on the GVA per head of ths sub-regon s clearly outweghed by the postve nfluence of the other three NUTS 3 areas, n partcular, by hgher than natonal average productvty n Surrey and Sussex and hgh actvty rates n Brghton and Hove and Surrey. Whle productvty and actvty rates worsened n Sussex and Brghton and Hove areas over the perod, nward commutng made a postve contrbuton to the GVA per head n Brghton and Hove snce. The above average GVA per head performance of Gloucestershre, Wltshre and Somerset NUTS 2 regon 32

3 Economc & Labour Market Revew Vol 4 No 2 February 2010 Regonal economc ndcators Fgure 1 Explanng the dfferences n GVA per head from the UK average n all NUTS 2 regons, Percentage dfference from UK average Inner London Tees Valley and Durham umberland and Tyne and Wear Cumbra Cheshre Greater Manchester Lancashre Merseysde Rdng and ern Lncolnshre Yorkshre Yorkshre Yorkshre Derbyshre and Nottnghamshre Lecestershre, Rutland and amptonshre Lncolnshre Herefordshre, Worcestershre and Warwckshre Shropshre and Staffordshre Angla Productvty GVA per hour worked Hours per job Employment Rate Commutng Rate Actvty Rate Source: 33

4 Regonal economc ndcators Economc & Labour Market Revew Vol 4 No 2 February 2010 Fgure 1 contnued Explanng the dfferences n GVA per head from the UK average n all NUTS 2 regons, Percentage dfference from UK average Bedfordshre and Hertfordshre Essex Outer London Berkshre, Bucknghamshre and Oxfordshre Surrey, and Sussex Hampshre and Isle of Wght Kent Gloucestershre, Wltshre and Somerset Dorset and Somerset Cornwall and Isles of Sclly Devon Wales and the Valleys Wales ern Scotland ern Scotland ern Scotland Hghlands and Islands ern Irleand Productvty GVA per hour worked Hours per job Employment Rate Commutng Rate Actvty Rate Source: 34

5 Economc & Labour Market Revew Vol 4 No 2 February 2010 Regonal economc ndcators Fgure 2 Explanng the dfferences n GVA per head from the UK average wthn selected NUTS 2 sub-regons, Percentage dfference from UK average (a) Surrey, and Sussex NUTS2 Brghton and Hove Sussex CC Surrey Sussex (b) Gloucestershre, Wltshre and Somerset NUTS 2 Brstol Cty of and Somerset, Gloucestershre Gloucestershre Swndon Wltshre CC (c) Inner London NUTS2 Inner London Inner London Productvty GVA per hour worked Hours per job Employment Rate Commutng Rate Actvty Rate Source: 35

6 Regonal economc ndcators Economc & Labour Market Revew Vol 4 No 2 February 2010 Fgure 3 Explanng the dfferences n GVA per head from the UK average wthn selected NUTS 2 sub-regons, Percentage dfference from UK average (a) Yorkshre NUTS 2 Bradford Leeds Calderdale, Krklees and Wakefeld (b) Derbyshre and Nottnghamshre NUTS 2 Derby Derbyshre and Derbyshre Nottngham Nottnghamshre Nottnghamshre (c) NUTS 2 Brmngham Solhull Coventry Productvty GVA per hour worked Hours per job Employment Rate Commutng Rate Actvty Rate Source: 36

7 Economc & Labour Market Revew Vol 4 No 2 February 2010 Regonal economc ndcators (Fgure 2 b) s largely drven by the economc performance of Swndon and Cty of Brstol where all the components are above natonal average over the and perod. Despte ncreasng outward commutng from Gloucestershre and and Somerset, Gloucestershre, ther productvty, employment and actvty rates remaned above the natonal average throughout the perod consdered. Actvty and employment rates were above average n Wltshre, however, lower than average productvty, and outward commutng causes ths sub-regon to have a lower than average GVA per head. Inner London, whch has by far the hghest GVA per head n the UK, also shows a dvde at the NUTS 3 level between Inner London and Inner London (Fgure 2 c). Inner London, whch ncludes the Cty of London, contrbuted most to the hgh performance of Inner London throughout and. The very hgh GVA per head n Inner London can be explaned largely by the area s strong nward commutng effect and hgher productvty. Snce, nward commutng has been decreasng whle productvty has been ncreasng contnuously. GVA per head n Inner London has also been above the UK average over ths perod but at a much lower level compared to Inner London. Its nward commutng s relatvely low, and ts above average GVA s largely explaned by ts productvty rather than nward commutng. Fgure 3 dsplays the economc performance of three NUTS 2 areas wth a negatve dvergence from the UK average GVA per head from to (the top bars represent and the bottom bars represent ). Fgure 3 a shows the decomposed GVA per head of Yorkshre and ts three NUTS 3 areas. The relatvely low GVA per head of ths regon s drven by Bradford and Calderdale, Krklees and Wakefeld. Whle the GVA per head contnued to decrease from to due to ncreasng outward commutng and declnng productvty and employment rates n the former, the relatve rankng of the latter has not changed sgnfcantly despte ts worsenng productvty because outward commutng has declned. From, productvty was below the natonal average n Leeds. However, sgnfcant nward commutng and above