Technology, Job Stability and Inter-Industry Labor Mobility

Transcription

1 Technology, Job Stability and Inter-Industry Labor Mobility Sumati Srinivas Department of Economics Radford University Radford, VA Abstract The effects of technology in raising the skill premium in the labor market, while increasing wage inequalities since the 1980s have been well researched. In this study, we look at another aspect of the labor composition of technology-intensive industries by studying the labor inflows into them from less technology intensive industries. Using data from the NLSY79, we construct a new measure that quantifies the technology- gap bridged by workers moving across industries the technology gap index, or the TGI. Tracking the statistical characteristics of this gap over the period from 1986 to 2000, we analyze the relationship between job stability and technology gap for workers who move into higher tech industries. Thus, we hope to gauge both the openness of high tech industries to workers from low tech industries and the success rate of workers who make such a move in keeping their jobs. Introduction Technology has profoundly altered the United States labor market over the past few decades. Researchers studying the impact of technological change on worker 1

2 earnings and mobility have found new trends that has significant welfare implications for those seeking and changing jobs in the current economic environment. Firstly, technology-intensive firms 1 utilize a workforce that is more skilled (Doms, Dunne and Troske (1997)), and firms introduce new technology, they increase their demand for highly-educated workers (Bartel and Lichtenburg (1987)). This suggests that technology-intensive industries, the current engines of growth, are less likely to absorb workers from industries that are use less technology. Secondly, since the mid-1980s, the US economy has witnessed a rise in wage disparities between low- and high- income workers (Gottschalk (1997)). This has been linked to technological change and a rise in the price of highly-skilled workers by various researchers (Gottschalk (1997), Addison and Teixeira (2001)). Finally, occupational mobility research has revealed increasing signs of occupational stratification in the US labor market since the 1980s, with no sign of sustained upward occupational mobility of low-wage workers (Gabriel (2003)). A theoretical basis for such trends has been found in the human capital model, where technological acceleration decreases skill transferability, thus increasing both within-job wage growth and wage losses on job separation (Violante(2002)). Given that occupational flexibility and opportunity for upward mobility has long been a distinguishing feature of the US labor market, these trends clearly cause concern for the welfare implications of technological change for low-skilled workers. In this study, we look at the effects of technology on the labor market from a new perspective, extending and adding to previous research in the following significant ways. 1 We use the terms technological intensity and rate of technological change interchangeably throughout this paper. 2

3 Firstly, we construct a new technology gap index (TGI) that quantifies the difference in technological intensities between industries when a worker changes jobs. This index provides us with a new and unique insight into the technology gap that a worker has to overcome during a job change. Even a simple analysis of the number of observations of a positive TGI and other statistical characteristics of this variable during each year of our sample period should provide us with valuable information about what is happening to the ability of workers to bridge technology gaps across industries. Secondly, we look at the relation between the job retention rate of workers who have changed industries and relate it to the TGI of their industry change. The previous research on the impact of technology on job retention has been sparse, as most of the recent work has focused on wage effects. Zavodny(2000) looks at the relationship between technological intensity and job retention rates in the period and finds that the relationship varies according to the proxy used for technology. Aaronson (1999) uses data on displaced workers to find that technological intensity is not significantly associated with job displacement or reemployment after displacement. Our study pursues a new line of research in examining the population of job changers who move into higher technology jobs, and analyzing the stability of their new jobs. We take as a hypothesis that the indeterminate relationship between technology and job stability in the secular population demonstrated in such studies may obscure the instability of job changers moving into higher tech industries. Thus, our study attempts to provide an important piece of the puzzle of how technology affects job retention rates in general. Thirdly, we use new instruments to measure technological intensity in nonmanufacturing industries, which has not been possible until recently due to measurement 3

4 problems. Using these instruments, we are able to achieve a wide amount of industry coverage which gives us insights into the effects of technology on labor mobility outside of manufacturing, which has been the traditional realm for this type of research. Previous Research Technological Change A major focus of the previous research into technological change has been changes in the labor composition of industries that undergo significant technological change and the implications of these changes for less-skilled workers. Studies into this aspect of technological change have attempted to determine the wage effects of such change, and generally found it to increase the skill premium for highly skilled workers. Most such studies, however, have focused on the effects of technology on workers already in a given industry or firm. Less attention has been paid to the effects of technological change to intra-industry labor mobility, particularly with respect to observing the welfare effects on workers who move into jobs in industries that employ a higher level of technology. Another gap in the research into technological change has been the fact that it has almost entirely been restricted to the durable manufacturing industries. Indeed the most commonly used measure of technological change is the total factor productivity growth series calculated by Jorgensen, Gollop and Fraumeni (1987). This method of calculating technological change, though it has led to path-breaking research in this area, is restricted to twenty two 2-digit industry levels that are mostly manufacturing industries. However, many of the major employers of technology have been service industries. Indeed, in the 4

