Chapter 1
The Limits of Private Electrical Modernization, 1919–1929

The Perspective of Enterprise

In the 1920s, American electrical utilities completed the final stage of their development. The commercial central station industry had evolved from small, local distribution systems at the turn of the century to regional transmission monopolies by World War I. Holding companies then consolidated regional utilities in the decade before the Great Depression. By 1929, five holding companies controlled 80 percent of the electrical generating capacity of the nation. Electrical operating utilities grew dramatically during the 1920s. The amount of electricity generated by central station utilities rose 228 percent, from 39,519 million kilowatts in 1920 to 90,076 million kilowatts in 1930. Revenues from sales of this electricity grew by 244 percent, from $882 million in 1920 to $2,155 million in 1930. Since the vast majority of ultimate consumers of electricity were households and since domestic consumption of electricity returned a significant percentage of utilities' revenue, observers assumed that utilities as an industry eagerly mass marketed home appliances during the decade to increase domestic consumption of electricity. This was not the case.^[1]

Most utilities lacked enthusiasm for the domestic market, because their bottom line showed little profit from it. The industry in fact subsidized much household service—because governments required them to do so. As late as 1930, eight million households, or 40 percent of the nation's private utility domestic customers, did not consume enough electricity to be profitable. The utilities subsidized this service by maintaining higher flat rates to relatively high value domestic users.^[2]

The utilities naturally wished to invest their capital in the most profitable projects. In the 1920s, the sale of electricity to industry using electric power motors brought the greatest profit. Between 1923 and 1929, the percentage of total electricity distributed in the United States taken by manufacturers rose from 48.2 to 52.9 percent. Central station electrification of industry brought high returns in two ways. First, many industrial plants and large stores had their own electrical generators and distribution systems. Utilities needed only to convince their management that they could buy central station energy more cheaply than they could generate their own power. The utility then could feed electricity to the industrial plant's electricity system, in some cases requiring installation of couplers to transform AC to DC current. The utilities' industrial campaign met remarkable success. After 1923 the aggregate power generated in the United States by isolated industrial power plants declined. The second reason for the high profit from industrial sales came from the geographical concentration of industrial plant energy consumption. Industrial motors used large amounts of electricity when measured by standard criteria, such as metered connections per mile of distribution line. Other factors also affected the relative profitability of the industrial and the domestic sectors of the business. For instance, domestic consumption's higher ratio of distribution mileage to meters meant greater losses of electricity from distribution lines. More meters brought higher overhead costs for service and management. An analysis conducted in 1923 of residential service in the Middle West revealed that the carrying cost of homes averaged $2 to $2.50 a month, at a time when the average revenue from homes was less than that.^[3]

In 1929, domestic customers constituted 82.7 percent of users of central station energy but returned only 29.4 percent of total revenues. Large and small industries, by contrast, constituted 17.3 percent of customers and returned 55.5 percent of total revenues. The 1920s saw tremendous growth in household electrification, but the percentage of total customers comprised by households in 1929 changed little from 1922. Managers of the electric utilities understood their profit situation. As the utilities pulled out of the postwar depression (1920–1921), they looked to industrial electrification and electrical modernization of capital equipment for the profits they needed to attract financing for the increased generating capacity they intended to install. In an editorial of January 1923, Electrical World, the major management journal of the industry, pointed to the fact that only one-third of machine horsepower in industry came from electricity. The editors recommended conversion of the remaining two-thirds of industrial horsepower as the outstanding opportunity for the utilities. Projections of future electrical loads pointed to greater increases in industrial power loads than domestic lighting loads. In a set of ten-year projections made in 1924, industrial power clearly ranked as the planning priority, though predicted increases in domestic power sales warranted notice.^[4]

An array of problems blocked increased sale of electricity to homes for the utilities. In the early 1920s, utility marketing analysts portrayed American homes

as two distinct markets. Home illumination comprised one market for electricity; the other market called for full electrical service. Under public regulation to extend service broadly, utilities had to make capital investment to electrify new houses and to retrofit nonelectrified housing stock with wiring for illumination. Utilities thought households might install one or at most a few small electric appliances on house light lines. Managers drew back, however, from investing in electrifying dwellings for heavy appliances, because heavier appliances required heavier wiring and different neighborhood transformers, with no assurance that the investment would produce significantly greater consumption of electricity. Most utility managers simply did not believe, through the 1920s, that the home heavy appliance market would be sufficiently profitable for them to divert massive capital to it. A marketing study of 584 residential customers conducted by the Philadelphia Electric Company in 1926 demonstrated that the average light customer worked within a rigid budget for electrical service. Though households might be willing to buy appliances up to their budget limits, they resisted raising the level of budgeted expenditures for electrical service. Electrical demand was not elastic: "It would appear that considerable resistance exists to an increase in the annual expenditure even in the face of continual educational effort as to the advantages and economic benefits of a more liberal use of electric service." On the basis of analysis of federal income tax data and utility marketing, another analyst concluded flatly, "The actual demand for electric service will always be a function of the amount of the family budget that can be spent for these domestic requirements."^[5]

As a result of the segmentation of the residential market, most utilities refused to build electrical load in the illumination sector that represented four-fifths of their customers. They focused their marketing efforts, instead, on increasing consumption by the few upper-income households who could pay for full electric service. This class represented a large market, no doubt, but not a mass market. A report by Electrical World in 1927 testified to this widespread strategy. An industry analyst surveyed 376 large utilities serving an aggregate population of 54,500,000 persons. The analysis demonstrated—in his opinion—that only 28 percent of the utilities promoted "full domestic service," that is, appliances as well as lights, across all their markets. They ignored lower-income households earning $2,000 to $3,000 a year. A review of 1928 revenues from fifty-seven operating utilities led Electrical World to conclude that only 10 to 20 percent of these utilities customers were "prospects for complete electric service at indicated competitive rates." The experience of the few utilities that did attempt to market full service indicated that average customers responded to lowered appliance prices and electric rates with increased demand; nonetheless, the overwhelming majority of utilities did not lower prices, refusing to believe that those households had sufficient income to make them worth the investment. Throughout the 1920s, the industry press was filled with self-confirming studies that middle- and lower-income households were not good markets for increased electricity consumption.^[6]

For all the claims utilities made in the 1920s of marketing prowess and of the necessity for encouraging diversity of electric demand (for purposes of "load management"), the evidence in their own trade literature testifies that they had little understanding of domestic consumers and marketing. They pigeonholed households in a few large-scale categories. They expressed little appreciation for the individuality of consumer tastes and buying habits. They had little concrete information about how households actually worked inside the home or made buying decisions. Most utilities lacked appreciation for the research on rural, farm, and urban households by women social scientists of the U.S. Department of Agriculture in the 1920s and for the whole range of household management literature by housekeeping reformers. Their ignorance of the home market was not unusual. Telephone companies, for instance, could not initially imagine that households might want to use phones simply for social conversation. As the study by C. F. Lacombe revealed in 1927, most utilities had little understanding of the competitive, mass consumer marketplace.^[7]

Some of the utilities' difficulty in assessing the residential market developed because they had no standard accounting methods to analyze the domestic rate structure. They could not determine how to charge the home user of appliances to make a profit. A survey in 1923 of thirty-five midwestern utilities revealed that in small operating utilities, with less than a thousand domestic customers, the utility managers had "almost a total lack of knowledge of what it costs to carry a residential consumer." The domestic rate posed an economic conundrum not solved by the utilities until the early 1930s. On the one hand, the cheap, flat rate enabled the utilities to fulfill their political mandate for universal electrification. On the other hand, the flat rate discouraged consumption of electricity. How did this situation arise? Samuel Insull's Chicago Edison system established the pattern of charging domestic customers by a cheap, flat rate in the 1890s. Insull adopted a two-tier rate that he thought would permit massive residential consumption, though this consumption did not develop as he expected. Ordinary household customers had a flat illumination rate. Large consumers had a second, lower, rate for consumption above a base amount. The two-tier billing structure reflected the utilities' segmented domestic market conception. Shortly before World War I, state regulatory commissions enshrined the flat rate as democratic social policy and thereby threw onto the utilities the burden of justifying a departure from this historic practice. Such justification proved extremely difficult. Not until the early 1930s did the utility industry agree on accounting procedures to determine profit and loss in delivering electrical service to the individual customer.^[8]

