Industrial Restructuring of the Prairie Labour Force: Spatial and Gender Impacts

Indstrial Restrctring of the Prairie Labor Force: Spatial and Gender Impacts M. Rose Olfert and Jack C. Stabler Department of Agricltral Economics University of Saskatchewan Saskatoon, SK S7N OWO The economic restrctring of the Prairies that has characterized the latter part of this centry, is nowhere more apparent than in the decline of the agricltre sector. Even dring periods of increasing vale of agricltral prodction, the farm poplation has been steadily declining in absolte terms (Stabler and Olfert 1992). Labor-saving technological change, together with incomeinelastic world demand for cereals, a relatively fixed qantity of land resorce, and the competitive market in which agricltral otpt is sold, have all contribted to a redction in farm labor reqirements. Of corse, the absolte decrease in agricltre employment that is occrring at the same time other sectors are expanding, translates into a sbstantial redction in the relative importance of agricltre. In addition to the traditional adjstment throgh exit, a partial adjstment of labor has recently appeared. That is, farm operators and members of farm families participate in the non-farm labor market while contining their farm prodction. Agricltre policy has been directed primarily at spporting and stabilizing income associated with farm prodction for those remaining in primary prodction. Mch less attention, and money, has been directed towards encoraging and facilitating the reallocation of labor ot of the indstry, either for those exiting the indstry or for those making the adjstment by combining non-farm employment with contined farm prodction. There are two exceptions -- the Canadian Rral Transition Program (CRTP) and the recently annonced (Agst 1993) Canadian Rral Opportnities Initiative (CROI). These are both retraining programs directed at the farm labor force. The former, CRTP, reqires that farm families exit from farming in order to qalify for assistance. The newer program, CROI, is designed to provide training, conselling, and bsiness development assistance to farm families ta increase «.1 Canadian Jornal of Regional Science/Reve canadienne des sciences régionales, XVII:2 (Smmer/été 1994), 133-152. ISSN: 75-458 Printed in Canada/Imprimé a Canada

134 OLFERT AND STABLER INDUSTRIAL RESTRUCTURING OF THE PRAIRIE LABOUR FORCE 135 and diversify their income withot having to leave farming. Sccessfl assistance to adjstment of the labor force ot of primary agricltre prodction into other sectors reqires consideration of the more general indstrial restrctring of the economy. Frther, if there is sorne vale attached to retaining the poplation in rral areas, the metro/non-metro distribtion of growth indstries is an important consideration. A rral poplation is necessary to spport the commercial and pblic services provided for farmers and those involved in resorce extractive activities. Farm women are consistently engaged in non-farm employment to a mch greater extent than farm men. The ability of the agricltre labor force to adjst to the redced on-farm labor reqirements by increased participation in non-farm employment will depend to a significant degree on the participation of women in the positive aspects of indstrial restrctring. It is the objective of this paper to describe the differential impacts of indstrial restrctring of the Prairie economy on metro and non-metro areas as weil as on males and females. The distribtion of employment growth over metro and non-metro areas and between males and females will help to define the opportnities faced by farm farnilies making either a complete or a partial adjstment ot of primary prodction agricltre. Other Stdies Indstrial restrctring is occrring globally and is the sorce of both economic growth and strctral nemployment, as the labor force adjsts with a lag. Restrctring is the reslt of a combination of changes in final and intermediate demand for goods and services, as weil as technological changes that have led to the sbstittion of capital for labor at differential rates among indstries. While the precise natre of the restrctring varies from one place to another, the predominant characteristic is the relative decline of the goodsprodcing sectors as sorces of employment and a dramatic increase in the importance of the service sectors. Urban and rral areas participate to different degrees in the restrctring process, depending on their existing economic bases and their ability to attract new firrns. Persistent differences in the indstrial distribtions of females and males sggest that the restrctring process may also have a gender bias. Rral commnities in agricltre-dependent regions are negatively affected by the poplation decreases associated with the redction in farm nmbers. In addition, changes in transportation, commnication, and distribtion technologies have led to consolidation of many bsiness fnctions into larger centres. The consmer preferences of rral dwellers, reflected in their shopping patterns, sggest a strong bias in favor of the larger