average actvty rates kept the GVA per head n Leeds above the UK average. Fgure 3 b shows the NUTS 2 regon of Derbyshre and Nottnghamshre and ts sx NUTS 3 areas. All sub-regons except Derby and Nottngham show sgnfcant outward commutng, however, commutng nto Nottngham has been declnng over the perod. Actvty rates n Nottngham meanwhle have been negatve but mprovng, reducng ther negatve mpact on GVA per head over the perod. Productvty n all the sub-regons except Derby was relatvely low and n a number of cases worsened through the tme consdered. In, sgnfcant nward commutng nto Brmngham and Solhull mpacted postvely on the regon s GVA per head (Fgure 3 c). Hgher than average (but declnng) productvty n Solhull, however, has been offset by lower than average productvty n Brmngham and Coventry. Ths resulted n the NUTS 2 regon havng a below average productvty between and. The slghtly declnng GVA per head rankng of the NUTS 2 regon was manly due to a declne n the GVA per head n Solhull due to a combnaton of worsenng productvty, nward commutng and actvty rates between and. As n prevous artcles, the analyss n ths secton has shown the mportance of dentfyng dfferences n economc performance at a sub-regonal level. By splttng GVA per head nto fve explanatory factors productvty (GVA per hour worked), hours per job, employment rate, commutng rate and actvty rate ths analyss has examned the economc performance of NUTS 2 regons and explaned the roles of specfc factors nfluencng economc performance. It has shown the performance of NUTS 3 areas wthn selected NUTS 2 regons. The results have shown that larger geographcal areas can hde consderable dfferences at lower scales due to specfc characterstcs of the respectve areas. Regonal overvew Key fgures on a regonal bass ndcate that: n 2008 London was the regon wth the hghest productvty, n terms of GVA per hour worked, at 33 percentage ponts above the UK average and dverged further from t whle ern had the lowest productvty, at 19 percentage ponts below the UK average. and of were the only other regons wth a productvty performance above the UK average (4 and 0.7 percentage ponts respectvely) n the total value of goods exports decreased n all the regons except n Scotland (up by 3 per cent), but there were sgnfcant dfferences among regons. had the largest percentage declne n the value of goods exports (down by 22 per cent). The smallest declne occurred n the (down by 0.4 per cent). the of had the hghest employment rate n the thrd quarter of 2009, at 77.2 per cent; ern had the lowest rate, at 66.1 per cent, compared wth the UK employment rate of 72.5 per cent. Headlne ndcators In order to gan an overvew of the economc performance of UK regons, ths artcle dscusses a selecton of economc ndcators. Currently, the most wdely used ndcator of regonal economc performance s Gross Value Added (GVA) per head. Polcymakers frequently use GVA per head as a headlne ndcator of regonal productvty and of regonal ncomes when comparng and benchmarkng regons that dffer n geographcal sze, economc output and populaton. However, as Dunnell (2009) has explaned, productvty and ncome are very dfferent concepts. GVA per head s calculated as the smple rato of the economc actvty n a regon dvded by the number of people lvng n a regon, whle productvty s defned as the rato of GVA dvded by the labour nput (jobs or hours worked) used to create t. GVA per head does not take account of: people commutng n and out of regons to work regonal dfferences n the percentages of resdents who are not drectly contrbutng to GVA, such as young people or pensoners, and dfferent labour market structures across regons, such as full- and parttme workng arrangements Therefore, GVA per hour worked or GVA per flled job are more approprate productvty ndcators. It needs to be noted that these ndcators also depend on prcng thus productvty can fall/ rse wth decreasng/ncreasng prces. As regonal prce deflators do not yet exst, GVA estmates used n productvty fgures 37

8 Regonal economc ndcators Economc & Labour Market Revew Vol 4 No 2 February 2010 are n nomnal, not real terms, therefore t s not possble to solate volume changes from prce changes. Smlarly, Gross Dsposable Household Income (GDHI) per head s a better measure of regonal ncomes than GVA per head. For example, due to commutng, resdents mght derve ther ncomes from economc actvty n another regon, whch s not captured by GVA per head of ther regon. They may also have sources of ncome whch are unrelated to current work, such as pensons and nvestment ncomes. GDHI, therefore, s one of the determnants of the welfare of the people n the regon. Regonal performance GVA s a good measure of the economc output of a regon. In December 2009, ONS publshed GVA estmates for 2008 and revsed estmates for prevous years. Table 1 shows the regonal economc performance n terms of workplace-based GVA and GVA per head and ther respectve average annual growth over the perod 1998 to Although GVA