Theoretically, the flat rate represented the average charge to the individual customer needed to cover capital and operating expenses, proportional to the aggregate ratio of capital and operating expenses for domestic customers as a whole. Being an average charge, domestic customers who consumed high amounts of electricity subsidized domestic customers who consumed low amounts. An analogous situation characterized the telephone industry before

1894. Bell systems charged users a universal flat rate. Only when telephone companies switched to per-message rates could average monthly charges drop sufficiently to make telephones attractive for homes, which used them less than did businesses. Before World War I, the electrical industry could tolerate the economics of the flat rate. Since sales to industry and electric railroads earned most of the utilities' profit, the loss due to subsidization was not great. It certainly did not seem worth the effort to put every domestic customer on a profit basis. During normal growth, the addition of profitable customers balanced the addition of subsidized customers. Subsidized customers had little reason to object. Profitable customers did not use enough kilowatt-hours of electricity monthly to make carrying a subsidized customer burdensome. Across the country, in the 1920s, the domestic flat rate ranged between nine cents and seven cents per kilowatt-hour (kwh). In 1922, a study by the Idaho Power Company demonstrated that to make a profit the company would have to charge 10.5 cents/kwh for the average domestic electric light customer (who used about 21 kilowatt-hours per month [kwh/mo.]). The flat rate did not provide for a decrease in charges as volume of consumption increased, until demand jumped to a higher step. For lower-income households, lack of incentive to increase consumption amounted to disincentive. The utility industry itself felt little motivated to try to change the flat rate structure, as long as economic prosperity continued. As late as 1928, according to an executive of the New York Power and Light Corporation, two of three cities in New York State charged a flat rate that discouraged increased residential consumption of electricity. As a result of this situation, the central station electrical generating industry halfheartedly expressed interest in the domestic market. As the editors of Electrical World put it in 1926, "As a matter of fact, frankly admitted, the central-station industry has never actually made serious, intelligent effort to develop the domestic market."^[9]

The industry closely watched the Hartford Electric Light Company's experiment with a split rate structure. Before 1922, the company attempted to induce higher domestic consumption of electricity by offering a lower flat rate of 3 cents to 4 cents/kwh for customers with heavier wiring. Only five hundred residences (Out of 15,000 served) fitted their homes with heavy wiring capable of taking heavy appliances. Prompted by this failure, in January 1922 the company offered a split schedule, whereby the customer paid 5 cents monthly per 100 square feet of floor area in the home and paid 6 cents/kwh for metered, delivered energy. For domestic customers with light wiring, this split schedule was cheaper than a flat rate of 10 cents/kwh. Over the decade, the company further reduced rates, so that in 1928, residences using over 200 kwh annually paid only 1.5 cents/kwh of delivered energy. By the company's own analysis, the experiment succeeded hugely. In 1921, only 10 percent of the domestic customers used over 480 kwh of energy a year; in 1928, 33.6 percent did.^[10]

On the basis of the Hartford experiment and other similar marketing experiments, the utility industry concluded that it had to change the domestic rate structure to reflect the distinction between capital equipment costs and operat-

ing costs. Capital investment referred to the expense of extending electrical service to a home, including distribution lines, step-down transformers, light lines, and meters. Operating expenses referred to the costs created by generating the electricity sent over the wires to the residence. If the utility could charge each domestic customer to pay for its capital investment to the residence, then it could encourage all customers to consume more electricity with delivered energy rates cheaper than the existing flat rates. By the early 1930s, utilities were applying to utility commissions for the right to bill a fixed charge to cover capital investment, with a variable electricity rate above that. This split schedule permitted utilities to drop charges as low as 3.5 cents/kwh for delivered energy. As it turned out, the split rate with a lower kilowatt-hour charge returned greater profit to the utilities than the old flat rate, because of economies of scale in central station production of energy.^[11]

In the late 1920s, scattered empirical evidence indicated that domestic demand for electricity would increase if prices fell significantly. The evidence convinced some utility managers of the potential of the domestic market. This internal persuasion in the industry never had the opportunity to ripen of its own logic, however; historical events intervened. The onset of the depression changed the context in which utilities thought about and acted in the domestic market. In response to general deflation, regulatory commissions forced domestic rates down. Because of rate experiments in the 1920s, the utilities were ready to respond to the demands of regulatory commissions. Accountants had devised standardized accounting procedures to analyze residential accounts. The new split schedule provided a convenient method for the companies to negotiate rate reductions with the commissions. Without doubt, the pressure of the regulatory commissions accelerated adoption of the new rate and accounting procedures inside the electric industry.^[12]

In the economic collapse, the electric utilities lost a huge portion of their industrial market. By 1937 the manufacturing share of distributed electricity had dropped (to 52.3 percent). Who would replace the industrial customer? They could not expect to shift their sales to commercial and retail customers. The depression halted the growth of the commercial market. Nor did electrified mass transportation promise to absorb electricity not sold to manufacturing. The shift of the nation's urban transportation away from electric streetcars in the 1920s accelerated after 1929, so streetcars, once the economic core of the utility business, offered no potential for increased sales. In desperation, the private utilities turned to the only market left—the home. Electrical World made the point clearly: "Interest has centered in the home this year.... Because of the depression, neither the factory, store nor office has presented a hopeful field for rapid development." Being compelled by the depression to promote the domestic market did not necessarily make the utilities enthusiastic about that market. A positive attitude came later, from the success of the great public power experiment of the 1930s that the utilities had fought successfully in the 1920s: the Tennessee Valley Authority. TVA provided a vast experiment in home electri-

cal modernization. Its extraordinary success in increasing domestic electrical usage among some of the poorest American households in the nation overwhelmingly proved that domestic demand for electricity could be elastic: the lower the price, the more households consumed. "While private utility managers believed in the value of electricity and its promotion, few realized that electricity demand had such a high elasticity. In other words, TVA planners demonstrated that usage could be stimulated dramatically with extremely low rates, which would be justified only if usage actually increased through promotion and appliance saturation." The depression and the TVA committed the private utility industry to the mass domestic market in a way that the prosperity of the booming 1920s could never do.^[13]

While utilities approached the home market hesitantly, manufacturers of electric appliances enthusiastically expanded production for the home market in the 1920s. The number of electrical manufacturing establishments increased from 1,333 in 1921 to 1,777 in 1927, and their sales increased from $833 million to $1.6 billion. The increase in manufacturers indicated the entrance into the domestic market of small start-up companies, which might annually build less than a thousand units of an appliance. Large and established corporations also saw opportunity in home sales. General Electric, the nation's largest electrical manufacturer, turned to the residential market to recover from the post-World War I depression, in which G.E. laid off a quarter of its workforce. When Gerard Swope became president of G.E. in 1922, he pushed the company into high-quality appliance lines, starting with refrigerators.^[14]

The complications of manufacturers' participation in the residential market, however, constrained its growth. The largest home appliance companies, General Electric and Westinghouse, were deeply involved with the electric power utilities and had to accept the basic segmentation of the residential market as strategized by the utilities. General Electric created the Electric Bond and Share Company in 1905, for instance, to finance expansion of operating utilities, which would purchase electrical equipment from their shareholder, G.E. The issue of the home electrical appliance market in the 1920s is not simply its size and dynamics. It is also the extent to which market categories prescribed by the utilities bound appliance manufacturers. A small historical literature and a few industry aggregate statistics indicate that the manufacturers, like the utilities, did not try to create a mass market for their products in the 1920s. Income segmentation of households raised market barriers for appliance manufacturers as for the utilities. For the appliance manufacturers, as for the utilities, the mass market came in the mid-1930s.^[15]

