Does manifestation actually work? What 40 years of research says

Manifestation is not one practice. Several separate habits travel under the one name, and the research treats them separately. The component with measured support is mentally rehearsing the specific process toward a specific goal, repeated over time: a medium effect (Simonsmeier, Andronie, Buecker and Frank, 2021: d = 0.431), and in the older pooling it came out smaller than physical practice (Driskell, Copper and Moran, 1994: r = .255 against .364). The component that reliably costs you something is vividly enjoying the finished result (Oettingen and Mayer, 2002). Since 2023 there is also research on the believers themselves, not only on the technique.

Is manifestation real? What the word actually covers

Manifestation is real as a belief (35% of one US online panel endorsed it: Dixon, Hornsey and Hartley, 2023, Study 1 N = 306) and untested as a mechanism. The word is a folk category rather than a single measurable procedure: it covers the law of attraction, vision boards, the “369 method”, writing the outcome out as though it had already happened, affirmations and plain mental rehearsal. Psychology has instruments for three of those: mental imagery, positive fantasies versus positive expectations, and implementation intentions, plus a thin and mixed literature on repeating self-statements. The rest of the bundle has no measurement tradition behind it.

The only study of believers themselves

The believers themselves have been studied once. Dixon, Hornsey and Hartley (2023) built an 11-item Manifestation Scale across three US Prolific samples, total N = 1,023 (306, 348, 369). The 35% is the share of the first sample whose mean score fell between 5 (slightly agree) and 7 (strongly agree) — 49.7% on the Personal Power subscale, 33.3% on Cosmic Collaboration. A paid panel, not a national sample.

What the belief travels with

The financial findings are associations, not causes. Study 3 analysed 369 of the 400 people recruited, and each model reports its own analytic sample. Manifestation scores went with belief in getting rich quickly (r = .43, beta = .40, the largest unique association among the psychological variables in that regression; N = 357).

Each one-point increase on the 1–7 scale went with 1.33 times higher odds of currently holding cryptocurrency (N = 354), no significant difference for having ever invested in stocks other than 401k retirement funds (OR 1.16, n.s., N = 355), and 1.42 times higher odds of reporting a personal or business bankruptcy (N = 353).

Correlational and not pre-registered, these studies show what the belief travels with, not what it causes. Believers rated themselves as more successful, yet their scores were unrelated to income or education attainment: the belief paid in confidence, not in measured outcomes.

What the studies actually measure

Imagined and executed movement run on an overlapping brain network, not on the same one. Hardwick, Caspers, Eickhoff and Swinnen (2018) pooled 303 motor-imagery experiments with 4,902 participants against 142 movement-execution experiments with 2,302 participants. Where the two maps overlap, they share a sensorimotor and premotor network extending to the putamen and the cerebellum, while primary motor cortex is recruited only when the movement is executed; the authors amend earlier models that claimed equivalence.

What gets tested under the word “visualization” is that imagery: a specific action imagined from the inside, in first person. Imagery is rehearsal of an action, so what gets rehearsed matters more than how vivid the picture is.

The correction retires a popular line. The claim that the brain treats an imagined action and a performed one as the same event is folk phrasing, not a finding of this literature, and the imaging is precisely what separates the two. Popular framings persist anyway: Stanford neurosurgeon James Doty's Mind Magic (Avery, 2024) is a trade book, not a study, and reports no sample of its own.

Is manifestation scientifically proven? The effect is real and modest

No. Manifestation as a bundle has never been tested. One component has been: mental rehearsal of a specific process, with a medium meta-analytic effect (Simonsmeier et al., 2021): d = 0.431, 95% CI [0.298, 0.563]. The narrower contrast in the same meta-analysis — imagery plus physical practice against physical practice alone — is much smaller: d = 0.156, 95% CI [0.002, 0.310], a lower bound just off zero. That headline number stacks on top of practice rather than substituting for it.

Earlier meta-analyses

The older estimates agree and add the limits. Driskell, Copper and Moran (1994) found mental practice improved performance at r = .255, against r = .364 for physical practice, and the benefit shrank as the interval between rehearsal and performance grew (r = −.216). Real, smaller than doing the thing, and decaying without repetition.