per head s not a good ndcator of regonal productvty or ncome, t does take account of varatons n geographcal sze among UK regons and therefore allows better comparsons than usng GVA n total. The estmates show that London had the hghest GVA ( bllon) and GVA per head ( 35,000) n 2008, followed by the ( bllon and 21,700, respectvely). London s GVA per head was 71 per cent above the average for the UK, whle that of was 6 per cent above the average. The generated the thrd hghest GVA ( 119 bllon), but was eghth n terms of ts GVA per head ( 17,300). ern had the lowest GVA n 2008, whle Wales had lowest GVA per head (26 per cent below the UK average). In terms of average annual percentage growth of nomnal GVA between 1998 and 2008, London, of,, and ern had the hghest GVA growth. Average annual percentage growth of GVA n these regons was equal to or above the UK growth. The lowest growth occurred n and. Average annual percentage growth of GVA per head between 1998 and 2008 was hgher than the UK average n London, Scotland,, and ern, whle and Yorkshre grew slowest over the same perod. Labour productvty To compare regons n terms of productvty, GVA per hour worked s the preferred ndcator. At lower levels of geography, hours worked estmates are not yet avalable and GVA per flled job should be used. These two measures of productvty dvde GVA by the labour nput, namely hours worked n all jobs or the number of jobs used to create t. GVA per hour worked and GVA per flled job take account of commutng effects and dfferent age profles, and the former also accounts for varatons n labour market structures, such as full- and part-tme workng arrangements and job share avalablty. In February 2010, productvty estmates for 2008 and revsed estmates for prevous years are beng publshed. These estmates make use of the GVA fgures presented n Table 4, and updated flled jobs and hours worked estmates. It should be noted that the productvty fgures presented here use unsmoothed GVA as ther output measure as opposed to headlne GVA, whch s calculated as a fve-year movng average. The unsmoothed measure s used to ensure consstency wth the labour nput data (Dey-Chowdhury et al 2008), but rases some concerns about ncreased volatlty of productvty estmates compared to those based on headlne GVA. The queston of whether to smooth productvty fgures after dvdng unsmoothed GVA by labour data, and presentng these as headlne estmates, s one whch wll be addressed by ONS n the comng months. Fgure 4 shows that n 2008 GVA per flled job and GVA per hour worked exhbted smaller dfferences from the UK average than the catch-all ndcator GVA per head. Ths s manly due to commutng patterns. London, for example, has a very hgh GVA per head, manly due to ncomng workers generatng a hgh GVA, whch s then dvded by a much lower resdent populaton. Productvty ndcators, on the other hand, dvde regonal GVA by the jobs or hours worked used to create t. Fgure 5 shows the regonal GVA per hour worked productvty ndex on a tme seres bass from 2000 to In 2008, London, the and the of were the only three regons wth a productvty performance above the UK average. The of saw the strongest mprovement n ts relatve performance from below the UK average n 2000 to above average n London contnued to mprove ts relatve performance, therefore dvergng further from the UK average. Relatve productvty n the weakened Table 1 Workplace-based gross value added and gross value added per head at current basc prces: by NUTS1 regon UK 1 Yorkshre of London Wales Scotland ern GVA ( mllon) ,500 26,600 78,500 58,000 49,900 63,200 66, , ,200 58,900 29,700 64,600 17, ,259,600 40, ,000 88,500 80,100 94, , , ,100 98,500 45, ,400 28,700 Average annual percentage growth GVA per head ( ) ,200 10,400 11,600 11,700 12,100 12,000 12,600 20,800 13,800 12,100 10,200 12,700 10, ,500 15,800 17,300 17,000 18,100 17,500 19,500 35,000 21,700 18,900 15,200 20,000 16,200 Average annual percentage growth Notes: 1 UK less Extra-rego and statstcal dscrepancy. 2 Provsonal. Source: Regonal Accounts, 38

9 Economc & Labour Market Revew Vol 4 No 2 February 2010 Regonal economc ndcators Fgure 4 Comparson of regonal economc ndcators: by NUTS1 regon, Indces (UK 2 =100) 180 Output per head Output per flled job Output per hour worked Yorkshre Notes: 1 Provsonal. 2 UK less Extra-rego statstcal dscrepancy. of London Wales Scotland ern Source: Fgure 5 GVA per hour worked: by NUTS1 regon Indces (UK 1 =100) Yorkshre of London Wales Scotland ern Notes: 1 UK less Extra-rego and statstcal dscrepancy. 