Internal business considerations also limited the interest of large manufacturing corporations in the home market. Domestic appliances represented only one product group in their diversified production. Electrical equipment manufacturers produced for a large market of value-added manufacturers. Home electrical appliance makers took only a small percentage of their production. In 1927, home heating and cooking appliances took only 4 percent of manufacturers'

sales; radio equipment, 12 percent. Manufacturers' sales to makers of electrical home appliances totaled only $73 million. The automobile and utility industries, by contrast, took a huge share of manufacturers' sales. Battery sales (9 percent) and generator sales (7 percent) grossed $265 million of revenue. Radio sales greatly exceeded all appliance sales. General Electric and Westinghouse built a wide variety of capital goods for the electric utilities and for other industries, as well as consumer appliances. In the early 1920s, General Electric marketed nearly 400,000 separate catalog items. Home market divisions of these corporations had to compete with other divisions for the corporation's investment capital. G.E.'s board of directors passed on all departmental requests for investment capital exceeding $100,000 for production purposes and $50,000 for research. This scrutiny generated intense interdepartmental competition at budget proposal time, driven by the same profit considerations as intercorporate competition. How did the home market appliance division's gross revenue, return on investment, percentage of profit to sales, and percentage of profit to value added compare to other divisions? The results of this comparison affected the capability and willingness of corporation directors to wage battle in the home marketplace, or to try more ambitiously to break down the utilities' segmentation of the residential market.^[16]

Languid price competition among appliance companies is a strong indicator of their lack of interest in breaking down segmentation of the home market. They kept prices high throughout the decade. In 1929, the average price for electric refrigerators advertised in Riverside, California, newspapers was $268, for ranges $111, for washing machines $ 16, and for radios $116. General Electric, Westinghouse, and Allis-Chambers divided the market and administered prices for industrial electrical equipment in the early 1920s. Despite Gerard Swope's exhortation that price competition in home appliances would benefit his company, manufacturers and retailers did not lower prices of large home appliances rapidly enough in the 1920s to make them affordable to the lower four-fifths of the nation's households. Nor did they have to. Deriving but a small percentage of their income from appliance sales, large diversified companies could afford to keep sales volume low and target households in the top income bracket with high-priced models. The impression is irresistible that the major manufacturers wanted to extend to the new home market the same oligopolistic, administered market structure they had for the market in industrial electrical equipment. General Electric used its patents to control the production and pricing of lightbulbs—by far the largest-grossing product for the home market in the 1920s. Westinghouse, for instance, used G.E. patents to manufacture its bulbs. They did not engage in the severe price cutting that would crush small companies and enlarge their market share at the expense of their large rivals.^[17]

Not until the depression did prices of refrigerators, ranges, and washers fall sufficiently to break the dual market barrier of the 1920s. Only when TVA created a prototype mass market did prices fall to the point at which nearly every household, as long as it had some regular income, no matter how small, could

buy a major appliance. Industry priced only small appliances—cookers and casseroles, desk fans, heating pads, irons, percolators, and toasters—for the mass market in the 1920s; yet, looking at price changes between 1921 and 1941, only the refrigerator took its largest price drop in the 1920s. Before 1934, few appliance manufacturers priced any models within reach of a true mass market. TVA compelled the large producers, particularly General Electric and Westinghouse, to develop mass market models for it; prices dropped as much as 50 percent below those of 1932. Seven of the eleven leading household appliances took their first or second greatest drop in price after 1932, almost certainly the result of New Deal pressures. At this point, aggregate sales began to soar.^[18]

The difficulties for home mechanical refrigeration illustrate the barriers to the domestic market for manufacturers. Engineers did not make key technological inventions until the 1920s. Once manufacturers commercialized these inventions, such as the best refrigerant, electric refrigerators began to flood the market in 1929, just before the depression. Problems of mass marketing of electric refrigerators were, however, just as important as the pace of technical invention in delaying the device's mass adoption. In particular, the manufacturers' acceptance of the utilities' segmentation of the residential market in the 1920s prevented them from trying to develop a truly mass market machine. Before the early 1920s, appliance manufacturers developed industrial and commercial markets for mechanical refrigeration. These markets offered greater profit per refrigerator unit than the home market. Hotels, restaurants, ice cream parlors, and meat and fruit wholesalers were early buyers of mechanical refrigerators. In agriculture in the 1920s, California, Texas, and Florida produce packers built large mechanical precoolers to prepare fresh fruit for long rail journeys to eastern markets by slowly lowering fruit temperature before railcar loading. These commercial and industrial refrigerators were usually of walk-in size and custom designed. Domestic refrigeration represented a significantly different market. Mechanical refrigerators for the home had to be much smaller than commercial refrigerators. In the home, cheap ice dominated refrigeration. In the cities, many households did without any refrigeration at all, preferring to shop as needed for fresh foods rather than storing them. In 1927, only 40 percent of the nation's households had refrigeration (nearly entirely ice refrigeration), and only 17 percent took ice year-round. Home marketing required more than simply sending salesmen out to knock on doors; it required a different product.

The major problem for manufacturers in developing a home refrigerator was lack of standardization of components used in making refrigerators and lack of market standardization in the design and capacity of the refrigerators themselves. Since the industry sold to large commercial and industrial establishments, they designed the motors, compressors, and other equipment to cool large volumes of air. The industry did not standardize fractional horsepower motors for small refrigerators until late in the decade. As late as 1925, motors required higher voltage to operate than many homes could offer. At the 1925 meeting of the National Electric Light Association, "some representatives [of refrigerator and mo-

tor manufacturers] said that they had frequently found the supply as low as 70 volts, which, of course, is altogether inadequate for this service, as well as for general service in the home. This was a very general complaint." Motor manufacturers invested in research on fractional horsepower motors reluctantly, because the market for home refrigerators was small. Refrigerator manufacturers similarly had difficulty standardizing the refrigerator itself because the market was so small. For instance, Commonwealth Edison of Chicago undertook a concerted sales campaign in 1923 for home electric refrigerators. The company contracted with a manufacturer for a single small standard box to sell to customers. In 1923 and eleven months of 1924, the giant utility installed only 279 refrigerators. Not until 1926 did the domestic refrigeration market take off, with national sales moving toward several hundred thousand units annually.

The small number of refrigerators sold in the mid-1920s meant that the product had not reached the stage of mass production. In 1925, Frigidaire, the largest manufacturer of refrigerators, had only 35,000 of its machines installed around the country. Kelvinator followed with 26,840 installed. Servel held third place with only 3,500 refrigerators in homes. In 1926, there were ninety-five different makes of home refrigerators on the market, many of them manufactured by companies with annual sales of less than twenty-five units. The distributor-dealers of large manufacturers, such as Kelvinator, Frigidaire, and General Electric, maintained such large repair shops for preservicing the units prior to installation in homes that we cannot call the machines mass produced at all. The low sales volume and the large preinstallation servicing kept prices high. Salesmen reported that potential customers resisted their sales pitch because of the high initial cost of the device. The high price meant that upper-income homes represented the only market for the device.