5 list of top ten users of Information Technology (IT) prepared by Triplett and Bosworth (2002), the first four are non-manufacturing industries. McGuckin and Stiroh (2000) show that three non-manufacturing industries use 76.6 percent of all computers in the US private sector. In such a situation, any analysis of technological change that is restricted to manufacturing must necessarily be incomplete. The reason for this focus on manufacturing can be traced measurement problems and lack of available instruments for measuring other industries. Grilliches (1994) details these data woes that make it difficult to quantify technological change in what he calls the unmeasurable sectors. This lack of reliable measures has forced most studies to be restricted to output measures of technology. However, there have been a few surveys in recent years that attempt to fill this gap, and provide reasonable (if partial) coverage of non-manufacturing industries. Labor Mobility Labor mobility in the US has been studied from various perspectives. The most common method of studying labor mobility is to look at employer-based job mobility, as in Davis, Haltiwanger and Schuh (1996). Another line of research focuses on the career of an individual across jobs. Sicherman and Galor (1990), for example, define a concept of upward or downward career mobility by focusing on various occupational aspects of a job. A closely related area of research is that of occupational mobility in which the aim, as in Gabriel (2003), is to study the occupational classifications of the job changes of an individual. Some of this research, such as Rytine (2000), also attempts to judge an individual s socioeconomic mobility during their careers. 5

6 The present study looks at labor mobility from a different perspective, borrowing from some of the approaches above. We are interested in finding the technological level of the industries that workers have been drawn to in recent years. There is a common belief that the so-called hi-tech industries have been main engines of job creation in the new economy, but there has been little research into their ability to absorb workers from industries with lower technological intensities. Our study attempts to breach this gap. In doing so, we also attempt to find out how workers fare after upgrading to a job at a hitech industry. Since the workers in our study are in their early to mid-careers, they could be giving up significant job tenure in changing industries, and our study tries to analyze the returns to this trade. Job Stability Previous research on the impact of technology on job retention has been sparse, as most of the recent work has focused on wage effects. Zavodny (2003) looks at the relationship between technological intensity and job retention rates in the period and finds that the relationship varies according to the proxy used for technology. Aaronson (1999) uses data on displaced workers to find that technological intensity is not significantly associated with job displacement or reemployment after displacement. Our study pursues a new line of research in examining the population of job changers who move into higher technology jobs, and analyzing the stability of their new jobs. We take as a hypothesis that the indeterminate relationship between technology and job stability in the secular population demonstrated in such studies may obscure the instability of job changers moving into higher tech industries. Thus, our study attempts to provide an important piece of the puzzle of how technology affects job retention rates in general. 6

7 Proxies for technology A number of technological proxies have been used by previous researchers. In addition to the extensively used total factor productivity growth (TFP) series calculated by Jorgensen, Gollop and Fraumeni (1987), other proxies used include patents obtained by the industry from the series calculated by Kortum and Putnam (1995) and the NBER TFP growth series calculated by Bartelsman and Gray (1996). Some studies such as that of Bartel and Sicherman (1999) use multiple proxies for technological change, in the hope that if the proxies show a similar result, it argues well for the robustness of the result. However, such agreement among technology proxies is more the exception than the rule, as can be expected from the fact that they capture different aspects of technology. In this study, we use two measures of technological change, both of which are input measures. The first is a US Department of Commerce survey of 58 industries which range across both the manufacturing and service sectors, that measure the IT Equipment share of total expenditures. This measure is similar to the Census of Manufacturing IT Equipment Share dataset which has sometimes been used in previous research, but it has the advantage of having wide coverage of non-manufacturing industries. The other proxy we use is the R&D to net sales ratio of R&D performing firms in an industry as calculated for the National Science Foundation. This is a more commonly used technology proxy, but the newer datasets from the NSF include limited amounts of non-manufacturing data which has not been exploited in previous research into technological mobility. These two measures explore different aspects of technological change since industries that do not perform much R&D may still be heavy 7