Feltz and Landers (1983) reached the same conclusion a decade earlier, pooling 60 studies and 146 effect sizes for an average effect of .48, with a warning that published studies showed larger effects than unpublished ones.

The evidence at a glance

What the table does not cover

A boundary goes with that table. The imagery meta-analyses pooled motor and sport tasks, not money outcomes, and none of them reports an income effect. The planning meta-analyses cover goals across many domains, but neither reports a separate estimate for financial goals.

Why imagining the result backfires

Rehearsing how you would do the thing helps; enjoying the finished result in advance costs effort. Pham and Taylor (1999) assigned 101 undergraduates to simulate the process of studying for a midterm, the outcome of getting the grade, or neither. The process group reported clearer plans, put in more study hours and finished 6.60 points above the class average; the outcome group finished 1.96 points below it, a gap to the control condition that was only marginal. What is solid is the distance between the two simulations on the exam itself, and that distance is the load-bearing result.

Outside the lab: the job market

Outside the lab the pattern holds. Oettingen and Mayer (2002) followed graduates entering the job market: those who reported frequent positive fantasies about professional life sent out fewer applications, received fewer offers and earned lower salaries over the next two years. Positive expectations grounded in past performance predicted the opposite for offers and salary. Only one of the two tracks what you have actually done before.

Energy and mood over time

Kappes and Oettingen (2011) induced positive fantasies and measured lower energization as a drop in systolic blood pressure (Study 1, n = 164, with no follow-up). In a separate experiment, lower self-reported energization predicted less accomplishment a week later (Study 3, n = 40).

Oettingen, Mayer and Portnow (2016) found across four studies that positive fantasizing predicted more depressive symptoms at follow-up and, in the fourth study, partly through lower effort and lower academic success. None of this makes the people who do it foolish; the failure mode is invisible from the inside.

Across these studies the same association recurs: the more the finished result is enjoyed in advance, the less effort follows. Only the Kappes and Oettingen experiments induce the fantasy rather than measuring it.

Does manifestation work without action?

No. The effects that survive on goal attainment, rather than on task performance, come from plans. Gollwitzer and Sheeran (2006) reviewed 94 independent tests of if-then plans naming when, where and how you will act: d = .65 on goal attainment, over and above holding the goal.

Wang, Wang and Gai (2021) tested pairing the desired future with the obstacle in front of it and attaching such a plan: g = 0.336 across 24 effect sizes from 21 studies with 15,907 participants, corrected down by the authors to g = 0.242 after adjusting for publication bias. No study compares a plan against a picture head to head.

Two findings come closer to money. Hershfield and colleagues (2011) showed people a photorealistic age-progressed image of their own future self; those participants allocated 6.17% of pay to retirement against 4.41% for people shown their current face (n = 18 against n = 20), an effect the authors call preliminary.

Cheema and Bagchi (2011) found that external visual representations making a goal easier to picture increased effort and savings commitment, but mainly when people were already close to the goal, and that proximity condition is part of the finding.

Manifestation vs visualization vs affirmations: what is the difference?

Three different things, with three separate evidence bases.

• Manifestation is a belief system, measured so far only in its believers.
• Visualization is a rehearsal technique with meta-analytic support, d = 0.431 in the most recent pooling, where the process-versus-outcome split decides whether it helps.
• Affirmations are a spoken or written self-statement, and their evidence is thinner and mixed.

Affirmations in the self-help sense

In the self-help sense the best-known experiment is Wood, Perunovic and Lee (2009) (Study 1 survey N = 249, Study 2 N = 68, Study 3 N = 116): in Study 2, people with low self-esteem who repeated “I am a lovable person” felt worse on a composite of mood and state self-esteem than controls, while people with high self-esteem gained only a little.

Flynn and Bordieri (2020), in two experiments with university samples of approximately 225 and 237, then found no differences between affirmation and control conditions on mood or state self-esteem, and, in the second experiment, none on goal completion either. Their own phrase for it: “a failure to replicate”. No pooled effect size for self-help affirmations exists, so none is printed here, and both papers measured students' mood, not money.