2 Provsonal. slghtly n 2008, but t remaned above the UK average over the perod. ern and Wales had the lowest relatve productvty compared to the UK average n Relatve productvty n most regons dverged from the UK average between 2000 and The strongest dvergence below the UK average productvty over ths perod was experenced n the, Wales and ern. Ths ndcates that these regons productvty grew by less than the UK average, therefore wdenng the productvty gap between regons. Income of resdents The prevous secton dscussed the economc actvty and productvty n the regons. Ths secton dscusses regonal ncomes, whch gve an ndcaton of the welfare of resdents lvng n the regon. Gross dsposable household ncome (GDHI) represents the amount of money avalable to households after taxes, Natonal Insurance and penson contrbutons, property costs and other nterest payments have been deducted. The estmates of GDHI, however, are at current basc prces and so do not take nflaton effects or regonal prce dfferences nto account. In order to make relable comparsons of regonal ncome levels, the analyss needs to take account of relatve szes of regons. Therefore, GDHI per head, whch s a resdence-based measure, s used as an Source: ndcator of the welfare of people lvng n the regon. The May 2009 edton of ths artcle dscussed the latest data on GDHI n detal, therefore ths secton presents a bref overvew of those analyses. Fgure 6 presents ndces of GDHI per head for 2001,, and, showng movements n regonal household ncome relatve to the UK average over tme. It s evdent that the GDHI per head s above the UK average only n the regons of the Greater. Of these regons, London has consstently had the hghest GDHI per head snce 2001 and s dvergng from the natonal average. The and of, on the other hand, are gettng closer to the 39

10 Regonal economc ndcators Economc & Labour Market Revew Vol 4 No 2 February 2010 Fgure 6 Headlne gross dsposable household ncome per head: by NUTS1 regon Indces (UK 1 =100) Yorkshre of London Wales Scotland ern Notes: 1 UK less Extra-rego. 2 Provsonal. Source: Fgure 7 Gross medan weekly pay of all full-tme employees: 1 by NUTS1 regon, Aprl all male female UK average Yorkshre of London Wales Scotland ern Note: 1 Resdents of the respectve regon. natonal average as they experenced the lowest growth n household ncome compared to other regons between 2001 and. Smlarly, mprovements aganst the natonal average are evdent n most regons wth lower household ncome, partcularly the and the devolved admnstratons. Ths ndcates greater party across regons n terms of household ncome. Gross medan weekly earnngs represent another ndcator of regonal welfare. Fgure 7 shows the gross medan weekly pay for all full-tme employees, splt nto female and male full-tme employees, lvng n each regon n Aprl As n prevous years, London was the regon wth the hghest gross medan weekly pay, at , followed by the, at and the of, at These were the only regons above the UK average of ( ), ern ( ), and Wales ( ) recorded the lowest earnngs n Aprl Females across the UK regons receved lower pay than males. In ern, the dscrepancy was smallest, whle t was largest n the and of. In terms of annual average percentage growth over the four years to 2009, pay for females outperformed that for males except n the. The hghest annual average growth rate for female pay was observed n the whle Scotland had the hghest annual average growth rate for male pay between and Source: Annual Survey of Hours and Earnngs, Drvers of productvty HM Treasury and Department for Busness, Innovaton and Sklls (BIS) formerly known as Department for Busness Enterprse and Regulatory Reform (BERR) have dentfed fve key drvers of productvty nvestment, nnovaton, enterprse, competton and sklls that can help explan dfferences n productvty across regons. Alongsde these fve key drvers, other factors, such as connectvty, ndustral structure and regon-specfc assets can have a strong nfluence on regonal productvty performance. Ths artcle uses expendture on Research and Development (R&D) by busnesses as a measure of nnovaton; the numbers of busness brths and deaths 40

11 Economc & Labour Market Revew Vol 4 No 2 February 2010 Regonal economc ndcators and survval rates as an ndcator for enterprse; UK regonal trade n goods serves as a measure of competton; and the qualfcatons of the current workngage populaton and those of young people, who represent the future workforce, to provde an ndcator for the sklls drver. Innovaton Innovaton s a necessary, although not suffcent, condton for economc success and s therefore recognsed as an mportant drver of productvty. Innovaton comprses, among others, the development of new technologes that ncrease effcency and the ntroducton of new, more valuable goods and servces. It also ncludes ntangbles such as new methods of workng and mprovements to servces. R&D represents one of the determnants to the nnovaton process and s defned by the Organsaton for Economc Cooperaton and Development (OECD) n ts Frascat Manual, whch proposes a standard practce for surveys on R&D, as creatve work undertaken on a systematc bass n order to ncrease the stock of knowledge, ncludng knowledge of man, culture and socety, and the use of ths stock of knowledge to create new applcatons. The OECD defnton of R&D covers the followng: basc research: expermental and theoretcal work to obtan new knowledge of the underlyng foundaton of