For prices to drop, manufacturing had to be concentrated in a small number of companies that could build refrigerators in sufficiently high volume to develop effective mass production and to gain economics of scale. Manufacturers had to increase the quality of the machines so that they could move from factory to home directly, without preinstallation services. When this happened, major retailers, such as Sears Roebuck, which did not want to finish manufacturing of low-quality boxes, could market them. At the same time, equipment manufacturers had to develop more efficient motors that worked off low-voltage domestic lines. This conjunction of events occurred in the early 1930s, and prices dropped rapidly. Nonetheless, even as late as 1935, upper- and middle-income families made up the majority of owners. A survey of Austin, Texas, showed that only 15 percent of all households had electric refrigerators but that half of all families with an annual income of $3,000 owned them. A true mass market did not exist until the federal government made refrigerators eligible for Title I insured loans under the National Housing Act and TVA forced manufacturers to build mass market models for the TVA sales programs.^[19]

In conclusion, electrical manufacturers did not attempt to break residential market segregation in the decade. Their ties to the utilities effectively bound them

to the latter's segmentation of the home market. Even Swope complained that the persistent high price for electricity doomed his ambitions for broadening the domestic market for G.E. during the decade. The depression broke the stranglehold of the utilities' conceptualization of the nation's electrical market. In 1929, electrical manufacturers sold $2.3 billion worth of products; in 1933, they sold only $549 million worth of goods. While general price deflation accounted for some of this decline, decreased sales of their industrial products accounted for most of it. In a long review of the industry in 1935, Electrical World 's editors argued that the domestic market represented the only growth market left. New Deal policies had made domestic electrical modernization a national priority. "Electrification is in popular favor. Electrical products and electric service are considered necessary to advance both economic and social standards in this country. Government spending and action are directed toward increased electrification." The electrical market created by the New Deal redirected investment of electrical manufacturing companies and enabled them to work their way toward prosperity.^[20]

Enthusiasm of appliance manufacturers for the home market produced an explosion of advertisements for appliances in the mass media. The deluge of advertisement would seem to imply that selling of consumer appliances had no obstacles. Careful study of merchandising in the 1920s shows, to the contrary, that problems in merchandising restrained electrical appliance retailing until the 1930s. The vice president of Commonwealth Edison of Chicago called the distribution of appliances so poor as to constitute a "joke." Three problems were particularly troublesome: the small number of retailers, lack of credit, and the lack of a service industry.^[21]

Five classes of retailers sold electrical appliances: the electric utilities, independent appliance specialty stores, hardware stores, department stores, and mail-order houses. The utilities and the mail-order houses used traveling salesmen to supplement walk-in trade. In 1923, electrical utilities sold 31 to 41 percent of total sales (in dollar volume) of electrical appliances, dominating retailing, especially of heavy appliances. Department store and mail-order retailing did not take significant market shares until late in the decade, when consumer advocates and other retailers politically pressured utilities to withdraw from retailing. This means that a seller—the utilities—that had little interest in the home dominated electrical appliance merchandising. Concentration of retailing in a halfhearted retailer throttled sales. Suppliers did not compete in price. Customers found shopping inconvenient. Retailers did not solicit the customer's attention. In contrast, in one of the few cities where the central station utility did not retail appliances, there were two additional classes of retailers—home furnishing stores and electric shops. Department stores had a slightly higher percentage of the trade, but retailing remained decentralized. Retailers competed in prices, a shopper could easily locate a store selling an appliance, and salesmen competed in service to the potential customer.^[22]

Retailing did not improve in the next two years. In 1925, Electrical World

surveyed the selling of electrical appliances in the geographic areas served by 174 operating utilities. Electrical utilities sold 42.5 percent of all electric appliances, in dollar volume, compared to 26.9 percent by electrical dealers (including specialty stores), 15.4 percent by department stores, and 15.2 percent by all other stores. In conclusion, the domination of retailing by the electric utilities in the first half of the 1920s constrained the market for appliances. Even Electrical World concluded that electrical modernization of the home would proceed more rapidly if electrical utilities did not dominate appliance merchandising. "If other dealers, both electrical and non electrical, could or would build up their sales volume so as to reduce the power companies' proportionately large share of the total, it would aid materially in a quicker and more complete electrification of the home." Consumers, too, wanted to end the utilities' domination of appliance merchandising. In 1931, Oklahoma and Kansas passed laws prohibiting utilities from selling appliances. With utility merchandising stopped, the number of appliance retailers increased dramatically over the next four years. Sales of appliances increased. In 1930, Oklahoma consumers bought only 2,085 clothes washers; with utilities out, in 1932 they bought 5,487 washers (in the worst year of the depression!), then increased purchases to 8,662 in 1933 and 11,067 in 1934. Sales of all other appliances similarly soared as dealers and merchants invigorated the market.^[23]

At the end of the decade, the utilities' domination of retail selling of appliances began to diminish. In 1930, they constituted only 30 percent of national aggregate sales of electric appliances. Observers thought department stores took an increasing share. In 1931, department stores made 24 to 30 percent of retail sales of appliances across the nation. While the number of stores specializing in electric appliances increased in the years 1925 to 1929, the depression reduced their ranks. The decision of Sears Roebuck, of Chicago, in 1931 to offer refrigerators through catalog sales typified the shift in merchandising. The depression rapidly completed this reorganization of the retailing structure of the domestic market for electrical appliances. By the mid-1930s, the utilities no longer dominated retailing and sales of electrical appliances rose dramatically.^[24]

The Consumer's Perspective

Powerful restraints reined in consumers' buying of electric appliances in the 1920s. Reduced capability to take on more installment debt, unreliability of the new technology, difficulty in obtaining repair service, incompatible home tenure status, competition from alternative technologies, and preference for services over savings suppressed purchases. Most households bought appliances on installment credit. The amount of credit consumers could acquire and the way they preferred to use it affected what they purchased. Consumers overwhelmingly preferred spending their money on automobiles rather than on household durable goods. Households spent 31 percent of their expenditures on durable goods on automobiles, and only 8 percent on household appliances. The extent

of household finances devoted to the automobile dampened purchase of other major goods. Contrast with the 1930s shows the preferences. When hard times reduced the ability of households to buy automobiles, they nonetheless increased their spending on household appliances. The share of household budgets consumed by auto purchases increased only 4 percent in the 1930s over the 1920s. The share of budgets consumed by appliance purchases increased 24 percent. The shift in consumer preferences is more dramatic when stated in dollars. In 1929, of every $100 of consumer debt, consumers spend $47 for automobiles, and only $.91 for home repair and modernization loans. In 1936, they spent $42 for automobiles, and $11 for home repair and modernization loans.^[25]

Although appliances were cheaper than automobiles, sales of automobiles soared over appliances. The annual retail sale of autos rose steadily from 1,457,000 cars in 1921 to 4,036,000 in 1929, totaling 26,957,000 vehicles over those years. The cumulative number of autos sold in the 1920s nearly equaled the number of households in 1929 (29,582,000). In nine years, the auto had attained nearly universal adoption. Henry Ford helped the nation to envision a true mass market for the automobile, and in the 1920s automobile manufacturers both created and saturated that market. The contrast with electric home appliances is telling. For every electric dishwasher Americans purchased in 1929, they purchased 336 cars; for every electric desk fan, 6 cars; for every refrigerator, 5 cars. Only the radio surpassed the popularity of the auto in 1929; Americans bought 4,438,000 radios that year. The electrical industry of the 1920s simply failed to grasp the nature of a mass market for their products.^[26]

The mass adoption of the radio in the 1920s reveals, as does mass adoption of the automobile, the willingness of households to spend money and, by contrast, their reluctance to spend money on electrical appliances. Commercial radio broadcasting began in 1920. Manufactured radio receivers were widely available by 1922. Montgomery Ward sold its first receiver for $49.50 in its 1922 fall catalog. Radio sets quickly became popular, but not until mid-decade did the radio industry overcome marketing barriers to offer a mass product. Early radios took power from a battery, offering as little as forty hours of reception before the owner had to recharge the battery at a store or utility. Replacement of short-lived vacuum tubes added to inconvenience. Although the radio industry touted its own growth, by the end of 1925, it had placed only 2,850,000 radio sets in American homes, barely 10 percent saturation. The accomplishment is even less impressive when we consider that no division between nonelectrified rural homes and electrified urban homes hobbled marketing of battery-powered radios.

