The Manufacture of Wants

Essay · society · September 2026

A century ago the people who built the consumer economy wrote down what they were doing: training people to want new things before the old ones wore out, keeping them reasonably dissatisfied, lending them the money, making consumption a way of life. The evidence that it worked is not as simple as the story, but it points one way: persuasion, credit and status competition made us buy more, and to buy more we worked more.

Reading time ≈ 13 min Bernays → Galbraith → the feed A follow-up to Where the Hours Went

The claim

Where the Hours Went asked why the productivity gains of the last century were taken as income rather than as time, and ended with a sentence it did not prove: that a good deal of the income was spent “on things we were persuaded to want”. This essay makes that case. It follows the persuasion from the people who designed it, through the mechanisms they used, to the evidence on whether it worked, and then to Portugal.

The argument is that wanting more is not a fixed trait of human nature that the economy simply serves. It was cultivated, deliberately, by an industry that said so in print; it was financed by credit that separated buying from paying; and it was sustained by a competition for status that no one can win, because the prize is relative. The strongest objections are real, and they are stated below. They limit the argument; they do not overturn it.

$1.14 tnworld advertising revenue in 2025, about 1% of world GDP, three-quarters of it digital

+6.9%hours worked due to advertising, in an estimated model of the US economy; households end up worse off

−3%life satisfaction for a doubling of national advertising spending, across a million Europeans

126%Portuguese household debt relative to disposable income in 2009, from 38% in 1995

In their words

They said so

The case does not rest on conspiracy. The men and women who built modern marketing described their aims in books, articles and speeches, and much of it is quoted accurately below, from the original texts; some famous versions are not accurate, and are corrected.[1] Figure 1 puts them in order.

Fig. 1 — Selected events in the history of advertising, obsolescence and consumer credit, 1919–2025. Quotations are from the original texts except Watson’s, which is from an archival address quoted by historians. Sources in the notes.

The first principle was that people could be steered without knowing it. Edward Bernays, Freud’s nephew and one of the founders of public relations, opened his 1928 book Propaganda with the sentence: “The conscious and intelligent manipulation of the organized habits and opinions of the masses is an important element in democratic society.” He asked, a few pages later, whether it was “not possible to control and regiment the masses according to our will without their knowing it”, and described how to sell pianos not by advertising pianos but by making the music room fashionable, so that the desire for a piano “will come to him as his own idea”.[2] The behaviourist psychologist John B. Watson left Johns Hopkins in 1920 for the J. Walter Thompson agency, and told its staff in 1922 to reach the consumer by telling him “something that will tie up with fear, something that will stir up a mild rage, that will call out an affectionate or love response”.[3]

The second was that needs are finite and desires are not. The Lehman Brothers banker Paul Mazur wrote in 1928 that a community “that can be trained to desire change, to want new things even before the old have been entirely consumed, yields a market to be measured more by desires than by needs. And man’s desires can be developed so that they will greatly overshadow his needs.” (The line often attributed to him, “we must shift America from a needs to a desires culture”, is a later paraphrase.)[4] Charles Kettering, head of research at General Motors, wrote in 1929 under the title “Keep the Consumer Dissatisfied” that a banker had once told him the only reason for research was “to keep your customers reasonably dissatisfied with what they already have”, and that “he was right”: “You must accept this reasonable dissatisfaction with what you have and buy the new thing, or accept hard times.”[5]

The third was that things should not last. Christine Frederick, in Selling Mrs. Consumer (1929), defined “progressive obsolescence” as, among other things, “a readiness to ‘scrap’ an article before its natural life of usefulness is completed”, and wrote that “America’s triumphs and rapidity of progress are based on progressive obsolescence”. Bernard London proposed in 1932 that the State set a legal life for every product, after which it would be “legally ‘dead’”, and complained that “people everywhere are today disobeying the law of obsolescence”. The light-bulb manufacturers had not waited for a law: in 1924 their cartel fixed the life of a bulb at 1,000 hours, down from 1,500 to 2,000, and fined factories whose bulbs lasted too long.[6]