phenomena and observable facts, wthout any partcular applcaton or use n vew appled research: work undertaken to acqure new knowledge, whch s drected prmarly towards a specfc practcal am, and expermental development: systematc work, drawng on exstng knowledge, whch s drected at producng new materals, products or devces, nstallng new processes, systems and servces, or at mprovng substantally those already produced or nstalled The OECD defnton excludes educaton, tranng and any other related scentfc, technologcal, ndustral, admnstratve or supportng actvtes. However, nnovaton depends on a wder set of nputs than R&D, ncludng sklls tranng, desgn, software and organsatonal nvestment by frms. HM Treasury Economcs Workng Paper No. 1 quantfes these broader knowledge economy nputs at UK level; more work s needed before these factors can be measured effectvely at regonal level. Fgure 8 presents statstcs on Busness Enterprse Research and Development (BERD), whch are consstent wth nternatonally agreed standards. Fgures for 2008 publshed on 11 December 2009 show busness expendture on R&D as a percentage of workplace-based GVA n 2000,,, and Ths s a measure commonly used n regonal comparsons as t takes account of the sze of regonal economes. The fgure shows that, snce 2000, the of has been the regon wth by far the hghest percentage of R&D expendture n terms of GVA, wth 3.7 per cent n The and the regons had the second hghest percentage (1.9 per cent) whch has, however, been declnng n the snce These three regons together also accounted for 62 per cent of the total expendture on R&D n London had the lowest R&D expendture as a share of ts regonal GVA n 2008 (0.4 per cent). Yorkshre and the, Wales and Scotland had the second lowest shares n the UK n 2008, at 0.5 per cent each. London s very low share of expendture on R&D does not necessarly suggest low levels of nnovaton but may be due to t havng a large concentraton of servce ndustres, whch may be less R&D ntensve (wthn the OECD defnton) f, for example, they rely heavly on human captal. It may also reflect the choce busnesses make over locatng ther R&D actvtes. Approxmately three quarters of the R&D expendture n the UK was made n the manufacturng sector n Fgure 9 shows that n all regons except London, ern and of the share of the R&D expendture on manufacturng was over 75 per cent of ther respectve expendture. The fgure also shows that of accounted for 26 per cent of the total R&D expendture n the UK n 2008 and had the hghest level of R&D expendture on both manufacturng and servces. Ths may suggest that some London R&D occurs n the surroundng regons such as Cambrdge technology start-ups. Enterprse Enterprse s another drver of productvty. It s defned as the sezng Fgure 8 Busness expendture on R&D as a percentage of workplace-based GVA: by NUTS1 regon Percentages Unted Kngdom 2 Yorkshre of London Wales Scotland ern Notes: 1 Provsonal. 2 UK less Extra-rego and statstcal dscrepancy. Source: Regonal Accounts and Busness Enterprse Research & Development, 41

12 Regonal economc ndcators Economc & Labour Market Revew Vol 4 No 2 February 2010 Fgure 9 Busness expendture on R&D by NUTS1 regon: broad ndustry groups, 2008 mllon 4,500 4,000 Servces Manufacturng and Other 1 3,500 3,000 2,500 2,000 1,500 1, Yorkshre of London Note: Source: Busness Enterprse Research & Development, 1 Other ncludes agrculture, huntng and forestry, fshng, extractve ndustres, electrcty, gas and water supply and constructon. The expendture on other ndustres across the UK was less than 2 per cent of the total expendture. Wales Fgure 10 Enterprse brths, deaths 1 and net change as a percentage of enterprse stock: by NUTS1 regon, 2008 Percentages Brths Deaths Net change Unted Kngdom Note: 1 Provsonal. Yorkshre of new busness opportuntes by both start-ups and exstng frms. New enterprses can brng nnovatve processes and technologes to the market, forcng exstng ones to mprove ther productvty n order to reman compettve. A relatvely large proporton of enterprses jonng and leavng the stock can be seen as desrable, as new enterprses enterng the market are consdered to brng nnovatve processes and technologes that drve up productvty and force unproductve enterprses to leave the market. The February 2009 edton of ths artcle focused on busness demography n UK regons, usng the newly publshed ONS seres of enterprse brths and deaths, whch ncludes enterprses regstered for VAT and also those regstered for pay-asyou-earn (PAYE). It needs to be noted that enterprse statstcs relate to the place of regstraton of the enterprse, even though the enterprse may consst of more than one local unt, possbly n dfferent regons. of London Scotland ern Wales Scotland ern Source: Busness Demography, Fgure 10 shows the number of brths and deaths of enterprses as a proporton of the actve enterprse stock n The dfference between the two represents the net change, whch s calculated as a proporton of total stock. In 2008, across all regons, the