Beginning in late 1925, political, technological, and cultural developments rapidly created a national mass market for radio. After a year of intensely competitive research, in the fall of 1925 ten manufacturers began marketing radio receivers that plugged directly into house current. They built all the components of the radio—receiver, amplifier, tuner, speaker, battery eliminator—into a single cabinet. A year later, RCA, General Electric, and Westinghouse, the major

corporations involved in American radio, established the National Broadcasting Company. NBC began the first nationwide continuous programming. In 1927, the federal Radio Act regularized (and thinned the ranks of) the hundreds of independent broadcasting stations, making the airwaves more useful for networked broadcasting. Significantly, by explicitly opposing monopoly, the new law indirectly forced RCA to license key patents to competitors, opening up the number of manufacturers of radios, increasing competition, and lowering the costs of radio sets. In just two years, from 1925 to 1927, manufacturers and broadcasters transformed radio, positioning the industry to satisfy a nearly insatiable demand by American households.

The eagerness with which Americans brought the house current radio into their homes places into perspective their laggard adoption of labor-saving household electric appliances. In the late 1920s, the radio receiver became the most widely adopted electricity-consuming domestic device, after the flatiron and the lightbulb. By 1933, Americans had purchased over 24,400,000 radios, enough for universal adoption by electrified households. The radio consumed enormous amounts of electricity, compared to flatirons and bulbs, and even compared to the vacuum cleaners and washing machines used by a minority of homes in 1928. The radio broke the long-standing resistance of domestic consumers to high consumption of electricity. The mass adoption of the very expensive automobile throughout the 1920s and the radio at the end of the decade revealed starkly the consumer choices made by Americans. They chose to modernize transportation and information, but not household labor. The electrical industry caught the comparison. Electrical World 's editors scolded their readers: "The more impressive thought is this—that if after all these years a normal household already spends as much for radio as it does for all the benefits of electricity, the electrical industry has not done a selling job of which it can be very proud."^[27]

Failings of appliance technology and inadequate repair service also blocked domestic electrical modernization in the 1920s. The technological optimism of advertising did not accurately reflect reality. Lack of standardization meant a consumer could not be sure that she could plug an appliance into her home. Industry standardized plugs and receptacles on the parallel-bladed plug only in the mid-twenties. She could not be sure that wiring and heating element connections were durable, or that she could conveniently make repairs. Major appliances were in their early stages of development with attendant problems of product reliability. An engineer cited as typical a washing machine in which the handle came off, two rivets popped, the motor's torque drove the bracing out of alignment, and the machine worked well only for a month. One service organization, the United Appliance Company and Electric Shop Company, of Jackson, Michigan, kept service records on 450 mechanical refrigerators under its service contracts. Over six years, those 450 refrigerators required a total of 1,440 service calls, an average of 3.2 per machine. Problems included faulty automatic controls, burned out motors, and leaking tanks. The National Electric Light Association calculated that refrigerator service cost as much as $24.39

per year. Detroit Edison Company—a large company serving a large city, admittedly—made 173,928 appliance repair calls in 1923. In August, the company's repair shop made 13,392 repairs on electric iron cords alone. Addressing the Electric Association of Chicago in 1926, Samuel Insull complained, "It is impossible to escape the conclusion that lack of standardization in appliance manufacturing, and the consequent effect upon selling prices, has been a large factor in retarding the use of domestic appliances." Faulty appliances deprived utilities of potential sales of electricity so large that one utility found it economical to establish regular service routes for inspection and maintenance of customers' appliances and to offer the first dollar of repairs free. As late as 1929, the problem of appliance quality remained sufficiently serious that the Association of Edison Illuminating Companies began an inspection and testing program. When utilities were slow to protect consumers from poor merchandise, merchants sometimes stepped in. In 1928, seventeen New York City department stores, in cooperation with insurers and electrical inspectors, organized an inspection program to remove unsafe and poorly manufactured electrical goods from their shelves. Consumers also organized to protect themselves from shoddy manufacturing and lack of standardization. By 1933, forty communities, most in western states, had passed ordinances prohibiting local merchants from selling unsafe appliances, requiring standardization in wiring, and providing for inspection.^[28]

Competing technologies and services erected a third major barrier to domestic electrical modernization in the 1920s. Adoption of electric appliances frequently required displacing existing nonelectric appliances, as electric cooking had to replace gas cooking. Despite evidence from home economics studies that cooking costs for gas and electric ranges were roughly equal, or could be made so by careful selection of an efficient electric range, gas ranges overwhelming dominated urban cooking. In 1930, 13.7 million homes used piped gas to cook, and less than a million used electricity to cook. Persistent use of cheap ice refrigeration slowed adoption of electric refrigerators. Once widespread adoption of the domestic mechanical refrigerator began, the ice industry greatly improved ice boxes, intensifying competition. It added, for instance, the finned metal grid for better circulation of cold air and better insulation.^[29]

As their incomes rose, households bought services, rather than labor-saving appliances. They ordered industrially canned and bottled foods from stores and prepared foods at delicatessens by telephone, or went out to restaurants instead of preparing food at home. Women went out to beauty parlors, rather than curl their hair at home with electric curling irons. Once bobbed hair became a popular women's hairstyle in the mid-twenties, the sale of electric curling irons declined, since the style required the gas iron available at parlors. In cities and towns, street peddlers sold seasonal fruits and vegetables and meats and fish door to door, thereby making long-term storage of fresh produce unnecessary. Dairies delivered fresh milk products to homes. We can easily imagine that a city household would be reluctant to spend money for a mechanical refrigerator

Plate 2.
Competition for Electrical Refrigerators . ([Photographer unknown.]
"The Meat Wagon." Keystone-Mast Collection. UCR/California
Museum of Photography, University of California, Riverside.)

when they had minor year-round cold storage needs. Scholars generally presume that street peddlers disappeared in the 1920s, as Americans bought cars and sped from store to store. This probably did not occur until the 1930s. The business licenses for Riverside, California, through the mid-1920S prove that street peddlers were plentiful and frequent, year-round. (See pl. 2.) As many as four delivery persons and peddlers visited each home every day. Plied from

horse-drawn carts and trucks, seasonal fruits appeared at every household's door: Washington navel oranges and grapefruit in the winter, Valencia oranges in the summer, apples from Oregon and Washington in September, apples from southern California mountains in October and November, truck crops from the neighboring arroyos and river bottom all year long. Ice came daily. Laundries made regular pickups and deliveries. In the winter of 1922, police issued forty-seven peddler's licenses, permitting sale of over twenty items, including hats, furs, clothes, raincoats, medicinals, belts, ice, milk, melons, fish, meat, fruit, vegetables, ladies garments, shoes, and walnuts. This situation did not change until the mid-1930s, when the great increase in floating unemployed population and homeless transients, many of whom begged door to door, cast general suspicion on street vendors. At the same time, adoption of refrigerators after 1935 enabled households to dispense with the services of peddlers.^[30]

The competition between commercial laundry service and domestic electrical washing machines epitomizes how competition from services dampened consumer demand for electrical appliances. Ownership of home electric clothes washers dramatically increased in the early 1920S. The washing machine apparently had "profound implications for the reorganization of work in the households of the prosperous. Possession of an electric washing machine meant that a 'decent' housewife could do her wash at home." By encouraging women to do laundry in the home themselves, instead of having servants do it or sending it out to commercial laundries, ownership of the home washer dramatically increased the effort, time, and onerousness of laundry duty. These new duties would have ended the Victorian ideal that the bourgeois woman should supervise labor, rather than be a laborer, and use time freed from labor to cultivate the home as a moral setting.^[31]

Did a laundry revolution really occur? Statistics on adoption of electric washing machines provide the strongest evidence that it did, but the statistics are misleading. For most of the decade the percentage of electrified homes owning washers increased little. In 1922, 34.3 percent of electrified households had electric washing machines; in 1929, 38.4 percent had them. Between 1921 and 1942, the nation's stock of washers never dramatically and quickly rose so as to indicate a popular revolution in housekeeping. When we look at the percentage of all homes (nonelectrified as well as electrified), the ownership of washers is less impressive. The saturation of all households rose from 10.2 percent in April 1922 to 20 percent in 1929. Four of five American homes did not adopt the electric washing machine in the booming twenties.^[32]