centres, again at the expense of smaller centres. Poplation redistribtion also reslts in the consolidation of pblic infrastrctre, albeit with a lag. The decline in agricltre employment has coincided with a rapid increase in female participation rates. Generally, female participation rates in rral areas are lower than in rban areas, althogh the difference is narrowing (Godwin and Marlowe 199; McCarthy et al. 1988). For women on farms, increased participation most often takes the form of non-farm employment. Female employment has conventionally been highly concentrated in the service sector. The service sector incldes both a high-skill end (managerial and professional occpations) and a low-skill end (clerical, sales and service occpations). Historically, female employment has been concentrated in the latter, althogh the health and edcation infrastrctre in rral areas has been a major sorce of professional employment for rral women. Empirical evidence sggests that women consistently commte shorter distances to work than men (Deseran 1989; Hanson and Johnston 1985; Madden 1981), so the consolidation of pblic infrastrctre into more distant, larger centres may pose serios constraints on the opportnity for rral women to participate in this employment. Information on employment changes by indstry and by gender is readily available. Sorne empirical work on changes in the indstrial distribtion of employment also distingishes between metro and non-metro areas. However, an analysis of the differential effect of indstrial restrctring for females and males in metro and non-metro areas is largely absent. A detailed examination, based on commodity-by-indstry inpt-otpt tables for the Canadian provinces, describes the natre of recent strctral change in Canada and the importance of services in the regional growth process (Stabler and Howe 1992). The reslts spport the argments of a prodction process increasingly dependent on service inpts, and growing interregional and international trade in services. A U.K. stdy examines the regional conseqences of the shift from goodsprodcing jobs to service-type jobs (Hall 199). This stdy concldes that restrctring has aggravated regional disparities -- regions that lost the most in manfactring jobs gained the least in (private sector) service jobs. Growth indstries are ths becoming more concentrated in the largest rban centres. Similar patterns of the growth in the relative importance of the service indstries and their regional distribtion in the 198s are docmented for the U.S. (Bernat and Frederick 1992; Deavers 1991; Hady and Ross 199; Harnrick 1991/92; Porterfield 199). For most regions, in both non-metro and metro areas the service sector has been the major sorce of employment growth, bt non-metro areas remain more dependent on goods-prodcing indstries than metro areas. 1 Frther, the service sector jobs that have gone ta 1. The metro/non-metro distinction is somewhat arbitrary as non-metro areas will inclde residents srronding metro areas that are really more rban than rral. In the Prairie region of Canada, however, non-metro areas are tr1y rral.

136 OLFERT AND STABLER INDUSTRIAL RESTRUCTURING OF THE PRAIRIE LABOUR FORCE 137 non-metro areas tend to be non-basic rather than basic and are characterized by lower-paying employment. The impacts of the decline of the agricltre sector are a special case of the general pattern of goods-prodcing indstries. An additional aspect of the agricltre restrctring process appears to be the phenomer..on of permanent, planned plriactivity of farm family members (Barkley 199; Barlett 1986; Brooks and Reimnd 1989; Olfert 1992; Schmitt 1989). On the Canadian Prairies, farm women are more likely than farm men to participate in the nonfarm labor market. Participating in the non-farm labor market means competing in the same labor market as the rest of the labor force. Farm family members, however, are sometimes faced with an additional constraint in terms of commting to the location of the (desired) non-farm job. To the extent that the sorce of growing labor demand (service sector) is more likely to locate in metro areas, farm family members located in non-metro areas are at a competitive disadvantage. Althogh improvements in transportation infrastrctre will redce this disadvantage, time and cost brdens imposed by distance will still impede their participation. Several stdies have shown that the evoltion of the strctre of rral and rban centres is the reslt of economy-wide changes and the location choices offirms in the growth indstries (Beck and Herr 199; Freshwater et al. 1991; Stabler and Olfert 1992). These stdies point ot that as smaller, remote, or poorly located commnities decline, they become increasingly less able to provide employment opportnities and less likely to attract growth indstries. However, these stdies do not address the qestion of gender bias in the restrctring process. What is relatively weil docmented are the persistent differences in the indstrial distribtions of females and males (Statistics Canada 199). In Canada, 53 percent of females find employment in the