8 users of IT Equipment. On the other hand, industries that perform large amounts of R&D are likely to require highly skilled workers. By using these two input measures to complement each other, we hope to get a more complete picture of the effects of technological change on labor mobility. Data and Instruments The NLSY is an ongoing survey that uses a nationally representative sample of 12,686 young men and women who have been surveyed annually every year from 1979 to 1994, and once every two years from 1994 to The NLSY respondents were aged 14 to 22 at the start of the survey. So, they were aged 21 to 29 in 1986, making them excellent candidates for the purposes of our study, since they are in their early and midcareers during our study period, and this is when most occupational upgrading takes place. The NLSY provides a rich set of job-related information, including the industrial classification for each job the respondent holds at the time of the interview. We use the industrial classification for the current or most recent job reported by the respondent (the so-called CPS job) in our study. This classification is a 3-digit code based on the 1980 Census classification. 2 Since one of our interests is to study the secular increase or decline in labor mobility to higher tech industries, we sample the NLSY data biannually from 1984 to 2002, so that the number of industry changes between our samples remains comparable. Also, we account for possible misreporting of industry change by discarding industry 2 For the survey year 2002, the NLSY switches to using the 2000 Census classification codes. We have constructed a mapping between the 1980 and 200 codes for use in our analysis. Details on this mapping are available from the author on request. 8

9 change data for a respondent when there is no corresponding employer change. 3 The number of observations industry changes in our dataset was more than adequate for our purposes as we show in the analysis section, a relatively large number of the NLSY cohort changed industries during every year of our survey period. 4 The NLSY, unlike certain other panel datasets like the Panel Study of Income Dynamics (PSID) makes it possible to directly relate the current employer code to the employer code reported in the last survey. This makes it possible to construct a continuous employment record for each respondent and directly note job separations, instead of having to infer them indirectly. We measure job stability by assessing if the respondent retains the job that triggered the industry change after two years. 5 Observations are dropped for individuals who do not report job information for the next survey period after their industry change occurred. Only data from the main NLSY files are used because data in the supplemental and military samples are not available for our entire survey period. We use two instruments to measure the technological intensity in a given industry. The first is a survey of 58 industries carried out as part of the US Department of Commerce s report on Economy-Wide and Industry Level Impact of Information Technology (1997) that contains industry-level data on IT Equipment expenditure as a share of total capital expenditure. This dataset is especially useful for our purposes since it contains a set of industries specifically chosen to be inclusive of non-manufacturing 3 The NLSY collects employment data on up to 5 employers. We record a change of the CPS employer as a job change. 4 This large number of industry changes can probably be attributed to the young age of our sample population. 5 This is in accordance with measures of job stability used in many previous studies such as Bernhadt et al (1999) and Zavodny (2003). 9

10 industries. The IT expenditure is measured approximately at the 2-digit SIC level, though with many exceptions that allow us to map it more exactly into certain 3-digit Census industry codes. We obtain a mapping from the industries used by the survey into NLSY industry codes that cover 208 three-digit industries, a much finer disaggregation than has been used in previous studies of industry-level technological intensity. Our second instrument, the National Science Foundation s report on R&D to net sales ratio, has started collecting data for non-manufacturing industries since We make use of this dataset to get industry-level technological intensities for 145 three-digit Census Industry codes. 10

11 Empirical Model For use in our empirical model, we first calculate the technology gap index, TGIi, t2= Ii, t2 Ii, t1 (1) where i is used to index individuals, I denotes the technology intensity proxy for a given industry, t 2 is the survey year when the industry change occurred and t 1 is the previous survey year. 6 We then compute the normalized TGI, TGIi, t2 = TGI ' i, t2/ TG Imax, (2) where TGI max is the maximum recorded TGI across all our observations for the given technology proxy. By using this normalization procedure we obtain a range -1 TGI 1, which allows us to compare the TGI variable across proxies. We then estimate the probability that an individual will retain the same job two years after an industry change by using the following binary logit model: Pr ob( Yijt Xit, TGIj, Ct + 2) Logit ( Xit, TGIj, Ct + 2), = = (3) where the response variable Y denotes the individual retaining the same job, X is a vector of individual characteristics, TGI denotes the normalized technology gap that is calculated as detailed above, C is a vector of variables that control for the economic 6 For the years 1984 to 1994, when annual employment data is available, an industry change in the survey year between t1 and t2 is included only if the change is observed in year t2 as well. 11

12 climate, and Logit denotes the logistic function. The index i is used to index individuals, j is used to index industry changes and t is used to index time. The individual characteristics that we use include age, sex, race, job tenure (and its square), marital characteristics and possession of a college degree. All the job-related individual characteristics (such as tenure) are taken at time t (before industry change), and they may be endogenous with job separation. The unemployment rate in the individual s labor market is used to control for economic climate variables used to control for economic climate. This variable is used for the year t+2 since our interest is in obtaining its effect on the job separation, if it occurred. We then change our model specification in order to better study a specific question of interest. We split the industry changing population into higher- and lowertechnology movers in order to focus on the former. To do this, we replace the variable TGI in equation (3) with the dichotomous variable T, which is given by Tj = 1, iftgij > 0 Tj = 0, otherwise We now use equation (3) again to study the effect of changing to a industry with a positive technology gap on job stability. We run both the model specifications above using the two different proxies.throughout our analysis, we apply sampling weights provided by the NLSY at the beginning of each survey period 12