The retail analyst Victor Lebow summed up the post-war programme in 1955: “Our enormously productive economy demands that we make consumption our way of life, that we convert the buying and use of goods into rituals, that we seek our spiritual satisfactions, our ego satisfactions, in consumption.” Lebow was describing the system, arguably with some unease, rather than cheering it; but his description was accurate.[7] Two years later Vance Packard’s The Hidden Persuaders documented the “motivational research” of Ernest Dichter, whose advice to advertisers was to “give moral permission to have fun without guilt”.[8] And in 1958 John Kenneth Galbraith drew the economic conclusion: “One cannot defend production as satisfying wants if that production creates the wants.”[9]

A market to be measured more by desires than by needs. And man’s desires can be developed so that they will greatly overshadow his needs.Paul Mazur, American Prosperity, 1928

How

The mechanisms

Five mechanisms run through that history, and each has been studied.

Dissatisfaction by design. The first is to make what people already own feel inadequate: the annual model change, the new phone, the fashion season. The world made roughly twice as many garments in 2015 as in 2000, and each was worn about a third fewer times before being thrown away, as Where the Hours Went showed. Obsolescence need not be physical; it is enough that the old thing becomes embarrassing.

Emotion over information. Watson’s fear, rage and love, and Dichter’s moral permission, are the method of most advertising still: it associates a product with a feeling or an identity, not with facts about it. The American Psychological Association concluded in 2004 that most children under about eight “do not comprehend the persuasive intent of advertising”, at a time when American children saw more than 40,000 commercials a year and marketers spent more than $12 billion a year to reach them.[10]

Status and emulation. Much consumption is a signal, and signals are relative: a bigger car is only bigger than someone else’s. The evidence here is the strongest. When a Dutch household won the postcode lottery, its next-door neighbours became almost five percentage points more likely to buy a new car.[11] In the United States, as incomes at the top pulled away, middle-income households spent more and saved less: had top incomes grown only as fast as the median, middle-income households would have spent 2.6 to 3.2 per cent less by 2005, and the national saving rate would have been almost twice as high. The effect was concentrated in visible goods, and the households most exposed reported more financial distress and filed for bankruptcy more often.[12] The Duty to Work showed the same comparison at work in hours: where inequality is greater, people work longer.

Credit. Desire needs money, and credit supplied it before the income arrived. General Motors created a finance company in 1919; by the late 1920s most American cars were bought on instalments. The economic historian Martha Olney argues that credit and advertising together produced a real shift in what Americans wanted; Lendol Calder replies that Americans had always borrowed, and credit only made it respectable.[13] In 1958 the Bank of America mailed 60,000 unsolicited credit cards to the residents of Fresno, California, and the revolving debt that followed made the gap between wanting and paying permanent.[13]

Attention. The newest mechanism is the most efficient: advertising that is paid for by the attention it captures, targeted person by person, on platforms designed to hold that attention as long as possible. World advertising revenue reached $1.14 trillion in 2025, about 1 per cent of world GDP, three-quarters of it digital.[14] When people in an experiment gave up Facebook for four weeks, they gained an hour a day and reported higher well-being, by about a quarter to two-fifths of the effect of therapy; when Facebook arrived at American colleges, students’ mental health worsened, driven by unfavourable comparison with others.[15] The Accelerant took that story further.

The evidence

Does it work?

That persuasion was intended does not prove that it worked. The question the earlier essay raised is whether it changed how much people consume, and so how much they work. The table sets out the evidence on both sides, with how strong each design is.

Study	What it measures	Finding	Design
Molinari & Turino	advertising in a model of the US economy	consumption +4.2%, hours +6.9%, welfare lower	estimated model
Cowling & Poolsombat	advertising and hours, US 1962–2001	more advertising, longer hours; larger than tax effects	time series
Michel, Sovinsky, Proto & Oswald	advertising and life satisfaction, 27 countries	doubling ad spend: −3% satisfaction	country panel
Bertrand & Morse	top incomes and middle-class spending, US	middle spends more, saves less, more distress	regional panel
Kuhn et al.	neighbours of lottery winners, Netherlands	more new cars; no change in happiness	natural experiment
Dittmar, Bond, Hurst & Kasser	materialism and well-being, 259 samples	r = −0.19	meta-analysis
Ashley, Granger & Schmalensee	aggregate advertising and consumption, US	no evidence advertising causes consumption	time series
Shapiro, Hitsch & Tuchman	TV ads and own-brand sales, 288 brands	median effect small; most not different from zero	quasi-experiment