Self-affirmation in research

Self-affirmation in research means something else: writing about a personal value before receiving unwelcome information. It has small but reliable effects on health behaviour (Epton et al., 2015): d+ = .17 for message acceptance across 34 tests, N = 3,433; .14 for intentions across 64 tests, N = 5,564; .32 for behaviour across 46 tests, N = 2,715, out of 144 tests in total. It shares a word with “I am wealthy” statements and nothing else.

The statistics that are not real

The three numbers most often quoted in favour of manifestation cannot be sourced, and the best known of them points to a study no librarian has been able to find.

• First, the Harvard/Yale goals study: 3% of graduates wrote down their goals and later earned ten times more than the rest. Harvard Library's reference answer (Ask a Librarian, FAQ 82314) states that its librarians have never found it and treat it as an urban legend, and that Yale's librarians have not found a trace of it either. A reference answer, not research, which is the point: there is no research to cite.
• Second, the written-goals percentages offered as the respectable replacement, which trace to Gail Matthews' goal-setting research. By the same reference answer's account those results, as far as Harvard's librarians know, were never published, so the figures are not reproduced here.
• Third, the sentence “X% of people who visualize achieve their goals”. No source reports it, which is an absence rather than a counter-statistic of our own.

Vision boards as a format have not been tested. We found no controlled trial of them; the published material is case descriptions, single-group studies without controls, and media commentary. Science has not disproved vision boards; it has not examined them. The evidence belongs to imagery, process simulation and if-then planning, so whether a board does anything depends on what is on it.

What this means for a money goal

Nothing in this literature tests a target balance. Three ingredients have been tested, and they set the conditions under which looking at a number could do anything.

• First, something concrete to look at rather than a mood: the closest money-specific evidence is Hershfield's age-progressed future-self image, 6.17% of pay allocated to retirement against 4.41%, which its authors label preliminary and which tested a picture of a future self, not a target number.
• Second, the process and the obstacle attached to it rather than the finished result: Pham and Taylor's process condition, and the mental contrasting plus if-then plan pooled by Wang, Wang and Gai (g = 0.336, corrected to g = 0.242).
• Third, repetition, because the benefit decays as the gap between rehearsal and performance grows (Driskell, Copper and Moran, r = −.216).

On paper that is four lines:

• the number and the date (€12,000 by 1 June);
• the step that moves it (transfer on payday);
• the obstacle that reliably stops you (an unplanned online order that evening);
• and the if-then sentence that meets it (if I open a shop after 9pm, then I close the tab and move the transfer first).

The limits stay attached. The meta-analytic effects are modest, the Dixon findings are correlational, and no controlled study has tested a board as a format, so nothing here is evidence that a dashboard works: what has been studied are its ingredients.

That is the line a product has to respect. A target number works as a cue for the plan behind it, and stops working the moment it becomes a scene you enjoy having already reached; used the second way, the studies above say it costs you effort.

WealthNow is a vision board with real numbers — a goal dashboard where the balance you're working toward becomes a five-second daily ritual. Visualization, not affirmations.

The free version shows a fixed example balance so you can feel what the daily ritual is like; setting your own number is Pro.

FAQ

Is manifestation scientifically proven?

Manifestation as a whole has never been tested. One of its components has: mental rehearsal of a specific process has meta-analytic support (Simonsmeier et al., 2021: d = 0.431, 95% CI [0.298, 0.563]; Driskell et al., 1994: r = .255 against r = .364 for physical practice). The belief itself has one peer-reviewed study of its believers (Dixon et al., 2023), which found no relationship with income or education attainment.

Does manifesting money work?

There is no evidence that it produces money. The one finding that touches money is Hershfield et al. (2011): people shown an age-progressed image of their own future self allocated 6.17% of pay to retirement against 4.41%, an effect the authors call preliminary. Dixon et al. (2023) found 1.33 times higher odds of holding cryptocurrency and 1.42 times higher odds of reported bankruptcy per one-point increase on the belief scale — correlational, on a paid US panel, so it cannot say the belief caused either.

How long before visualization has an effect?

No study gives a date. The benefit shrinks as the interval between rehearsal and performance grows (Driskell et al., 1994: r = −.216), which favours short repeated rehearsal over one long session, and imagery effects stack on physical practice rather than replacing it, though that added increment is small (Simonsmeier et al., 2021: d = 0.156, 95% CI [0.002, 0.310]).