First, why does it *feel* wrong? A. the top 10% get about 50% of *pre-tax* income B. they get 30 to 37% of disposable income after tax&transfers (depending on source) 3. and we know the rich have a *higher* savings rate, so consume a *lesser share* of their income (US, China)

Taken together, these facts mean that the top 10% cannot be 50% of consumption. They do not even receive 40% of disposable income, and with consumption-income ratios *decreasing* in income, must account for < than 40% of consumption. So how does Moody's get to 50? Here is how:

They: 1. Compute (in the SCF) the share of asset i held by income group k 2. Multiply it by "net acquisition of i" in the quarterly Financial Accounts, Delta(A_i) 3. Sum across all assets & liabilities 4. Call this "savings" of k 5. Subtract from income of k to get "consumption"

Simplifying, this amounts to saying: - the top 10% (by income) own X% of wealth, and thus do *exactly* X% of all savings *every quarter*. - they receive X% of income - so they do Z% of consumption with: Z=(Y*national income-X*national savings)/(national income-national savings)

Using a national personal savings rate (from the Financial Accounts) of about 7%, this can be rearranged as: Z=(Y/0.93-X*0.07/0.93) But this is completely wrong (and by construction not a measure that can be released every quarter...).

For example, think of the excess savings after Covid. We *know* these were primarily a feature of lower-income households, suddenly receiving stimulus checks and saving way + than usual. Moody's methodology would by construction miss that, attributing most to high-income hhs.

Overall, all other sources suggest that: - the top 10% receive c. 35-40% of disposable income - their savings rate is c.20-25%, relative to a national of 7% - so they represent between: Z=(0.35*0.75)/(0.93)=28% and Z=(0.40*0.8)/(0.93)=35% of all consumption.

Overall, the 50% is: - an overestimate - extremely unlikely from just basic accounting relationships - a pretence of a quarterly number that is definitely not one

Might be of interest to @kingofthecoastt @M_C_Klein @BrianCAlbrecht @jmhorp @MattBruenig

@LevyAntoine Excellent analysis. Is the overall time trend of increasing consumption share of the top 10% income at least somewhat correct (despite its massive overestimate)?

@florianederer Best estimate is the BLS and I think the answer is mostly no. In the Moody’s imputation, it would be a feature of the assets that the rich own disproportionately seeing lower net acquisition rates (or conversely for liabilities) every quarter.

@LevyAntoine But you are aware that US personal income statistics DO NOT include even realized capital gains, which every year amount to close 8-9% of GDP and low almost exclusively to the top 10%?

@mnmntklphrsn Moody's defines consumption as the residual (after subtracting their "savings") from the personal disposable income measure from the BEA (along with a very strange attribution by the SCF, which does not have a proper disposable income concept, but that's a distinct problem).

@LevyAntoine @grok is this correct?

@LevyAntoine Billionaires are responsible for a million times more greenhouse gas emissions than the average person, according to Oxfam. & they are also actively destroying democracy.

@LevyAntoine Do the wealthy who do not work count as earners? If so, and they were included, would that change the graph?

@LevyAntoine Thanks for doing the analysis!

@LevyAntoine Unroll @threadreaderapp pls

@LevyAntoine These viral econ charts always sound dramatic-like a snare drum solo. Love a fact check before jumping on the bandwagon!

@LevyAntoine @VladTheInflator Also if stock markets fall , these 10% may cut back on negative wealth effect and this ripples through economy

@LevyAntoine There is a missing word here… “consumer”. This is not an analysis of all consumption, only a part, “consumer”, which is ill defined. Friedman used to stress empirical evidence. The top 10% do not consume 50% of healthcare, education, energy, housing, etc. Self evident.

@LevyAntoine “How to lie with statistics”.

@LevyAntoine Nice analysis digging in. Generally agree. But what about wealth effect on consumption. The fact that rich own more may mean they sell to consume or maybe their savings rates are going down?

@LevyAntoine If we included corporations as earners, how worth that change the graph? I only ask because I think we could get higher growth by abolishing all national taxes and only have a very progressive consumption tax. The national consumption profile is important to that assertion.

@LevyAntoine What about people who consume without earning, by borrowing on stock?

@LevyAntoine @Heminator Right; the assumptions in their formula lead to incorrect results. [I might have missed it; but charity wouldn’t be consumption and doesn’t seem to be captured by the formula at all.] But it is also not a helpful way of understanding spending. Exclude the top 1-2%, which are in

@LevyAntoine He never specified which country. Surely one can find among the 190 countries in the world one where the top 10% of "earners" constitute 50% of measured consumption.

@LevyAntoine nice explanation

@LevyAntoine Le BLS a fait le boulot, c'est 23.4%

@LevyAntoine The thing is they should account for 10% of consumption and anything else above it is a flaw in the system.

@LevyAntoine Is it top 10% earners or top 10% asset owners? Assets appreciation >>> Wage growth. Debt can be taken a low interest rates than Asset appreciation % and can be used to buy Yacht, beach houses, Highlands, private jets...etc.. Avoids many taxes. Top earners are losers. Great game.

@LevyAntoine Why do you think facts matter to Marxists?

@LevyAntoine @besttrousers The numbers don’t jive. Certainly consumption is a smaller part of income for the wealthy than for working Americans. The wealthy generate a ton of surplus passive income and for the most part are much less charitable than working class people. Get the wealthy to give more to ch