service indstries compared with 25 percent of males. Women have not been weil represented in the goods-prodcing indstries, which are concentrated in nonmetro areas. The indstrial distribtion of female employment acconts for a large part of their persistently lower eamings. The literatre docments the growing importance of the service sector as a sorce of employment for both metro and non-metro areas, althogh metro areas are better able to attract high tech service indstries. While it is clear that females remain dependent on the growing service sector, empirical stdies do not docment the extent to which metro and non-metro females are reliant on this sector or the differential impact on females and males of the indstrial restrctring. Data Sorces and Methodology Indstrial distribtions ofe experienced labor force by gender for 1971, 1981, and 1986 provide the data base for this stdy (Statistics Canada 1988, 1989). Indstries are aggregated to nine grops which correspond to one or two-digit SIC levels.2 Unspecified indstries are deleted from consideration. The experienced labor force incldes nemployed and ths may not be a precise representation of the employed labor force, althogh given the relatively long time intervals over which change is being considered, this is not a serios problem. The metro/non-metro disaggregation is available for only 1981 and 1986, and the distinction between agricltre and other primary cannot be made. Metro areas are defined as the sm of Censs Metropolitan Areas (CMAs).3 The non-metro labor force was determined as a residal. An index of restrctring, which shows the changes in relative importance of each indstry to total employment, is sed to compare the degree to which metro and non-metro areas, and females and males, have participated in the restrctring process on the Canadian Prairies. The index is compted by dividing each indstry's share of employment change by that indstry's share of the employment base. 4 This index measres the extent to which the relative importance of individal indstries will change over time as the reslt of crrent changes in labor demand. Where aggregate employment is increasing, 2. A finer level of disaggregation wold lend more precision to the analysis at the expense of being able to sccinctly present the reslts. 3. Metro areas for each province are aggregated from the Censs Metropolitan Areas (CMAs) from Statistics Canada censs data. CMA is defined by Statistics Canada as comprising one or more censs sbdivisions (CSDs) that meet at least one of the following criteria: a. the CSD falls completely or partly inside the rbanized core (of at least 1, poplation); b. at least 5 percent of the employed labor force living in the CSD works in the rbanized core; or c. al least 25 percent of the employed labor force working in the CSD lives in the rbanized core. In Manitoba, Winnipeg is the only CMA; in Alberta, CMAs are Calgary and Edmonton; in Saskatchewan, Saskatoon and Regina. For this stdy, nonmetro areas are calclated as a residal by sbtracting the CMA total from the provincial total. 4. Indstry restrctring indices are compted for a particlar region as follows: (Empl.it - Empl.it_l) 1 (Total empl.i - Total empl.t_l) Empl.it_1 1 Total Empl.'_1 where i represents the indstry for which the index is compted and t the time period. As expressed here, t is the second or terminal year and t-i represents the initial point in time. Indstries for which the indices were calclaled are those nine sectors indicated in Figre 1. The nmerator is a ratio of employment growth (decline) in an indstry to total employment change in the region; in other words, the indstry's share of employment change between two points in time. The denominator is a ratio of total employment in the indslry in that region to total employment (ail indstries) in the region in the initial period, before the observed employment change, that is, the denominator shows the indstry's share of total employment in the base year.

138 OLFERT AND STABLER INDUSTRIAL RESTRUCTUR1NG OF THE PRAIRIE LABOUR FORCE 139 individal indstries may be affected positively (index vales greater than one) or negatively (vales less than one inclding negative vales). A vale of one serves as a benchmark. A vale greater than one implies the indstry is growing in relative importance. A vale of less than one (bt greater than zero) means the indstry is experiencing positive growth, bt is declining in relative importance. Negative nmbers signal an absolte decline in employment in that sector. 5 Restrctring indices are compted first for the entire Prairie region, then for metro and non-metro areas and, finally, for females and males. Indstrial Restrctring in Metro and Non-metro Areas Prairie-wide changes in the experienced labor force between 1971-81 and 1981-86 (Table 1) show that growth dring both periods was concentrated in the service sector, more so in the latter period than the former. 