13 Preliminary Analysis and Results We recorded 9025 observations of industry change using the R&D/Sales technology proxy and observations using the IT Equipment Share proxy, both indicating a relatively high level of activity in our sample population. 7 An analysis of the descriptive statistics of our data shows some interesting findings. Table 1 shows the descriptive statistics of normalized TGI grouped by each of our survey years when measured using the IT Equipment proxy. The mean TGI for our entire population of industry changers seems to show a small but significant upward. This would be in line with expectations given the secular increase of technology usage across industries. More interestingly, though the numbers of industry changers are declining at a greater rate than the attrition rate of our sample, we find that the column Nh/N, which measures the proportion of industry changers moving to a higher tech industry has registered a steady increase over this period. Table 2 shows these same statistics measured using the R&D/Sales technology proxy. Strikingly similar trends are seen using this proxy, with the mean TGI and the proportion of industry changers moving to higher tech industries both registering increases during the 1990s, and both these measures reach their highest point at the end of our survey period. Graphs 1 and 2 map the data in tables 1 and 2 against each other allowing us to compare the two technology proxies. As can be seen, there is a significant level of agreement in the trends for both the mean and higher tech industry changers between the two proxies. 7 Throughout this section, it must be noted that the two proxies provide different levels of industry coverage. Thus, many observations of industry change that are made using one proxy may not covered with the other. This difference in coverage, taken in conjunction with the different qualities measured by the proxies, adds a measure of robustness to results when they converge. 13

14 TABLE 1: Descriptive Statistics Using IT Equipment Proxy % of Year workers Mean TGI Std Dev N moving to a (Normalized) higher tech industry Min Max TABLE 2 Descriptive Statistics Using R&D to Sales Proxy % of Year Workers Mean TGI Std Dev N moving to a (Normalized) higher tech industry Min Max

15 GRAPH 1: MEAN TGIs: 1986 to TGI Year Mean TGI_IT Mean TGI_RD GRAPH Year % of Workers moving to a higher tech industry (IT equip proxy) % of Workers moving to a higher tech industry (R&D proxy) 15

16 TABLE 3: Maximum Likelihood Analysis for Job Stability using IT Equipment Share as Proxy. 8 Variable Estimate Error Intercept TGI Age Age^ Female Black Tenure Tenure^ E-06 Married Divorced College TABLE 4: IT Equipment Share Split Level Maximum Likelihood Analysis Standard Variable Estimate Error Intercept Industry change to higher technology Age Age^2/ Female Black In tables 3 to 6, it should be noted that only a selection of interesting covariates are shown. Other variables used in the regression include year dummies for each survey year, unemployment rates, high school attainment and graduate degree attainment. 16

17 Tenure Tenure^ E-06 Married Divorced College TABLE 5: Maximum Likelihood Analysis for Job Stability using R&D Share as Proxy. Variable Estimate Error Intercept TGI Age Age^2/ Female Black Tenure Tenure^ E-06 Married Divorced College TABLE 6: R&D to Sales Split Level Maximum Likelihood Analysis Standard Variable Estimate Error Intercept Industry change to higher tech industry Age Age^2/ Tenure

18 Black Female Married College Degree Tables 3 to 6 show some of the results of the logistic regression with our two model specifications and two technology proxies. With the IT Equipment proxy, the technology gap index turns out to be significant and has a relatively strong positive association with job stability. This result holds true with the split population as well with industry changers moving to higher technology jobs being more likely to have job stability. However, the TGI is not significant when regressed with the R&D to Sales technology proxy. Again, a similar result holds with the split population for this proxy. Other significant variables in all four tables include job tenure, age, and being married which are positively associated with stability, and being female which is negatively associated These results for these other variables are we would expect from theory and previous studies. Conclusions and Future Work The rise in mean TGI and increase in the proportion of industry changers with positive TGI using both our proxies point towards companies that invest in technology being increasingly targeted by industry changers. This augurs well for the index we have created in this paper, TGI, being a variable of interest to explore various labor market phenomena. The positive association of TGI with job stability using the IT Equipment proxy and its insignificance when using the R&D to Sales proxy seems to indicate that the increase in welfare effects such as job stability when an industry changer bridges a 18

19 technology gap has more to do with the level of IT expenditure in an industry than the amount of research performed in it. This in turn has a significant implication for many service industries, which are among the heaviest IT users, but may not have a proportional investment in R&D. Future work can involve further exploratory work using the TGI to study wage effects for industry changers who move to higher technology industries. Another line of research could look at the top stratum of technology using industries and study the technology gap bridged by workers who move to such industries and their subsequent welfare effects. Further work also needs to be done in exploring the demographics of industry changers with high TGIs. 19