Instead of laundering clothes at home, families used wet-wash and dry cleaning services. By the end of the 1920s, nearly two-thirds of households in cities larger than 50,000 persons sent laundry to commercial laundries or to laundresses outside the home. While the ability to send laundry out depended to a certain extent on household income, a study of 12,096 wage-earning families in 1918–1919 showed that 43.4 percent of all families earning less than $900 a year sent laundry out. Isabel Lord, who directed her budgeting textbook toward

the small percentage of upper-class families who paid a federal income tax in the early 1920s, assumed they would send laundry out. A 1928 study of 306 families, most of whom were college educated and lived in midwestern small towns, showed that 22 percent of them used commercial laundries or hired laundresses. Detailed study of six of the forty-two cities included in the Bureau of Labor's 1934–1936 household survey confirms the resistance of households to doing their own laundry. Among lower-middle-income households, even in the depression, as many as a quarter to a half of families paid to have their laundry done. In Denver and Los Angeles, over 40 percent of all households paid for laundry service.^[33]

When wet laundries opened in a neighborhood, sales of washing machines declined. In Chicago, electric washing machine sales were high during the early 1920s but dropped off at mid-decade after forty-five wet-wash services set up in the city. By 1926, these washes were sending Out 350,000 bundles a week, servicing a sizable minority of the city's households. National total receipts for commercial laundering and dry cleaning rose from $362 million in 1925 to $541 million in 1929. Southern appliance dealers long believed that the inexpensive services of African-American laundresses represented a real obstacle to increased sales of electric washers in their region. Phyllis Palmer verified this folk wisdom for one southern city, Jackson, Mississippi, in the early 1930s. She found that 8 1 percent of all white households in Jackson either had black laundresses in to do laundry or sent out their laundry to black home laundresses or commercial laundries. Half of all white families also paid black housekeepers to iron their laundry. A concerted campaign by laundry owners in 1934 to slow the loss of their business to home electric washing machines produced striking success. Lowering the price of damp wash to forty-nine cents for a fifteen- to twenty-pound bundle brought a 150 percent increase in laundry business in Milwaukee and Oklahoma City, a 100 percent increase in business in Denver, and a 250 percent increase in Kansas City.^[34]

Home tenure provided a fourth major constraint on electrical modernization of the household. Until the post-World War I period, most American households did not own their home. Not owning, they did not have the right to renovate the home as needed for electrical modernization. They also would not have the home equity with which to secure modernization loans. Even if they could renovate, they had less interest in doing so than owners. Renters moved more often than owners. Modernization (as by incorporation of an electric range) might encumber the opportunity to move. Mildred Wood, a home economist, reflected this concern in her 1932 discussion of the advantages of owning a home, when she mentioned, as a reason to own rather than rent, that owners could accumulate possessions without having to worry whether they would fit in the next place they lived.^[35]

Appliance Sales and Household Formation

Two additional factors limited adoption of appliances during the 1920S: household formation and dwelling technology. Analysis of appliance sales of six electric appliances—flatiron, heating pad, vacuum cleaner, washing machine, range, and refrigerator—shows the influence of household formation. Sales of these appliances did not grow uniformly in the 1920S. In the first half of the decade sales increased rapidly, but sales in the second half of the decade slowed. The slowing is clear in growth rates. Growth rates express more complex demand dynamics than annual sales figures. Sales can increase year to year in absolute numbers, but the rate at which they increase can decline, reflecting some underlying change in demand. In the 1920S, the refrigerator's growth peaked in 1926, when sales were 120 percent greater than the previous year. Range growth peaked in 1926, also, when its sales were 26 percent greater than the previous year's sales. Clothes washer and vacuum cleaner sales growth peaked in 1923, when washers sales were 29 percent greater and vacuum cleaner sales 24 percent greater than the previous year. Heating pad sales growth peaked in 1927, at 71 percent greater than the previous year. Well before the financial crash in fall 1929, electric home appliance sales growth peaked.

The slowing of sales growth coincides with a well-known, national demographic slowdown. In the mid-1920s, the number of marriages, new households, and persons under four years of age began to decline, not to resume an upward trend until the mid-1930s. The growth rate of household formation declined from 2.63 percent in 1921 to 2.02 percent in 1926 and 1.56 percent in 1929. New housing starts in urban areas peaked in 1925 (at 752,000 units) and declined to little more than half that number (400,000 units) in 1929, before the onset of the depression. The growth rates in sales of refrigerators, ranges, vacuum cleaners, and heating pads strongly correlate to growth rates of new households over the years 1921–1941. (Fig. 1.4 illustrates the comparative growth rates for refrigerator sales, households, and housing starts.) After the economic collapse, sales began to rise in 1933. The increase in appliance sales rode the New Deal and the demographic wave upward, following the increase in housing starts and in marriages, new households, and births. These changes in sales rates reflect the significant relationship between the growth in appliance sales and the growth in new households. The stimulation of the economy by the New Deal after mid-1933 increased the nation's capability to form new households. In this way, the appliance sales boom from 1933 to 1941 reflects improving economic conditions resulting from governmental stimulation of consumption.^[36]

Normalizing the annual appliance sales against the number of households captures the close relationship between sales and household formation. Normalized data show clearly that the market dynamics of appliance sales in the 1920S were qualitatively different from the 1930s (after 1932). Except for the flatiron and the vacuum cleaner, normalized sales in the 1920s were lower than

in the 1930s. Some of the normalized sales rates of the 1930s are as high as or higher than the normalized rates of the 1940s and 1950s, the postwar economic boom years. Only flatiron sales grew most rapidly in the 1920s (mean rate = 21.8 per 100 households per year during 1921–1929). Refrigerator sales grew greatest in 1933–1941 (rate = 10.3). Sales of ranges (rate = 11.4), washers (rate = 9.7), vacuum cleaners (rate 7.2), and heating pads (rate = 4–5) grew most rapidly in 1946–1964. Only the electric range sold (at normalized rates) in postwar years significantly more than before the war, reflecting the significant increase in suburban households who elected to cook with electricity rather than gas and the electrification of the South.^[37]

Without knowing the trends of the mean sales growth rates, we cannot interpret the means unambiguously. The unqualified mean rate of sales falling from ten to one million units a year is the same as the mean rate of sales rising from one to ten million units a year, yet no one would say that the averages are the "same." We need to examine the sales trends by direction of change, or vector. Doing so, we see immediately the importance of the 1930s. In terms of the sales growth of the household electric appliances, the decade saw a vector growth in sales rates, while the earlier or later periods saw vector decline in sales rates. Relative to the number of households living in dwellings technologically eligible to employ them, the 1930s were more important for appliance sales than the 1920S or the postwar years. A true mass market in electrical household goods existed for the first time. Sales after the war, from 1946 to 1964, trended downward. In the immediate postwar years, the pent-up demand for appliances expressed itself all at once. The years 1946, 1947, and 1948 experienced spectacular increases in the growth of sales. After a brief lull, in 1950 sales rates rose dramatically for the last time. The outbreak of the Korean War at midyear stimulated a splurge in buying by households fearful that the war would end consumer production, as had World War 11. After 1950, growth in sales of appliances never again reached the height of the immediate postwar years. Viewed as a whole, therefore, the period 1945–1964 represented a long-term decline in growth of appliance sales, reflecting the slowing of the postwar baby boom. (See table 1.3.)