6 Between 1971 and 1986, the labor force grew by 878,35. Goods-prodcing indstries acconted for 14 percent while non-goods prodcing sectors acconted for 86 percent of this growth.7 Restrctring indices show the impact of absolte changes in employment on the relative importance of each sector in the regional economy (Figre 1). Between 1971 and 1981, trade, finance, insrance, and real estate (FIRE), other (non-agricltre) primary, and services all had restrctring indices greater than one. Constrction also had a vale in excess of one, althogh this is more a cyclical phenomenon than a long-term adjstment. Ali other sectors had indices of less than one, with only agricltre experiencing an absolte decline. Between 1981 and 1986, only services, pblic administration, and agricltre had restrctring indices of greater than one. Manfactring and constrction had negative indices, reflecting absolte declines in employment. The index for agricltre reflects the fact that there was a sbstantial increase in the nmber of women reported in the 1986 agricltre labor force. The large absolte size of the service-sector index reflects the importance of this indstry in providing new employment between 1981 and 1986. II shows that the service sector's share of new jobs was 3.31 times as great as that sector's share of 1981 base employment. 5. In the event that aggregate employment shold decline the interpretation of the signs of the index wold be exactly the opposite. 6. Growth and decline of the constrction indstry are more reflections of cyclical changes than longer-term strctral change. 7. Goods-prodcing indstries inclde agricltre, other primary (consisting of logging, forestry, fishing, trapping and mining); manfactring; and constrction. Non-goods-prodcing sectors inclde transportation, commnication and tilities (TCU); finance, insrance, and real estate (FIRE); commnity, bsiness, and personal services; and pblic administration. TABLE 1 Indstrial Distribtion of Labor Force Change by Sex, Prairie Region, 1971-81 and 1981-86 Labor Force Change Indstry 1971-81 1981-86 Males Females Tota1 3 Males Females Tota1 3 Agricltre -15,97-1,29-26,255 2,995 13,5 16,45 Other Primary 42,145 14,865 57,4 885 1,31 2,175 Manfactring 4,35 22,15 62,415-7,96-46 -8,425 Constrction 76,56 13,97 9,515-32,235-3,49-35,72 Transportation 4,51 21,8 62,3 1,38 3.35 4,415 Trade 61,98 74,965 136,96 9,35 6,535 15,88 F1RE 17,195 39,62 56,815 69 1,325 2,25 Services 83,5 16,37 243,875 45,36 9,35 135,39 Pblic Admin. 13,935 34,895 48,83 3,14 1,895 14,25 TOTAL 36,16 372,3 732,495 23,65 122,235 145,81 Sorce: Statistics Canada. 1988.1ndstry Trends, 1951-86. Cataloge No. 93-152. Ottawa: Minister of Spply and Services. a. Totals will not always eqal the sm of males and females de to ronding. TABLE 2 Indstrial Distribtion of Labor Force Change by Metro/Nonmetro and Sex, Prairie Region, 1981-86 Labor Force Change Indstry Metro Areas Nonmetro Areas Males Females Total 3 Males Females Total 3 Ali Primary 2,445 2,145 4,575 1,435 12,215 13,645 Manfactring -7,855-8 -8,65-15 34 Constrction -23,6-2,785-26,38-8,635-75 -9,34 Transportation -57 1,11 54 1,95 1,925 3,875 Trade 6,39 4,6 1,45 2,96 2,475 5,43 FIRE 59 28 865 1 1,45 1,16 Services 33,8 55,27 88,33 12,28 34,765 47,6 Pblic Admin. 3,115 7,165 1,28 25 3,73 3,745 TOTAL 13,595 66,445 8,1 1,1' 55,79 65,8 Sorce: Statistics Canada. 1989. Censs Metropolitan Areas. Cataloge No. 93-156. Ottawa: Minister of Spply and Services. a. Totals will not always eqal the sm of males pls females de to ronding. 225

14 OLFERT AND STABLER INDUSTRIAL RESTRUCTURING OF THE PRAIRIE LABOUR FORCE 141 4. l - 4. 3.lX) l 1. 2. x ' 2.lX) 1..5 CJ) :> g O.lX) Q<: ' P- c - i-- U r-j c- ni _1971-kl, f- o 19k1-k6 x '.5 CJ). :::> :> UX). Q<: ' -1. U,-+- --+ W litj --Jo- I_Morro,1 :--=o Nonmerro -l.lx) -2. f -2.lX) -3. -3. 2 3 c:..: ' E cl: <5 ' 'C: ü '' ::E - c. f ' fo- UJ li: ' E. Vl Ë..: '. : C. -4. ' 'C: 'C:.9 ' ü ê f li: c. :;;: '' ::E f i'j Ë Vl..: : c. FIGURE 1 Restrctring Indices by Indstry, Prairie Region, 1971-81 and 1981-86 The changes for the 1981-86 period (Table 2) reveal the importance of the service sector as a sorce of growth for both metro and non-metro areas, althogh in absolte terms non-metro growth was only half that of metro growth. Trade and pblic administration were the two other sectors that showed absolte growth over this period. 8 8. The apparent increase in primary sector employment is de to an increase in the nmber of women reporting agricltre as their indstry. This is more a fnction of a change in reporting conventions rather than a real increase. FIGURE 2 Restrctring Indices by Indstry, Prairie Region, Metropolitan and Non-rnetropolitan Areas, 1981-86 The restrctring indices shown in Figre 2 permit a comparison of the way in which the absolte growth and decline in individal indstries changes the indstrial strctre in metro and non-metro areas. For metro areas, sectors that are ndergoing a decline in relative importance are, in decreasing order, two sectors that experienced absolte redctions -- constrction (-3.56) and manfactring (-.92) -- and three sectors where employment increased bt their share of the total increase was smaller than their share of the base -- trade (.72), FIRE (.17), and transportation-commnication-tilities (.7).