20 Appendix 1 Industries Used for Measuring IT Equipment Share Census Industry Code Description (SIC code) 40 Metal mining (10) 41 Coal mining (11,12) 42 Crude petroleum and natural gas extraction (13) 50 Nonmetallic mining and quarrying, except fuel (14) 60 Construction (15, 16, 17) 100 Meat products (201) 101 Dairy products (202) 102 Canned and preserved fruits and vegetables (203) 110 Grain mill products (204) 111 Bakery products (205) 112 Sugar and confectionery products (206) 120 Beverage industries (208) 121 Miscellaneous food preparations & kindred products (207, 209) 122 Not specified food industries 130 Tobacco manufactures (21) 132 Knitting mills (225) 140 Dyeing and finishing textiles, except wool and knit goods (226) 141 Floor coverings, except hard surface (227) 142 Yarn, thread, and fabric mills (228, ) 150 Miscellaneous textile mill products (229) 151 Apparel and accessories, except knit ( ) 152 Miscellaneous fabricated textile products (239) 160 Pulp, paper, and paperboard mills ( , 266) 161 Miscellaneous paper and pulp products (264) 162 Paperboard containers and boxes (265) 171 Newspaper publishing and printing (271) 172 Printing, publishing, and allied industries, except newspapers ( ) 180 Plastics, synthetics, and resins (282) 181 Drugs (283) 182 Soaps and cosmetics (284) 190 Paints, varnishes, and related products (285) 191 Agricultural chemicals (287) 192 Industrial and miscellaneous chemicals (281, 286, 289) 200 Petroleum refining (291) 201 Miscellaneous petroleum and coal products (295, 299) 210 Tires and inner tubes (301) 211 Other rubber products, and plastics footwear and belting ( , 306) 212 Miscellaneous plastics products (307) 220 Leather tanning and finishing (311) 221 Footwear, except rubber and plastic (313, 314) 222 Leather products, except footwear ( , 319) 20

21 230 Logging 231 Sawmills, planing mills, and millwork (242,243) 232 Wood buildings and mobile homes (245) 241 Miscellaneous wood products (244,249) 242 Furniture and fixtures (25) 250 Glass and glass products ( ) 251 Cement, concrete, gypsum, and plaster products (324, 327) 252 Structural clay products (325) 261 Pottery and related products (326) 262 Miscellaneous nonmetallic mineral and stone products (328, 329) 270 Blast furnaces, steelworks, rolling and finishing mills (331) 271 Iron and steel foundries (332) 272 Primary aluminum industries (3334, pt 334, , 3361) 280 Other primary metal industries 281 Cutlery, hand tools, and other hardware (342) 282 Fabricated structural metal products (344) 290 Screw machine products (345) 291 Metal forgings and stampings (346) 292 Ordnance (348) 300 Miscellaneous fabricated metal products (341, 343, 347, 349) 301 Not specified metal industries 310 Engines and turbines (351) 311 Farm machinery and equipment (352) 312 Construction and material handling machines (353) 320 Metalworking machinery (354) 321 Office and accounting machines (357, except 3573) 322 Electronic computing equipment (3573) 331 Machinery, except electrical, n.e.c. (355, 356, 358, 359) 332 Not specified machinery 340 Household appliances (363) 341 Radio, TV, and communication equipment (365, 366) 342 Electrical machinery, equipment, and supplies, n.e.c. (361, 362, 364, 367, 369) 350 Not specified electrical machinery, equipment, and supplies 351 Motor vehicles and motor vehicle equipment (371) 352 Aircraft and parts (372) 360 Ship and boat building and repairing (373) 361 Railroad locomotives and equipment (374) 362 Guided missiles, space vehicles, and parts (376) 370 Cycles and miscellaneous transportation equipment (375, 379) 371 Scientific and controlling instruments (381, 382) 372 Optical and health services supplies (383, 384, 385) 380 Photographic equipment and supplies (386) 381 Watches, clocks, and clock work operated devices (387) 382 Not specified professional equipment 390 Toys, amusement, and sporting goods (394) 391 Miscellaneous manufacturing industries (39, exc. 394) 392 Not specified manufacturing industries 400 Railroads (40) 401 Bus service and urban transit (41, except 412) 402 Taxicab service (412) 21