net changes were postve due to hgher proportons of enterprses jonng the stock than leavng t. These proportons were largest n London (4.7 per cent), followed by the (2.4 per cent). The lowest rate of net change was n Wales (0.6 per cent). These rates were manly drven by small enterprses wth fewer than 5 employees whch s approxmately 80 percent of the total enterprse stock As well as analysng brths and deaths of enterprses, t s useful to look at how long these enterprses survve. The Busness Demography seres contans data showng the number of years survved by enterprses born n the years to. Fgure 11 shows the proporton of enterprses born n, and that survved for at least three years each. It shows that, overall n the UK, survval rates ncreased from 63.6 per cent of enterprses born n to 65.3 per cent of those born n and went back down slghtly to 64.7 per cent of those born n. Patterns were smlar across regons. In most regons enterprses born n had the hghest three year survval rates compared to and. ern had the hghest three year survval rates whch were above the UK average for the enterprses born n all three years whle London stands out as the regon wth the lowest rates. Fgure 11 has shown that London had the hghest percentage of brths and deaths of enterprses and that survval rates were relatvely low. They could be an ndcaton of London s ablty to explot short-term busness opportuntes. At the same tme, t may suggest that many of the new enterprses born wll not provde long-term growth and employment. Competton Vgorous competton enhances productvty by creatng ncentves to nnovate and ensure that resources are allocated to the most effcent frms. It also forces exstng frms to organse work more effectvely through mtatons of organsatonal structures and technology. One ndcator of competton s the volume of exports. Even though exports do not represent competton wthn a regon, they stll provde an ndcaton of how nternatonal regons are n ther outlook, and how able they are to face global competton. HM Revenue & Customs (HMRC) publshes statstcs on regonal trade n goods to the EU and non-eu destnatons by statstcal value. Trade n goods by defnton excludes trade n ntangbles and servces. The statstcal value of export trade s calculated as the value of the goods plus the cost of movement to the country s border. Table 2 presents the latest quarterly estmates up to September quarter The total value of UK goods exports to all destnatons decreased by approxmately 9.0 per cent between September 2008 and September The total value of goods exports also decreased n all the regons except n Scotland, but there were sgnfcant dfferences among regons. had the largest percentage declne n the value of goods exports 42

13 Economc & Labour Market Revew Vol 4 No 2 February 2010 Regonal economc ndcators Fgure 11 Percentage of unts survvng three years: by year of brth and NUTS1 regon Percentages Unted Kngdom Yorkshre of London Wales Scotland ern Source: Busness Demography, Table 2 UK regonal trade n goods statstcal value of exports: 1 by NUTS1 regon Exports Unted Kngdom Yorkshre of London Wales Scotland mllon ern EU Exports Q4 32,952 1,557 2,854 1,725 2,058 2,314 3,196 2,152 4,891 1,725 1,331 1, Q1 2 34,980 1,634 3,182 1,744 2,196 2,405 3,314 2,304 4,937 1,817 1,485 1, Q2 2 37,251 1,629 3,365 1,885 2,119 2,506 3,595 2,438 5,354 1,937 1,631 1, Q3 2 35,742 1,619 3,283 1,913 2,013 2,137 3,222 2,850 5,096 1,707 1,647 1, Total to September ,925 6,439 12,684 7,268 8,386 9,362 13,327 9,744 20,278 7,185 6,093 6,048 3, Q4 32,677 1,442 2,859 1,826 1,904 1,993 2,895 2,377 5,156 1,562 1,329 1, Q1 2 31,098 1,334 3,093 1,615 1,851 1,791 2,822 2,435 4,908 1,666 1,188 1, Q2 2 29,295 1,310 2,959 1,462 1,782 1,697 2,894 2,391 4,364 1,566 1,178 1, Q3 2 29,837 1,353 2,891 1,479 1,673 1,622 2,920 2,704 4,524 1,416 1,158 1, Total to September ,907 5,439 11,802 6,382 7,211 7,104 11,531 9,907 18,953 6,209 4,853 5,385 3,097 - Non-EU exports - Q4 25,138 1,261 2,462 1,762 1,784 1,801 2,001 3,595 4,125 1, , Q1 2 23,867 1,164 2,452 1,641 1,743 1,767 2,167 3,195 3,892 1, , Q2 2 27,803 1,335 2,862 1,712 1,941 1,989 2,509 3,660 4,993 1,178 1,074 2, Q3 2 28,265 1,357 2,936 1,707 1,914 2,142 2,267 3,577 5,173 1,373 1,312 2, Total to September ,073 5,118 10,712 6,822 7,381 7,699 8,943 14,026 18,183 4,758 4,167 7,896 2, Q4 28,181 1,112 2,807 1,522 2,089 1,900 2,252 3,749 5,430 1,306 1,298 2, Q1 2 22, ,766 1,260 1,958 1,209 1,893 2,711 4,090 1,149 1,074 1, Q2 2 24, ,540 1,263 1,995 1,504 2,001 2,933 4,722 1,164 1,241 2, Q3 2 24,993 1,014 3,383 1,365 1,751 1,588 1,955 2,883 4,654 1, , Total to September ,894 3,984 11,495 5,410 7,792 6,201 8,101 12,276 18,896 4,698 4,545 9,040 2,376 Total Exports Q4 58,090 2,819 5,316 3,488 3,842 4,114 5,197 5,747 9,015 2,879 2,242 3,421 1, Q1 2 58,847 2,798 5,634 3,385 3,939 4,171 5,480 5,499 8,829 2,869 2,354 3,327 1, Q2 2 65,054 2,964 6,227 3,596 4,060 4,495 6,104 6,098 10,347 3,114 2,705 3,558 1, Q3 2 64,008 2,976 6,219 