142 OLFERT AND STABLER INDUSTRIAL RESTRUCTURING OF THE PRAIRIE LABOUR FORCE 143 In non-metro areas all sectors, except services, are declining in relative importance. Even so, the importance of this sector in non-metro areas is increasing less rapidly than it is in metro areas. In smmary, the faster-growing sectors in the Prairie economy are concentrated in metro areas and these sectors are, with few exceptions, growing faster than they are in non-metro areas. This implies that there is an increasing concentration of slower-growing or declining indstries in non-metro areas. 9 Indstrial Restrctring by Gender Jst as the changing composition of labor demand is not shared proportionately between metro and non-metro areas, female and male employment is affected differently by strctral change. This is becase their base indstrial distribtions differ and becase they participate in the changes in labor demand differently. Males constitted approximately 66 percent of the experienced labor force on the Prairies in 1971 bt this proportion had decreased to 58 percent by 1986. Over the lo-year interval from 1971 to 1981, 51 percent of the increase in the experienced labor force was acconted for by women, and over the following 5 years, the comparable figre was 84 percent. The indstrial distribtion of employment changes between 1971 and 1981 (Table 1) shows an increase for both genders across all indstries except agricltre. Females dominated increases in the service, trade, pblic administration, and FIRE sectors. The agricltre sector experienced an absolte decline, with the decrease in the nmber of males in this indstry somewhat greater than the decrease in nmber of females. Ali other sectors experienced growth, with the major share acconted for by males. The pattern of employment change by gender for the 1981-86 period, also shown in Table 1, is different than that observed for the earlier period. The dominance of the service sector is evident for both females and males, althogh the absolte increase for females is almost doble that for males. Females also dominated in the pblic administration, agricltre, and TeU indstries. Only in the trade sector is the increase in employment larger for males than for females, reversing the pattern of the previos decade. For the two declining indstries -- manfactring and constrction -- the loss of jobs by males otnmbered females by 9 ta 1. Restrctring indices constrcted individally for females and males indicate the changing strctre of indstrial employment for each gender. Althogh in absolte terms the service sector has been of paramont importance to the female labor force and females dominate this sector, a restrctring index of 9. The reslts presented here are based on employment changes. Focsing on vale of prodction may lead to a different presentation if prodctivity gains vary by sector. greater than one wold appear only if female employment is becoming even more concentrated in that sector over time. The index of restrctring by gender indicates which indstry is becoming relatively more important for females (males) as a reslt of captring a disproportionate share of employment growth. Indstrial restrctring of the female labor force between 1971-81 and 1981-86 is shown in Figre 3. For comparison, the vales of the restrctring indices for the indstry as a whole (males and females combined) are fond in parentheses nder the indstry names. For females, indstries with restrctring indices greater than one for 1971-81 were other primary, constrction, TeU, trade, FIRE, and pblic administration. This implies that the importance of these indstries to employment growth for females was greater than their representation in the 1971 labor force distribtion. The fact that services does not have an index vale of greater than one indicates that, althogh the service sector is the largest sorce of employment growth, the relative importance of this sector to the female labor force did not increase dring this period. In fact, its restrctring index of almost one indicates that the growth of female employment in the service sector was approximately eqal to this sector's share in the base employment. This wold be partly explained by the fact that women are participating to a greater degree in employment growth in other sectors as they make inroads (althogh small in absolte nmbers) into traditionally male-dominated indstries. The only sector with a restrctring index of sbstantially less than one for 1971-81 was agricltre. Restrctring indices for females may be compared with those for the entire labor force (males pls females), which appear below the table in parentheses for each indstry. For pblic administration, the restrctring index for the entire labor force for the 1971-81 period is less than one (.8), bt it is greater than one for women. TeU employment declined in relative importance as a whole bt increased in relative importance for females. On the other hand, the service sector, had an index greater than one for total employment bt slightly less than one for females. Trning to the 1981-86 period, restrctring indices for females show a different pattern than that of the preceding decade. The index for the service sector is greater than one, as are the indices for agricltre and pblic administration. Dring this period these sectors gained relative importance for the female labor force, while most other indstries declined in relative importance. Restrctring indices for the male labor force, shown in Figre 4, reveal other primary, constrction, trade, FIRE, and services gaining modestly in relative importance dring the 1971-81 period, while agricltre and pblic administration experienced smal1 relative declines. The fact that restrctring indices differed only modest1y from a vale of pls one indicates that the indstrial distribtion of male employment changes dring this decade closely approximated the indstry shares of male employment in the base period.