Dwelling Fitness for Electrical Modernization

The dwellings in which most households lived also affected their adoption of electric appliances in the 1920s. Electrification did not lead directly to electrical modernization of homes. Because a dwelling had electricity did not mean modernization could proceed simply by plugging in appliances and using them. Conflating electrification with electrical modernization confuses two distinct technological systems. Before federal home improvement programs during the New Deal, wired dwellings differed greatly in their capability to take electric appliances. The number and placement of outlets, the style of receptacles for

plugs, and the type, load capacity, and insulation of wiring restricted the number and type of appliances that households could connect to the lines. Heat-producing appliances, such as ranges and ovens, required heavy, insulated wiring. Early refrigerator motors required higher wattage than was normally available to homes. Households could plug a small power appliance into lines, but only if a room had the appropriate receptacle. Until the 1930s, contractors usually did not build dwellings to take a full range of electrical appliances; they wired most dwellings for illumination only. Looking at the bleak prospect for putting significantly more appliances into the substandard American housing stock, the editors of Electrical Merchandising put the issue succinctly in 1926. "Some six or seven million new residence customers have been connected up in the United States since the close of [World War I]. These houses have been 'wired for electricity,' so the layman says. Yet every electrical man knows that on the average these homes contain not more than one-quarter to one-half of the wiring equipment which their owners should have installed in justice to themselves and to their families." The vast extent of inadequate residential wiring so alarmed Electrical Merchandising that the editors complained in italicized print: "Electrical articles are the only ones which cannot be taken home and put to use by the purchaser, when, where, and as he pleases! "^[38]

The nation would not easily remedy the material inadequacy of the American dwelling. Despite the new houses built in the housing boom of the 1920s, the electrical industry continued to complain about the substandard electrical infrastructure of the American home well into the next decade. The general manager of the Virginia Electric & Power Company's Electric Department summarized industry opinion in 1933:

Residential wiring has not kept pace, particularly as regards capacity, with the improvement in utilization devices. Interior wires and convenience outlets of adequate size are not available in any but a small percentage of residences. Many of the present popular appliances require electric currents exceeding the capacity of the common standard convenience outlets and the wires supplying them. Among these appliances the following are mentioned: Automatic flatirons, ironing machines, hot plates, portable ovens, space heaters, small water heaters, sun lamps, etc.^[39]

A nation ill-housed did not provide a mass market for an "electrical age."^[40]

We can distinguish between four classes of housing from the point of view of electrical service: dwellings with no electricity; dwellings with electricity only for illumination; those with the additional wiring for small appliances; and those fully wired to take stationary heating appliances. These divisions in the housing stock developed historically as builders constructed housing within the regulatory context of municipal building codes. The building codes evolved in response to changing building technologies and housing politics. Local communities around the United States wrote their own building, plumbing, and electrical codes under state enabling legislation. Standards reflected each commu-

nity's particular environmental hazards, its building history, and the nature of its building stock. Before 1923, cities usually wrote electrical codes only to require electrical illumination and to ensure that electrical wiring, fixtures, and apparatus would not be a fire hazard. Most cities did not require builders to install wiring for electrical appliances. The 1923 electrical code for Bakersfield, California, for instance, required only that ceilings be wired to provide for a minimum number of "lamps" (bulbs). The city legislated that ceiling outlets in living rooms and dining rooms be wired to service at least four forty-watt light-bulbs and bedrooms be wired to supply electricity for at least two forty-watt bulbs. The 192 2 electrical code for Pasadena, California, required that each room have a ceiling outlet wired to support at least two forty-watt bulbs. To ensure fire safety, both Bakersfield's and Pasadena's ordinances stipulated that all electrical installations must meet the requirements of the National Electrical Code of the National Board of Fire Underwriters.^[41]

The National Electrical Code originated in 1897. A consortium of private manufacturers and insurers, the National Conference on Standard Electrical Rules, wrote the code to reduce fire hazard from early electrical installations and thereby to promote electrification. After disbanding of the conference, the National Fire Protection Association took over the function of testing electrical equipment and revising the code, which it did every two years. Members of the National Fire Protection Association included the American Electric Railway Association, the American Institute of Electrical Engineers, the Association of Edison Illuminating Companies, the National Board of Fire Underwriters, and the National Bureau of Standards (a U.S. government agency). The code had no legal force in itself; it constituted recommendations only. To provide legal force for the National Electrical Code, cities had to incorporate it into municipal ordinances. The 192 3 revision of the code was particularly important for domestic modernization, because the association made rules more flexible for plugging in appliances. The 1920 code limited small motors attached to a heavy wire branch circuit to a maximum of 660 watts and limited the number of lamp sockets on a single circuit to sixteen. To meet the code at minimal cost, builders would install only a few illumination outlets (to serve lamp fixtures with multiple sockets), requiring only one circuit. This practice made the safe use of appliances nearly impossible. The 1923 code increased the maximum number of outlets per branch circuit, thereby reducing the cost of wiring while increasing electrical service. In a further effort to make dwellings safe for plugging in appliances, the code provided that circuits for small motor appliances should be of larger wire than lamp circuits and electricity provided through specially designated "convenience outlets." Heavy appliances required a third separate circuit, of still larger wire. The code discouraged use of socket receptacles for appliances, because plugging appliances into lamp sockets could burn out the lamp wiring and start a fire. Some cities did not require convenience outlets; Bakersfield and Pasadena did not, and undoubtedly many dwellings in those cities had no convenience outlets. In such dwellings, households wishing to use an electrical

appliance would have to remove a lamp from an outlet and screw in a socket receptacle—a widespread practice around the United States (and still a frequent source of New Yorker cartoons). The failure of building or electrical codes to require convenience outlets partially accounts for the finding of a survey in the mid-1920s that the average dwelling in the United States had only three convenience outlets. Since builders frequently installed only a single ceiling outlet per room for a lamp fixture, households tended to fill base outlets, if they had any, with portable lamps, leaving no outlets open for appliances. A few cities took advantage of the code revision to require separate circuits for appliances. Chicago's 1923 electrical code revision required that convenience receptacles for appliance plugs be provided in living rooms, dining rooms, and kitchens. The 1923 version of the National Electrical Code thereby implicitly established categories of building quality in terms of electrical service .^[42]

Meeting code requirements for small motor appliances, such as vacuum cleaners and sewing machines, added cost to wiring installation. Installation costs further rose if dwelling owners wanted to use heavy appliances. The code required that heavy heating appliances (such as stoves and water heaters) should be on a separate circuit with larger wire than either illumination or small motor circuits. Heating appliance circuits should also have indicator lamps to indicate whether the appliances were on, switches to shut off electricity to the appliance in an emergency, special grounding, and special fuse boxes. Wiring into kitchens with indoor plumbing required rigid, sealed, metal conduit to protect the heavy wiring from mechanical and water damage. Dwellings electrified for illumination only would not, therefore, have the approved number of circuits, size of wiring, safety devices, or outlets needed for safe use of appliances.^[43]

The kind of protection required for wiring also differentiated the housing stock into classes. The National Electrical Code recommended different wiring protection corresponding to three, different, major wiring methods: rigid conduit, flexible conduit, and knob and tube. Rigid conduit involved armoring insulated wiring with rigid metal pipe and sealing it against moisture. Flexible conduit referred to sheathing insulated wiring with helically wound metal stripping. The metal protected wire against mechanical injury but—because the conduit had movable seams—not against water. Knob and tube installation left rubber-insulated wires exposed. Porcelain "knobs" carried wires through the dwelling like tiny power poles. Rigid porcelain tubes or flexible rubber tubes protected wires as they passed through partitions. Many cities, including Bakersfield, Pasadena, Chicago, Denver, and Kansas City, required metal conduit for concealed wiring in large buildings (e.g., office and public buildings and apartment houses with more than six units). Though the municipalities allowed flexible conduit or rigid conduit, metallic conduit was extremely expensive for retrofit installation, because installers had to open walls, ceilings, and floors. The author of one practical treatise on retrofit wiring delicately put it, "It is very difficult ... to wire an old building throughout with rigid conduit without a great deal of cutting and disfigurement of walls and ceilings." Utilities and builders

reluctantly installed more rigid conduit than the minimum required in dwellings, because most households used so little electricity that they gained little profit. Communities with a lot of inexpensive frame housing usually permitted knob and tube installation, because the cost of the housing would not warrant the more expensive methods. Nonetheless, knob and tube wiring for lights only provided minimal electrical service; safe addition of small motor appliances and heavy appliances still required upgrading the wiring. Electrical World estimated in 1931 that home owners could raise the average dwelling (of that year) with substandard wiring and fixtures to a "minimum standard" for a cost of $70, including materials, labor, and cost of lighting fixtures. From the point of view of the wiring required for electrical modernization, lights-only wiring was therefore substandard.^[44]