3,620 3,927 4,279 5,490 6,426 10,269 3,080 2,959 3,639 1,498 Total to September ,998 11,557 23,396 14,090 15,768 17,060 22,271 23,770 38,460 11,944 10,261 13,944 5, Q4 60,858 2,555 5,666 3,349 3,993 3,893 5,147 6,126 10,586 2,868 2,627 3,742 1, Q1 2 54,007 2,311 5,859 2,874 3,809 3,001 4,715 5,146 8,998 2,815 2,262 3,305 1, Q2 2 54,106 2,191 5,498 2,725 3,777 3,201 4,895 5,324 9,086 2,730 2,419 3,561 1, Q3 2 54,829 2,366 6,274 2,844 3,423 3,210 4,875 5,588 9,178 2,494 2,091 3,815 1,166 Total to September ,801 9,423 23,297 11,792 15,003 13,305 19,632 22,183 37,849 10,907 9,398 14,424 5,473 Notes: Source: UK Regonal Trade n Goods Statstcs, HM Revenue & Customs 1 Components may not sum to totals as Regonal Trade Statstcs ncludes estmates made for EU trade below the Intrastat threshold whch are ncluded n the unknown regon and not dsplayed n ths table. 2 Provsonal. 43

14 Regonal economc ndcators Economc & Labour Market Revew Vol 4 No 2 February 2010 (down by 22 per cent), followed by (down by 18 per cent) and Yorkshre and the (down by 16 per cent). The smallest declne occurred n the (down by 0.4 per cent). As the European Unon (EU) s the man export destnaton for UK goods, the table separates exports to EU and non-eu destnatons. In the UK as a whole, the value of exports to the EU dropped by 13 per cent between September 2008 and September Wth the excepton of London (up by 2 per cent), all the regons recorded decreases n the value of goods exports to the EU. reported the hghest drop, by 24 per cent. The total value of the UK exports to the rest of the world declned by 4 per cent from September 2008 to September 2009, wth the hghest drop occurrng n the (down by 22 per cent). The rse n exports to non-eu countres, however, masks wde regonal varaton. Fve out of the 12 regons had an ncrease n the value of goods exports to non-eu countres wth Scotland leadng the way (up by 14 per cent). The number of exporters n the UK for the September 2009 quarter compared wth the same quarter last year, decreased by 5.5 per cent to 48,454. London had the largest decrease of 7.5 per cent to 8, There were no regons where the number of exporters ncreased. Fgure 12 shows the value of exports of goods as a percentage of workplacebased regonal GVA n 2000, and 2008, whch takes account of the dfferng szes of regonal economes. In 2008, the value of goods exports relatve to the sze of the regonal economy was greatest n the and lowest n London. It needs to be noted that these fgures show exports of goods as a percentage of headlne GVA whch also ncludes servces and therefore s lkely to underestmate the export performance of some regons wth a large share of servces ndustres such as London. In terms of ths ndcator s change over tme, exports relatve to GVA were lower n all the regons n than n 2000, wth some recovery n 2008 except n, London and Scotland. In Scotland, exports as a percentage of regonal GVA dropped sgnfcantly between 2000 and, but remaned farly stable over the last four years to had the largest ncrease n relatve export performance, followed by ern between and Sklls The sklls of workers nfluence productvty as they defne the capabltes that the labour force can contrbute to the producton process. The concept of sklls ncludes attrbutes of the workforce, such as softer or nterpersonal sklls, whch are dffcult to measure or to compare n dfferent stuatons or over tme. Therefore, qualfcatons are often used as proxy ndcators. By examnng the qualfcatons, such as degree or equvalent, of the current workforce as well as those of young people, who represent the future capabltes of the labour market, a vew of how sklls are changng over tme and ther potental mpact on productvty can be analysed. However, as characterstcs of local economes dctate whch labour sklls are requred, comparablty between regons mght be dffcult. An alternatve approach s to compare the percentage of the workng-age populaton that has no recognsed qualfcatons. Fgure 13 shows the proporton of the workng-age populaton that has no qualfcatons n each regon, alongsde the UK average, for ern had the hghest proporton of the populaton wth no qualfcatons (9.1 percentage ponts above the UK average); whereas the and the had the lowest proportons, 3.8 and 3.7 percentage ponts below the UK average, respectvely. Above average proportons of workngage people wthout a qualfcaton do not necessarly mean that regons have the most unqualfed workforce. Due to dfferng regonal skll requrements, people wth recognsed qualfcatons mght mgrate nto other regons, where demand for ther qualfcatons s hgh, whle those wthout any recognsed qualfcatons mght mgrate out of these other regons. Also, f employers have a strong demand for lower sklls and a good supply of approprate workers, a low skll equlbrum s created n a regon. Regonal Sklls Partnershps (RSPs) are groups brought together by Regonal Development Agences n each regon of n response to the Natonal Sklls Strategy. RSPs am to strengthen regonal structures to make sklls provson more relevant to the needs of employers and ndvduals, coverng prvate, publc and voluntary sectors of the economy. They Fgure 12 Value of total export goods as a percentage of workplace-based GVA: by NUTS1 regon Percentages Unted Yorkshre Kngdom 2 of London Wales Scotland ern Notes: 1 Provsonal. 2 UK less Extra-rego and statstcal dscrepancy. Source: HM Revenue & Customs, Regonal Trade Statstcs and 44