22 410 Trucking service (421,423) 411 Warehousing and storage (422) 412 U.S. Postal Service (43) 420 Water transportation (44) 421 Air transportation (45) 422 Pipe lines, except natural gas (46) 432 Services incidental to transportation (47) 440 Radio and television broadcasting (483) 441 Telephone (wire and radio) (481) 442 Telegraph and miscellaneous communication services (482, 489) 460 Electric light and power (491) 461 Gas and steam supply systems (492, 496) 462 Electric and gas, and other combinations (493) 470 Water supply and irrigation (494, 497) 471 Sanitary services (495) 472 Not specified utilities 500 Motor vehicles and equipment (501) 501 Furniture and home furnishings (502) 502 Lumber and construction materials (503) 510 Sporting goods, toys, and hobby goods (504) 511 Metals and minerals, except petroleum (505) 512 Electrical goods (506) 521 Hardware, plumbing and heating supplies (507) 522 Not specified electrical and hardware products 530 Machinery, equipment, and supplies (508) 531 Scrap and waste materials (5093) 532 Miscellaneous wholesale, durable goods (5094, 5099) 540 Paper and paper products (511) 541 Drugs, chemicals, and allied products (512, 516) 542 Apparel, fabrics, and notions (513) 550 Groceries and related products (514) 551 Farm-product raw materials (515) 552 Petroleum products (517) 560 Alcoholic beverages (518) 561 Farm supplies (5191) 562 Miscellaneous wholesale, nondurable goods (5194, 5198, 5199) 571 Not specified wholesale trade 580 Lumber and building material retailing (521, 523) 581 Hardware stores (525) 582 Retail nurseries and garden stores (526) 590 Mobile home dealers (527) 591 Department stores (531) 592 Variety stores (533) 600 Miscellaneous general merchandise stores (539) 601 Grocery stores (541) 602 Dairy products stores (545) 610 Retail bakeries (546) 611 Food stores, n.e.c.(542, 543, 544, 549) 612 Motor vehicle dealers (551, 552) 620 Auto and home supply stores (553) 22

23 621 Gasoline service stations (554) 622 Miscellaneous vehicle dealers (555,556,557,559) 630 Apparel and accessory stores, except shoe (56, except 566) 631 Shoe stores (566) 632 Furniture and home furnishings stores (571) 640 Household appliances, TV, and radio stores (572, 573) 641 Eating and drinking places (58) 642 Drug stores (591) 650 Liquor stores (592) 651 Sporting goods, bicycles, and hobby stores (5941, 5945, 5946) 652 Book and stationery stores (5942, 5943) 660 Jewelry stores (5944) 661 Sewing, needlework, and piece goods stores (5949) 662 Mail order houses (5961) 670 Vending machine operators (5962) 671 Direct selling establishments (5963) 672 Fuel and ice dealers (598) 681 Retail florists (5992) 682 Miscellaneous retail stores (593, 5947, 5948, 5993, 5994, 5999) 691 Not specified retail trade 700 Banking (60) 701 Savings and loan associations (612) 702 Credit agencies, n.e.c. (61, except 612) 710 Security, commodity broker age, and investment companies (62, 67) 711 Insurance (63, 64) 712 Real estate, including real estate-insurance-law offices (65,66) 721 Advertising (731) 722 Services to dwellings and other buildings (734) 730 Commercial research, development, and testing labs (7391, 7397) 731 Personnel supply services (736) 732 Business management and consulting services (7392) 740 Computer and data processing services (737) 741 Detective and protective services (7393) 742 Business services, n.e.c. (732, 733, 735, 7394, 7395, 7396, 7399) 750 Automotive services, except repair (751, 752, 754) 751 Automotive repair shops (753) 752 Electrical repair shops (762, 7694) 760 Miscellaneous repair services (763, 764, 7692, 7699) 762 Hotels and motels (701) 770 Lodging places, except hotels and motels (702, 703, 704) 771 Laundry, cleaning, and garment services (721) 772 Beauty shops (723) 780 Barber shops (724) 781 Funeral service and crematories (726) 782 Shoe repair shops (725) 790 Dressmaking shops (pt 729) 791 Miscellaneous personal services (722, pt 729) 800 Theaters and motion pictures (78, 792) 801 Bowling alleys, billiard and pool parlors (793) 802 Miscellaneous entertainment and recreation services (791, 794, 799) 23

24 812 Offices of physicians (801, 803) 820 Offices of dentists (802) 821 Offices of chiropractors (8041) 822 Offices of optometrists (8042) 830 Offices of health practitioners, n.e.c. (8049) 831 Hospitals (806) 832 Nursing and personal care facilities (805) 840 Health services, n.e.c. (807, 808, 809) 841 Legal services (81) 842 Elementary and secondary schools (821) 850 Colleges and universities (822) 851 Business, trade, and vocational schools (824) 852 Libraries (823) 860 Educational services, n.e.c. (829) 861 Job training and vocational rehabilitation services (833) 862 Child day care services (835) 870 Residential care facilities, without nursing (836) 871 Social services, n.e.c. (832, 839) 24