Table 1 — The main studies on whether advertising and emulation raise consumption, hours and dissatisfaction, and against. The first six support the argument; the last two are the strongest evidence against it. Full references in the notes.[16]

The case for is strongest where it is least about advertising. An estimated model of the American economy finds that advertising raises long-run consumption by about 4 per cent and hours worked by about 7, and that households are “unambiguously worse off because the overworking effect more than compensates for the expansion in consumption”; they would give up to 3 per cent of their consumption to have advertising banned.[17] Time-series studies find that more advertising goes with longer hours in the United States and Britain.[18] Across a million Europeans in 27 countries over three decades, a doubling of national advertising spending goes with a 3 per cent fall in life satisfaction, about half the effect of marriage in the other direction.[19] And the evidence on emulation, above, is among the best-identified in the field.

The case against is strongest on advertising itself. The classic test of whether aggregate advertising drives aggregate consumption found no evidence that it does; if anything, spending drives advertising, because firms advertise more when sales are good.[20] The most careful modern study of television advertising, across 288 brands, finds that most campaigns barely move their own brand’s sales, and that the median brand loses money on its advertising.[21] And American advertising cost roughly 2 per cent of GDP for most of the twentieth century, with no upward trend, as The Duty to Work noted; in Portugal it is far less, about 0.3 per cent. The leading survey of the economics concludes, fairly, that “more extensive formal empirical studies are necessary before a stand can be taken”.[22]

These objections are weaker than they look, for three reasons. The brand studies measure whether one brand’s campaign takes customers from another, not whether the whole weight of advertising, year after year, shapes what people think a normal life requires; a campaign can fail at the first and the industry succeed at the second. The flat share of GDP measures what advertising costs, not what it does: the price of reaching a person has fallen with every new medium, so the same share of GDP now buys far more attention. And the mechanisms that the evidence supports most clearly, emulation, credit and obsolescence, are the ones the persuasion industry was built to exploit. Advertising is the visible part of a system whose effect runs mostly through products designed to be replaced, credit designed to be easy and a culture in which what you own says who you are.

Portugal

The Portuguese version

Portugal arrived at mass consumption late and quickly, and the clearest sign of it is debt. Household debt was 38 per cent of disposable income in 1995. With the euro came low interest rates and eager banks, and by 2000 it had doubled; at its peak in 2009 it was 126 per cent, close to Spain’s 134 and far above Germany’s (Figure 2). Since then it has fallen, to about 76 per cent.[23] Most of it was borrowed to buy homes, but not all: the stock of consumer credit nearly doubled between 2003 and 2009, and after a fall in the crisis new consumer loans rose from €2.0 billion in 2012 to €7.1 billion in 2025.[24]

Fig. 2 — Household (and non-profit institutions’) debt as a percentage of gross disposable income, Portugal, Spain and Germany, 1995–2025 (latest years provisional). Data: Eurostat, nasa_10_ki.

Saving moved the other way (Figure 3). Portuguese households saved 14 per cent of their disposable income in 1995 and 7 per cent in 2007 and 2008, while the EU average stayed around 12. It rose in the crisis, fell back to below 7 per cent in 2017, and has recovered since the pandemic to about 12.[23]

Fig. 3 — Gross household saving rate, per cent of gross disposable income, Portugal, EU-27 (from 1999) and Germany, 1995–2025 (latest years provisional). Data: Eurostat, nasa_10_ki.

What the money bought changed less than the stories of consumerism suggest (Figure 4). Between 1995 and 2022 the share of Portuguese household spending on housing and energy rose from 13 to 17 per cent and on restaurants and hotels from 11 to 15, the latter inflated by tourists’ spending, while clothing, recreation and cars fell as a share, because they became cheaper.[25] The Portuguese spend less of their budget on things and more on places: where they live and where they eat. That fits the argument of Where the Hours Went, that the dear things are positional, better than it fits a simple story of more gadgets.