144 OLFERT AND STABLER INDUSTRIAL RESTRUCTURING OF THE PRAIRIE LABOUR FORCE 145..: on 'ç: Ü ' [ 1 3. r- 2.IX) f- f-- 1. f-4 C----' -. I-- -1. 1_1971-81 1981-86 U >-J L L -2. I 2 :;.:; :t E ;E.s 'c fi.e Iii ::; ':::J (R.I. 1971-81) (-.21) (2.3) (.84) (1.82) (tl.95) (1.16) (2.5) (1.32) (.8) (R.I. 1981-86) (1.11) (.3) (-.6) (-2.81) (.34) (.64) (.27) (3.31) (1.24) ;;; t c. ' ' ;;Z :: L ' <n r.; «, :ë,t! 12. 11. 1. 9. 8. 7 6. 5.!Xl 4. 3,IX)..: 2.m on 1. 'ç:. 1 1--- Ü -1. 1--- -2.tlO 1-- ' -3. -4JXl -5. -6.lX) -7. -8. -9. -IO.IX) ' -ll.lxi -12.()() 2 :; 'C «E ;E :: -5 2 fi.e Iii ::; 'c U LJ '2 1971-81J 1981 86 JI. n (R.I. 1971-81) (-.21) (2.3) (.84) (1.82) (.9S) (1.16) (2.S) (1.32) (.8) (R.I. 1981-86) (1.11) (.3) (-.6) (-2.81) (.34) (.64) (.27) (3.31) (1.24) ' c. fo el f :: Ë <n.5. «:ë FIGURE 3 Restrctring Indices of the Femaie Labor Force by Indstry, Prairie Region, 1971-81 and 1981-86 FIGURE 4 Restrctring Indices of the Male Labor Force by Indstry, Prairie Region, 1971-81 and 1981-86 For the 1981-86 period, the most striking featre is the increase in the importance of the service sector and the decline of most other sectors, especially constrction and manfactring. In addition to services, only trade and pblic administration had restrctring indices of greater than one. The implication is that the limited employment growth for males dring this period was highly concentrated in the service sector and that this concentration is a signifjcant departre from the base indstrial distribtion. Are females becoming more concentrated in the service sector? In sm- mary, yes, bt so is the entire labor force. In fact, this sector is increasing (relative to other sectors) at a slower rate for females than for males. For the 1981-86 period in particlar, the service sector restrctring indices for both females and males are greater than one, showing that this sector acconts for an increasing share of total employment for both. However, the absolte vale of the index for males is mch higher than for females. Females, whose jobs are already highly concentrated in the service sector, are fjnding new employment in all other sectors as weil. Male employment, on the other hand, is

146 OLFERT AND STABLER INDUSTRlAL RESTRUCTURING OF THE PRAIRIE LABOUR FORCE 147 expanding at a very slow rate bt what growth there is consists a1most exclsively of service-sector employment. In this way, males may be considered more dependent on the service sector for employment growth than females. Is the proportion of females in the service sector, the traditionally femaledominated indstry, growing? Again, yes, the female share of employment in the service sector is increasing. Over time, females will contine to accont for an increasing share of the service-sector employment, bt the same holds tre for female representation in most other sectors. The overall rate of growth of the female labor force is large relative to growth in the male labor force. That is, at the same time that the service sector is increasing in relative importance as a sorce of employment for females, females are increasing their share of total service-sector employment. For males, servïce-sector employment is increasing in relative importance as a sorce of employment bt their share of service-sector employment is decreasing. Indstrial Restrctring by Gender and Metro/Non-metro Areas Combined As demonstrated above, restrctring affects metro and non-metro areas differently, with metro areas captring a disproportionate share of employment growth in the non-goods-prodcing sectors. Declining or slower-growth indstries sch as the primary sector and manfactring are losing employment more slowly in non-metro than in metro areas. Over time this may not bode weil for non-metro areas. The implication is that total employment growth in non-metro areas will contine to lag behind that in metro areas, where employment in the growth indstries is crrently concentrated and is becoming more concentrated.. The change in employment by sector for non-metro females, metro females, non-metro males, and metro males is shown in Table 2. Metro areas captred more than 6 percent of the increase in female and abot 7 percent of the increase in male service-sector employment. Metro areas a1so dominated in trade-sector employment growth for both males and females. The concentration of pblic administration employment in metro areas is apparent for both males and females. Only in transportation and the primary sectors did nonmetro-area employment growth exceed that of metro areas. In metro areas, manfactring employment declined for both males and females; in non-metro areas, female employment in manfactring increased while male employment decreased, reslting in a small overall net increase. The comparison of non-metro and metro females shows that service sector employment growth was concentrated in metro areas. This was a1so the case for pblic administration and trade. Decreases in manfactring and constrc- tion-sector employment was more concentrated in metro areas, while increases in female employment in agricltre, other primary, TCU, and FIRE were greater in non-metro areas. The net reslt has been that, compared to metro females, the non-metro female labor force has not benefited proportionately from employment increases in sectors of growing labor demand. Non-metro males are less Iikely than their metro conterparts to find new employment in the service sector, pblic administration, trade, and FIRE. Only in the primary sectors and TCU are employment increases for males concentrated in non-metro areas. Constrction and manfactring decreased in both metro and non-metro areas, bt the loss of jobs was greater in metro areas. A comparison of indstry restrctring indices for metro and non-metro females dring the 1981-86 period (Figre 5) shows that the non-metro female labor force is becoming more concentrated (restrctring indices of greater than one) in the primary and service sectors, while the metro female labor force is becoming more concentrated in the service and pblic administration sectors. However, the restrctring index for non-metro females in the service sector is smailer than it is for metro females. Non-metro females were not able to captre as mch of the service-sector growth (relative to their base participation in this sector) as metro females. This is consistent with the more general observation made above that the servïce-sector growth is concentrated in metro areas. Frther, the pblic administration sector, which is a1so increasing in overall importance as a sorce of employment, benefits metro females bt not non-metro females. For those indstries with overall restrctring indices of less than one (everything except the service and pblic administration sectors), the vale of the indices is greater for non-metro females than for metro females. This sggests that downsizing in declining indstries is occrring less rapidly for non-metro females than for metro females, ths leaving a higher concentration of non-metro females employed in declining indstries. Restrctring indices for metro and non-metro males, shown in Figre 6, reveal a similar pattern. Compared to non-metro males, metro males are becoming more concentrated in indstries with restrctring indices greater than one. This is the case for both services and pblic administration. Trade, which is a1so gaining importance for the total male labor force, is a1so increasing in relative importance more rapidly for metro males than for non-metro males. Indstries that are generally declining in relative importance in the economy are losing importance less rapidly for non-metro males than for metro males. In smmary, metro females have participated to a greater extent than nonmetro females in the indstrial restrctring as shown by their larger indices for the growing indstries and smaller (negative) indices for the declining indstries. In the non-metro labor force, however, maies have been more affected by the indstrial restrctring than non-metro femaies.