25 Appendix 2 Industries Used for Measuring R&D to Sales Ratio Census Industry Code Industry Description 121 Miscellaneous food preparations & kindred products (207, 209) 122 Not specified food industries 130 Tobacco manufactures (21) 132 Knitting mills (225) 140 Dyeing and finishing textiles, except wool and knit goods (226) 141 Floor coverings, except hard surface (227) 142 Yarn, thread, and fabric mills (228, ) 150 Miscellaneous textile mill products (229) 151 Apparel and accessories, except knit ( ) 152 Miscellaneous fabricated textile products (239) 160 Pulp, paper, and paperboard mills ( , 266) 161 Miscellaneous paper and pulp products (264) 162 Paperboard containers and boxes (265) 171 Newspaper publishing and printing (271) 172 Printing, publishing, and allied industries, except newspapers ( ) 180 Plastics, synthetics, and resins (282) 181 Drugs (283) 182 Soaps and cosmetics (284) 190 Paints, varnishes, and related products (285) 191 Agricultural chemicals (287) 192 Industrial and miscellaneous chemicals (281, 286, 289) 200 Petroleum refining (291) 201 Miscellaneous petroleum and coal products (295, 299) 210 Tires and inner tubes (301) 211 Other rubber products, and plastics footwear and belting ( , 306) 212 Miscellaneous plastics products (307) 220 Leather tanning and finishing (311) 221 Footwear, except rubber and plastic (313, 314) 222 Leather products, except footwear ( , 319) 230 Logging 231 Sawmills, planing mills, and millwork (242,243) 232 Wood buildings and mobile homes (245) 241 Miscellaneous wood products (244,249) 242 Furniture and fixtures (25) 250 Glass and glass products ( ) 251 Cement, concrete, gypsum, and plaster products (324, 327) 252 Structural clay products (325) 261 Pottery and related products (326) 262 Miscellaneous nonmetallic mineral and stone products (328, 329) 270 Blast furnaces, steelworks, rolling and finishing mills (331) 271 Iron and steel foundries (332) 25

26 272 Primary aluminum industries (3334, pt 334, , 3361) 280 Other primary metal industries 281 Cutlery, hand tools, and other hardware (342) 282 Fabricated structural metal products (344) 290 Screw machine products (345) 291 Metal forgings and stampings (346) 292 Ordnance (348) 300 Miscellaneous fabricated metal products (341, 343, 347, 349) 301 Not specified metal industries 310 Engines and turbines (351) 311 Farm machinery and equipment (352) 312 Construction and material handling machines (353) 320 Metalworking machinery (354) 321 Office and accounting machines (357, except 3573) 322 Electronic computing equipment (3573) 331 Machinery, except electrical, n.e.c. (355, 356, 358, 359) 332 Not specified machinery 340 Household appliances (363) 341 Radio, TV, and communication equipment (365, 366) 342 Electrical machinery, equipment, and supplies, n.e.c. (361, 362, 364, 367, 369) 350 Not specified electrical machinery, equipment, and supplies 351 Motor vehicles and motor vehicle equipment (371) 352 Aircraft and parts (372) 360 Ship and boat building and repairing (373) 361 Railroad locomotives and equipment (374) 362 Guided missiles, space vehicles, and parts (376) 370 Cycles and miscellaneous transportation equipment (375, 379) 371 Scientific and controlling instruments (381, 382) 372 Optical and health services supplies (383, 384, 385) 380 Photographic equipment and supplies (386) 381 Watches, clocks, and clock work operated devices (387) 382 Not specified professional equipment 390 Toys, amusement, and sporting goods (394) 391 Miscellaneous manufacturing industries (39, exc. 394) 392 Not specified manufacturing industries 400 Railroads (40) 401 Bus service and urban transit (41, except 412) 402 Taxicab service (412) 410 Trucking service (421,423) 411 Warehousing and storage (422) 412 U.S. Postal Service (43) 420 Water transportation (44) 421 Air transportation (45) 422 Pipe lines, except natural gas (46) 432 Services incidental to transportation (47) 440 Radio and television broadcasting (483) 441 Telephone (wire and radio) (481) 442 Telegraph and miscellaneous communication services (482, 489) 460 Electric light and power (491) 461 Gas and steam supply systems (492, 496) 26