15 Economc & Labour Market Revew Vol 4 No 2 February 2010 Regonal economc ndcators Fgure 13 Workng-age populaton wth no qualfcatons: by NUTS1 regon, 2008 Percentages Yorkshre of Note: Source: Labour Force Survey, 1 For summary of qualfcatons and equvalents see asp?vlnk=836. also am to gve regons the flexblty to tackle ther own ndvdual challenges and prortes. Table 3 presents the RSP core ndcators, whch help to montor the health of regonal and local labour markets and progress towards natonal sklls targets such as those documented n the Letch Report. These core ndcators wll be supported by local, more specfc, ndcators dentfed by ndvdual RSPs. The choce of 19 to state penson age for some of the ndcators n Table 3 has been nfluenced by: the ncreased emphass on educaton and tranng after the age of 16; the plan to rase the standard school leavng age to 18; and algnment wth ndcators specfed n the Local Area Agreements. In order to assess the future capabltes of the labour force, the percentage of pupls achevng fve or more grades A* to C at GCSE level or equvalent n each London UK average Wales Scotland ern Fgure 14 Pupls achevng fve or more grades A* to C at GCSE level or equvalent n () all subjects and () subjects ncludng Englsh and Mathematcs: by NUTS1 regon, 2008/09 1 Percentages Yorkshre All subjects 2 average all subjects of London Subjects ncludng Englsh and Mathematcs 2 average subjects ncludng Englsh and Mathematcs Notes: Source: Department for Chldren, Schools and Famles 1 Revsed data, ncludes attempts and achevements by these pupls n prevous academc years. 2 The average ncludes all schools, not only local authorty mantaned schools. Englsh regon can be used as an ndcator 2. Recent focus on lteracy and numeracy has led to a new measure beng publshed, of fve or more GCSEs grade A* to C n subjects ncludng Englsh and Mathematcs. Fgure 14 shows the percentage of pupls achevng at least fve grades A* to C at GCSE level or equvalent n any subjects, and n subjects ncludng Englsh and Mathematcs. In 2008/2009, the average for pupls n all schools achevng fve or more grades A* to C n any subjects was 70.0 per cent, whle t was down to 50.9 per cent f the subjects ncluded Englsh and Mathematcs. These were ncreases of 4.7 and 3.3 percentage ponts from /08, respectvely. Across all Englsh regons, the percentage of pupls achevng at least fve grades A* to C n subjects ncludng Englsh and Mathematcs was substantally lower compared wth achevng the same n any subjects. Also, regonal dfferences were more pronounced when subjects ncluded Englsh and Mathematcs. In the the percentage of pupls achevng fve or more grades A* to C n any subjects was 2.8 percentage ponts above the average, but the percentage dropped 2.8 ponts below the average when the subjects ncluded Englsh and Mathematcs. The opposte held for the and the of, where the proporton of pupls achevng at least fve grades A* to C ncreased above the average f the subjects ncluded Englsh and Mathematcs whle t dropped below natonal average for achevng fve or more grades A* to C n any subject. London and were the only two regons whch performed above the natonal average on both measures. Investment Investment n physcal captal, such as machnery, equpment and buldngs, enables workers to produce more and hgher qualty output. Therefore, nvestment can have a sgnfcant postve mpact on productvty. Due to qualty concerns regardng the regonal allocatons of nvestment, whch s recorded at the level of the enterprse and not at the local level, ths artcle does not currently nclude data on nvestment. Nevertheless, as Dunnell (2009) has ponted out, nflows of foregn drect nvestment (FDI) projects and estmated numbers of assocated jobs by regon can serve as a narrow ndcator of nvestment. However, FDI does not cover all nvestment n a regon and there s no requrement to notfy UK Trade & Investment when undertakng FDI. The labour market Table 4 shows the seasonally adjusted employment rate, the number of people of workng age n employment, expressed as a proporton of the populaton, from the Labour Force Survey (LFS). In quarter three (July to September) of 2009, the UK employment rate was 72.5 per cent, down 1.9 percentage ponts from a year ago and down 0.2 percentage ponts from quarter two (Aprl to June) of Regonal rates vared from 77.2 per cent n the of to 66.1 per cent n ern. All UK regons experenced annual falls n the employment rate between September quarters 2008 and The largest fall was n ern at 3.7 percentage ponts whle the smallest 45