148 OLFERT AND STABLER INDUSTRIAL RESTRUCTURING OF THE PRAIRIE LABOUR FORCE 149 >< '=' 2. 1.5 1-.-- fi-4- -+ LOO.----+- r+--.----+- r+-- Ur.5 - f-- 1--.= en t-. 1-- tî :::l -.5-1. -1.5-2. f - H_MWO f J o NonmerTO 'C '' 8.:: ;; ê a i::: ti: t 1 2 t:..:., :;;: :c ' ;j cs ê., 1- (.84) (-.6) (-2.82) (.34) (.64) (.27) (3.31) (1.24) 12. 11. 1. 9.1JO 8. 7. 6. 5. 4. 3. >< 2. '='.= 1. 1)..: 1. -2. -3. -4. -5.lXl -6. -7. -8. -9. -1. -11. -12. -13. _Metro,t o Nornetro = I J:lrr f--- f--- 1-- '----- - '' rj 'C.5 '':: ê ti: E <t il..: :;;: '.1 ;j :c 1 6': (.84) (-.6) (-2.82) (.34) (.64) (.27) (3.31) (1.24) FIGURE 5 Indstry Restrctring Indices, Fema1e Labor Force, Prairie Region, Metropolitan and Non-Metropolitan Areas, 1981-86 FIGURE 6 Indstry Restrctring Indices, Male Labor Force, Prairie Region, Metropolitan and Non-Metropolitan Areas, 1981-86 Conclsion One of the most visible indications of global restrctring is the changing indstrial distribtion of the labor force. The Prairie labor force has participated in the restrctring process. Goods-prodcing and resorce-extraction sectors, along with those indstries that directly spport them, have declined in relative importance and occasionally even in absolte terms. At the same time, service sectors have expanded rapidly in absolte tens and have greatly increased their relative importance as a sorce of employment. This trend toward increasing relative importance of the service sector was apparent dring the 1971-81 period, a period of general expansion, and more prononced dring the 1981-86 period when most other sectors grew slowly and sorne even declined. Restrctring indices in metro and non-metro areas show the growing relative importance of the service sector in both, althogh this sector is gaining importance more rapidly in metro than in non-metro areas. In general, inds

15 OLFERT AND STABLER INDUSTRIAL RESTRUCTURING OF THE PRAIRIE LABOUR FORCE 151 tries gaining relative importance on the Prairies as a whole are gaining more rapidly in metro areas. AIso, indstries declining in relative importance on the Prairies are declining more rapidly in metro areas. This indicates that metro areas are restrctring more rapidly. Metro areas are benefitting from an increasing concentration of growth indstries, while dec1ining indstries become more concentrated in non-metro areas. Separate restrctring indices for males and females reveal that, dring the 1971-81 period females gained in employment in sectors that have traditionally been male-dominated. At the same time, the service sector, while by far the largest sorce of emp1yment, did not gain in relative importance for the female labor force. The 1981-86 period showed services and pblic administration gaining relative importance in the female labor force while others experienced a relative decrease. Between 1971-81 the indstrial distribtion of the male labor force was very stable. In the following period, the service and pblic administration sectors gained significantly for both the male and female labor force. For males, however, net gains in employment were a1most exclsively in the service sector. Finally, indstrial restrctring for females and males in metro and nonmetro areas was examined. The pattern of mor gains by the service sector as a sorce of employment persists for ail for grops -- metro males, non-metro males, metro females, and non-metro females. Generally, growth sectors in the economy have favored metro females over non-metro females and metro males over non-metro males. The indstrial restrctring of the labor force -- with the non-goodsprodcing sectors gaining in relative importance and goods-prodcing sectors dec1ining -- is proceeding in non-metro as well as in metro areas. Non-metro areas are ths participating in economic restrctring, and the non-metro labor force, like the metro labor force, is being transformed into one that is more dependent on the service sectors. The rate of restrctring appears to be slower for non-metro areas. However, to pt a positive interpretation on this experience, dec1ining indstries are dec1ining less rapidly in non-metro than in