Fig. 4 — Portuguese household final consumption expenditure by purpose (COICOP), share of the total, 1995 and 2022, selected categories. Restaurants and hotels includes spending by non-residents in Portugal. Data: Eurostat, nama_10_co3_p3.

Portugal has also chosen to limit some of the persuasion. Its advertising code has, since 1990, forbidden advertising that exploits the inexperience of minors or encourages them to pester their parents; since 2019 advertising of food high in fat, sugar or salt is banned within 100 metres of schools, around children’s programmes and on websites and apps aimed at under-16s.[26] Those rules protect children from the advertising that the evidence says works best on them. More than half of Portuguese pupils from the fifth year up spend four or more hours a day in front of screens on school days.[27]

What follows

Wants are made

The argument, then, is this. The consumer economy was built by people who said, in print, that they meant to train desire beyond need, to keep people dissatisfied and to make things that would be replaced. They had tools that worked: credit to spend income not yet earned, products designed to date, and a competition for status that every rise in inequality intensifies. The evidence that these raised consumption and hours is strongest for emulation and credit and weakest for any single advertisement, and the cost in well-being is measurable. Keynes’s grandchildren did not choose freely to take their productivity as goods rather than time; they chose inside a system designed to make that the likelier choice.

The strongest reply is Friedrich Hayek’s, to Galbraith: almost every want beyond food and shelter is learned from others, including the want for music, literature and education, and the fact that a desire is taught does not make it false.[28] That is true, and it marks the limit of the argument: nobody can say which wants are “real”. But the case here does not depend on that distinction. It depends on the fact that one set of wants, those that can be sold, has an industry of a trillion dollars working on it, and the other set, for time, rest and each other’s company, has none. Societies that want a different balance have ways to get it: ad-free public space, as São Paulo decided in 2007; limits on advertising to children, as Portugal has started; a right to repair, which the EU adopted in 2024; and, above all, the choice to take some of the next rise in productivity as time.

On method and tools

This piece was written collaboratively with Claude Opus 5.5 (Anthropic): human specification, editorial direction and critical review; machine data research, analysis and drafting. Quotations were checked against scans or transcriptions of the original texts where these could be found; where a popular version differs from the original, the original is given and the difference noted, and quotations seen only in secondary sources are identified as such. The figures are computed by scripts/wants.py from Eurostat’s sector accounts and consumption statistics; results are in docs/wants-results.json and the figures in docs/wants-figures.html. The studies in Table 1 were read in their working-paper or published versions; effect sizes are quoted from them with page references in the research notes. The world advertising share of GDP is the author’s calculation from WPP Media and World Bank figures.

The cover photograph is Toward Los Angeles, California by Dorothea Lange, March 1937, for the Farm Security Administration; public domain, via Wikimedia Commons (Library of Congress), cropped.

Authored by: Luis Matos Ferreira — Physicist, Developer, Writer

Related essays on this blog

1. Where the Hours Went — who got the income, what it bought, and where the hours of work go.
2. The Duty to Work — how the centrality of work was instilled, and whose interests it serves.
3. The Accelerant — social media, attention and what it does to a society.
4. The Fifteen-Hour Week — Keynes’s prediction, productivity and the hours that were not given up.