I have estimated the changing proportions of the nation's housing stock in the four classes of housing: nonelectrified, "substandard," "standard," and "modern." Substandard dwellings had illumination wiring and minimal illumination only. Standard dwellings had many lights, higher-wattage bulbs, and power wiring for nonheating electrical appliances. Modern dwellings added heavy circuits for heating appliances to the standard wiring. In 1922, 80 percent of the nation's dwellings had no electricity or only substandard wiring. Only 19 percent had standard lights and wiring. By 1932, the situation had improved. Sixty-three percent of dwellings had no electricity or substandard wiring only while 34 percent were in the standard class. (See table I.I and pl. 3.) In other words, in the 1920s, nearly two-thirds of the nation's dwellings were not technologically capable of being electrically modernized by the simple installation of better illumination and power appliances. Substandard housing alone, among other factors, prevented a household revolution through electrical modernization. By contrast, at the end of the 1930s, two-thirds of all dwellings were technologically capable of being modernized with appliances. The New Deal accomplished this upgrading in the electrical status of the nation's dwellings. (See fig. 1.1.)^[45]

In conclusion, underlying social, economic, and demographic constraints on demand limited adoption of electric household appliances in the 1920s. An insufficient number of households adopted appliances to constitute a social revolution, or even a significant social change in the mass of households. The nation's capitalist economy distributed income highly unequally during the decade. Mainly the upper fifth of households, ranked by income, bought appliances. We have no evidence that upper-class women underwent a social revolution—whether liberation or proletarianization—from using the appliances they bought. The evidence indicates that adoption of appliances conservatively reinforced social status and social relationships. Prices remained high during the decade, as utilities and manufacturers sought to satisfy the luxury market. They failed to create a true mass market for the domestic devices. Average households had limited economic capability to adopt appliances, largely because they

Plate 3.
Substandard Urban Housing . (Clare L. Brian, photographer.
"Urban poverty in Bloomington, c. 1935–1940. ['Sanitation pictures']."
Courtesy of McLean County Historical Society, Bloomington, Illinois.)

had committed their debt to the automobile. Most of their dwellings were not technologically capable of serving multiple minor appliances or heavy appliances. Households did not choose to invest money in upgrading their dwellings' electrical wiring, undoubtedly because most households rented. Growth in aggregate demand for household appliances slowed in the latter half of the 1920s, as household formation declined and the number of new houses being built also declined. For a mass market in domestic electric appliances to exist, the nation had fundamentally to restructure the economy of its homes. The New Deal did.

Appendix

Definitions

Dwellings with no electricity: Urban and rural dwellings with no electrical service. This category includes dwellings that the 1933–1936 real property inventories and the 1940 and 1950 censuses identified as dilapidated and providing inadequate physical shelter.

Substandard dwellings: Substandard urban dwellings as a percentage of all occupied dwellings. The substandard dwelling has only wiring for illumination. Because of the light wiring, households could not install additional outlets for power appliances (directly off the illumination circuit). Frequently, a single 40- or 50- watt lightbulb illuminated each room. This category also includes all urban dwellings, even when the state of their electrical service is unknown, identified by the census with substandard plumbing, defined consistently by the real estate inventories and national censuses to be lack of private toilets or lack of indoor toilets (for urban dwellings only) and lack of running hot water. The assumption is that dwellings with substandard plumbing are unlikely to have more than substandard electrical service.

Standard dwellings: Urban dwellings with wiring for illumination, a branch circuit for convenience outlets, and more than average number (3.4 in 1924) of convenience outlets, and all rural and farm dwellings with electricity. Urban standard structures were usually not wired for heating appliances, such as the range and water heater. Urban dwellings had private plumbing and running hot water. I class farm dwellings as standard if they had electricity, regardless of the state of their toilets, because farm structures were usually

wired for power at the same time they were wired for lights. Farmers frequently took electrical service, since it enhanced farm productivity, before they built indoor toilets.

Modern dwellings: Urban dwellings with electric ranges. Dwellings have standard light wiring and above standard high conduit, permitting installation of a large number of outlets, especially in the kitchen. In the absence of national statistics of the number of urban dwellings with heavy wiring, I have used the estimated stock of domestic electric ranges as an indicator of dwellings with heavy wiring. Given that builders did not, as a matter of course, install heavy conduit branch circuits in new housing until after 1945, it is likely that most civilian housing before 1942 received such wiring by retrofit when the dwelling owner wished to install a heavy appliance. As such, the percentage of dwellings with such wiring is probably the same as the percentage of dwellings with installed electric ranges. After 1945, the range becomes a less reliable indicator of the presence of heavy wiring, because contractors built most new dwellings ready for heavy appliances. If a dwelling did not have a range, it may have had an electric water heater or air conditioner.

Methodology

Dwellings without electricity and standard dwellings determined by linear estimation from cross-sectional surveys and decennial census statistics from the U.S. Bureau of the Census. Modern dwellings determined by national stock of electric ranges, as explained above. Substandard dwellings determined by subtracting the estimated number of non-electrified, standard, and modernized dwellings from the total number of occupied dwellings. Although the procedure for determining the percent of substandard dwellings is indirect, its result is close to Edith Wood's survey of the nation's housing. Edith Elmer Wood, a housing reformer, estimated that in 1930, 34 percent of the nation's urban dwellings were substandard; this compares to my estimate of 31.7 percent in 1932. (I calculated the unitary urban substandard dwelling statistic from Wood's data in Edith Elmer Wood, Recent Trends in American Housing [New York: Macmillan, 1931], 40.) I do not distinguish between standard and superior farm dwellings. Most farms (87 percent) were electrified after REA electrification began, when federal intervention increased the wiring standards. In addition, most farmers intended to put heavy motors or farm electrical equipment on the lines that required heavy branch wiring. Wiring costs would quadruple ($100–$200) the cost of retrofit wiring of urban dwellings ($26–$50) at this time. See, for example, J. C. Damon, "Lower Cost Wiring," Electrical World 105 (September 14, 1935): 32, for wiring costs of a typical urban house. For these reasons, when farms were electrified, they usually were given power wiring (60-amp. capacity). See Donald S. Stophlet and Tom F. Blackburn, "Selling Appliances on New Rural Lines—2 Reports," Electrical Merchandising 61 (May 1939): 8 f.

Because independent estimations were made for the different classes of dwellings in table 1.1 and because of rounding in chain calculations, the columns do not cumulate to 100 (percent). A full account of the housing model is provided in Ronald C. Tobey, "Statistical Supplement to Technology as Freedom " (unpublished ms., 1995).

Figure 1.1.
Dwellings Ineligible for Electrical Modernization. 1920–1954

Source:
See Introduction, table I. 1.

INTERPRETATION :
The trend line after 1932 projects the decline of ineligibility of
dwellings for electrical modernization by the rates of private
modernization through 1929. The observed line begins in 1935.
The difference between the projection of the private trend line and
the observed line estimates the impact of New Deal housing programs.

Figure 1.2.
Heavy Electrical Appliances Annual Sales, 1919–1965

Source:
See Introduction, table I.1

Figure 1.3.
Small Electrical Appliances Annual Sales, 1919–1965

Source:
See Introduction, table I.1

Figure 1.4.
Comparative Growth Rates, 1922–1940

Source:
For refrigerators and households, see introduction, table I.1. For housing
starts, see Series N 157, "New Housing Units Started, by Ownership, Type of
Structure, Location, and Construction Cost: 1889–1970," U. S. Bureau of the
Census, Historical Statistics of the United States, Colonial Times to 1970,
Bicentennial Edition, Part 1  (Washington, D.C., 1976), 2: 639.

Note:
Refrigerator rate is 3-year moving average of sales
growth rate; r  = LN (p2/p 1)/t* 100.