27 462 Electric and gas, and other combinations (493) 470 Water supply and irrigation (494, 497) 471 Sanitary services (495) 472 Not specified utilities 700 Banking (60) 701 Savings and loan associations (612) 702 Credit agencies, n.e.c. (61, except 612) 710 Security, commodity broker age, and investment companies (62, 67) 711 Insurance (63, 64) 721 Advertising (731) 722 Services to dwellings and other buildings (734) 730 Commercial research, development, and testing labs (7391, 7397) 731 Personnel supply services (736) 732 Business management and consulting services (7392) 740 Computer and data processing services (737) 741 Detective and protective services (7393) 742 Business services, n.e.c. (732, 733, 735, 7394, 7395, 7396, 7399) 750 Automotive services, except repair (751, 752, 754) 751 Automotive repair shops (753) 752 Electrical repair shops (762, 7694) 760 Miscellaneous repair services (763, 764, 7692, 7699) 761 Private households (88) 762 Hotels and motels (701) 770 Lodging places, except hotels and motels (702, 703, 704) 771 Laundry, cleaning, and garment services (721) 772 Beauty shops (723) 780 Barber shops (724) 781 Funeral service and crematories (726) 782 Shoe repair shops (725) 790 Dressmaking shops (pt 729) 791 Miscellaneous personal services (722, pt 729) 800 Theaters and motion pictures (78, 792) 801 Bowling alleys, billiard and pool parlors (793) 802 Miscellaneous entertainment and recreation services (791, 794, 799) 812 Offices of physicians (801, 803) 820 Offices of dentists (802) 821 Offices of chiropractors (8041) 822 Offices of optometrists (8042) 830 Offices of health practitioners, n.e.c. (8049) 831 Hospitals (806) 832 Nursing and personal care facilities (805) 840 Health services, n.e.c. (807, 808, 809) 841 Legal services (81) 842 Elementary and secondary schools (821) 850 Colleges and universities (822) 851 Business, trade, and vocational schools (824) 852 Libraries (823) 860 Educational services, n.e.c. (829) 861 Job training and vocational rehabilitation services (833) 862 Child day care services (835) 27

28 870 Residential care facilities, without nursing (836) 882 Engineering, architectural, and surveying services (891) 890 Accounting, auditing, and bookkeeping services (893) 891 Noncommercial educational and scientific research (892) 892 Miscellaneous professional and related services (899) 28

29 References Addison, John T. and Paulino Teixeira Technology, Employment and Wages. Review of Labour Economics and Industrial Relations. 15. Autor, D.H, Katz and A.B.Krueger Computing Inequality: Have Computers Changed the Labor Market? Quarterly Journal of Economics. 113(4). Doms, Mark, Timothy Dunne and Kenneth R. Troske "Workers, Wages, and Technology." Quarterly Journal of Economics Bartel, Ann P. and Frank R. Lichtenberg "The Comparative Advantage of Educated Workers in Implementing New Technology." The Review of Economics and Statistics. LXIX Bartel, Ann P. and Nachum Sicherman Technological Change and Wages: An Inter-Industry Analysis. Cambridge, MA: National Bureau of Economic Research Inc. Working Paper No Bartelsman, E. J. and W. Gray The NBER Manufacturing Productivity Database. National Bureau of Economic Research Technical Working Papers Gabriel, Paul E An Examination of Occupational Mobility among Full-time Workers. Monthly Labor Review. September. Gordon, Robert J Does the "New Economy" Measure up to the Great Inventions of the Past? The Journal of Economic Perspectives. 14:4, pp Gottschalk, Peter Inequality, Income Growth, and Mobility: The Basic Facts. The Journal of Economic Perspectives. 11:2, pp Griliches, Zvi Productivity, R&D and the Data Constraint. American Economic Review. 84, pp

30 Jorgenson, Dale W. and Kevin J. Stiroh U.S. Economic Growth at the Industry level. American Economic Review. 90:2 pp Jorgenson, Dale W. and Kevin J. Stiroh Information Technology and Growth. American Economic Review. 89:2, pp Jorgenson, Gollop and Fraumeni Productivity and US Economic Growth. Cambridge, Mass. Harvard University Press. Kortum, Samuel S Research, Patenting, and Technological Change. Econometrica. 65:6, pp Kortum Samuel S. and Jonathan Putnam Predicting Patents by Industry: Tests of the Yale-Canada Concordance. Manuscript. Boston. Boston University. Lipsey, R.G The Productivity Paradox: A Case of the Emperor's New Clothes. ECWP. no.158, Toronto: CIAR. McGuckin, Robert H and Kevin J. Stiroh Do Computers Make Output Harder to Measure? Journal of Technology Transfer. Rytina, Steven Is Occupational Mobility Declining in the U.S.? Social Forces 78:3, pp Sicherman, Nachum and Oded Galor A Theory of Career Mobility. The Journal of Political Economy. 98:1, pp Triplett, Jack E. and Barry P. Bosworth Baumol s Disease has been Cured: IT and Multifactor Productivity in U.S. Services Industries. Texas A&M Conference. Violante, Giovanni L Technological Acceleration, Skill Transferability and the Rise in Residual Inequality. Quarterly Journal of Economics. 117:1, pp

31 Zavodny, Madeline Technology and Job Separation among Young Adults, Economic Inquiry. 41:2, pp