Sources

1. Research notes on the history of advertising, obsolescence and consumer credit, with the wording, edition and page of each quotation and the status of each (original text, transcription or secondary source).
2. Edward Bernays, Propaganda, Liveright, 1928, pp. 9, 47 and 54–56.
3. John B. Watson, address to J. Walter Thompson staff, 1922, Library of Congress, as quoted in Gondra, 2014; Buckley, Mechanical Man: John Broadus Watson and the Beginnings of Behaviorism, 1989.
4. Paul M. Mazur, American Prosperity: Its Causes and Consequences, 1928 (London, Cape edition), pp. 37–38.
5. Charles F. Kettering, “Keep the Consumer Dissatisfied”, Nation’s Business 17(1), January 1929, pp. 30–31 and 79.
6. Christine Frederick, Selling Mrs. Consumer, Business Bourse, 1929, ch. 25, pp. 245–253; Bernard London, Ending the Depression Through Planned Obsolescence, New York, 1932; Markus Krajewski, “The Great Lightbulb Conspiracy”, IEEE Spectrum, 24 September 2014.
7. Victor Lebow, “Price Competition in 1955”, Journal of Retailing 31(1), Spring 1955; on Lebow’s intent, Benton, Journal of Historical Research in Marketing, 2020.
8. Vance Packard, The Hidden Persuaders, McKay, 1957, quoting Ernest Dichter.
9. John Kenneth Galbraith, The Affluent Society, 1958, ch. 11, “The Dependence Effect”.
10. American Psychological Association, Report of the APA Task Force on Advertising and Children, 20 February 2004.
11. Kuhn, Kooreman, Soetevent & Kapteyn, “The Effects of Lottery Prizes on Winners and Their Neighbors: Evidence from the Dutch Postcode Lottery”, American Economic Review 101(5), 2226 (2011).
12. Bertrand & Morse, “Trickle-Down Consumption”, Review of Economics and Statistics 98(5), 863 (2016); NBER working paper 18883, pp. 3–4.
13. Martha Olney, Buy Now, Pay Later: Advertising, Credit, and Consumer Durables in the 1920s, University of North Carolina Press, 1991; Lendol Calder, Financing the American Dream, Princeton University Press, 1999.
14. WPP Media, This Year Next Year, December 2025; World Bank, World Development Indicators, world GDP 2025; author’s calculation.
15. Allcott, Braghieri, Eichmeyer & Gentzkow, “The Welfare Effects of Social Media”, American Economic Review 110(3), 629 (2020); Braghieri, Levy & Makarin, “Social Media and Mental Health”, American Economic Review 112(11), 3660 (2022).
16. Research notes on the evidence, with page references for each finding in Table 1.
17. Molinari & Turino, “Advertising and Aggregate Consumption: A Bayesian DSGE Assessment”, Economic Journal, 2018; working paper version, Universidad Pablo de Olavide, 2015.
18. Cowling & Poolsombat, “Advertising and Labour Supply: Why Do Americans Work Such Long Hours?”, Warwick Economic Research Paper 789, 2007, p. 13; Fraser & Paton, 2003, as cited in Benhabib & Bisin (2011).
19. Michel, Sovinsky, Proto & Oswald, “Advertising as a Major Source of Human Dissatisfaction: Cross-National Evidence on One Million Europeans”, CAGE working paper 397, 2019, p. 7.
20. Ashley, Granger & Schmalensee, “Advertising and Aggregate Consumption: An Analysis of Causality”, Econometrica 48(5), 1149 (1980).
21. Shapiro, Hitsch & Tuchman, “TV Advertising Effectiveness and Profitability: Generalizable Results from 288 Brands”, Econometrica 89(4), 1855 (2021).
22. Benhabib & Bisin, “Social Construction of Preferences: Advertising”, in Benhabib, Bisin & Jackson (eds.), Handbook of Social Economics, vol. 1A, 2011.
23. Eurostat, Key indicators, households (nasa_10_ki): gross saving rate and gross debt-to-income ratio of households, 1995–2025.
24. European Central Bank, balance sheet items and MFI interest rate statistics, Portugal, loans to households for consumption, stock and new business, 2003–2026.
25. Eurostat, Final consumption expenditure of households by consumption purpose (nama_10_co3_p3), Portugal, 1995–2022.
26. Código da Publicidade, Decreto-Lei n.º 330/90, art. 14; Lei n.º 30/2019, de 23 de Abril (arts. 20.º-A and 20.º-B).
27. Observatório da Saúde Psicológica e do Bem-Estar nas escolas, 2024, as reported by CNN Portugal, 9 December 2024.
28. Friedrich A. Hayek, “The Non Sequitur of the ‘Dependence Effect’”, Southern Economic Journal 27(4), 346 (1961).