metro areas. Even thogh an indstry is dec1ining in relative importance as a sorce of employment, positive absolte growth contribtes to income levels and poplation stability in non-metro areas. There may be a window of opportnity for non-metro areas with locational advantages for indstries that are down-sizing in metro areas. For non-metro areas, a sstained, althogh slow, rate of growth in indstries that are global1y declining in relative importance may be an indication of their comparative locational advantage. References Barkley, A.P. 199. The Determinants of the Migration of Labor Ot of Agricltre in the U.S., 194-1985, American Jornal of Agricltral Economies, 72: 567-73. Barlett, P. 1986. Part-time Farming: Saving the Farm or Saving the LifestYle?, Rral Society, 51: 289-313. Beek, R.J., and W. McD. Herr. 199. Employment Linkages from a Modified Shift-Share Analysis: An Illinois Example, The Review of Regional Stdies, 2: 38-45. Bernat, A. and M. Frederick. 1992. Goods-Prodcing Indstries Added More Rral than Urban Jobs in 1989, Rral Conditions and Trends, vol. 3. Rockville, Maryland: USDA, Economie Research Service. Brooks, N.L., and D.A. Reimnd. 1989. Where do Farm Hoseholds Eam their Incomes?. Washington, D.C.: USDA, Economie Research Service. Deavers, K.L. 1991. 198s, A Decade of Broad Rral Stress, Rral Development Perspectives, vol. 7. Rockville, Maryland: USDA, Economie Research Service. Deseran, F. A. 1989. Part-time Farming and Commting: Determinants of Distance to Off-farm Work for Loisiana Farm Coples, in W.W. Falk and T.A. Lyson (eds.), Research in Rral Sociology and Development: Rral Labor Markets. Greenwich, Conn.: JAl Press Inc. Freshwater, D., P. Ehrensaft, and L.P. Apedaile. 1991. The Implications of Global Restrctring of Primary Indstries on the Development of Rral and Single-Indstry Towns. Employment and Immigration Canada, Annal Rral Policy Seminar. Coaticook, Qebec: April 17-19, 1991. Godwin, D.D., and J. Marlowe. 199. Farm Wives' Labor Force Participation and Earnings, Rral Society, 55: 25-43. Hady, T.F., and P.J. Ross. 199. An Update: The Diverse Social and Economie Strctre of Nonmetropolitan America. Washington, D.C.: USDA, Economie Research Service. Hall, P. 199. Strctral Transformation of the Regions of the United Kingdom. University of California, Berkeley: Institte of Urban and Regional Development, Working Paper 57. Harnrick, K.S. 1991/92. Employment Mix Will Change by 25, Rral Conditions and Trends, vol 2. Rockville, Maryland: USDA, Economie Research Service. Hanson, S. and 1. Johnston. 1985. Gender Differences in Work-Trip Length: Explanations and Implications, Urban Geography, 6: 193-219. Madden, J.F. 1981. Why Women Work Closer to Home, Urban Stdies, 18: 181-194. McCarthy, M.R., P. Salant, and W.E. Sape. 1988. Off-Farm Employment by Married Farm Women: Research Review and New Evidence from Wisconsin, in W.G. Haney and J.B. Knowles (eds.), Women in Farming: Changing Roles, Changing Strctres. Bolder, Colorado: Westview Press. Olfert, M.R. 1992. Nonfarm Employment as a Response to Underemployment in Agricltre, Canadian Jornal of Agricltral Economies, 4:

152 ülfert AND STABLER 443-458. Porterfield, S. 199. Service Sector Offers More Jobs, Lower Pay, Rral Development Perspectives, vol. 6. Rockville, Maryland: USDA, Economie Research Service. Schmitt, G. 1989. Farms, Farm Hoseholds, and Prodctivity of Resorce Used in Agricltre, Eropean Review of Agricltral Economics, 16: 257-84. Stabler, J.e. and E.C. Howe. 1993. Services, Trade and Regional Strctral Change in Canada, 1974-84, Review of Urban and Regional Development Stdies, 5: 29-5. Stabler, J.C. and M.R. Olfert. 1992. Restrctring Rral Saskatchewan: The Challenge of the 199s. Regina, Saskatchewan: Canadian Plains Research Centre. Statistics Canada. 1988. Indstry Trends, 1951-86. 1951-1986 Censs of Canada, Cal. No. 93-152. Ottawa: Minister of Spply and Services.. 1989. Censs Metropolitan Areas. 1986 Censs of Canada, CaL No. 93-156. Ottawa: Minister of Spply and Services. _. 199. Women and the Labor Force. 1986 Censs of Canada, CaL No. 98-125. Ottawa: Minister of Spply and Services.