The second paper in this series compressed a long history into one sentence. The nineteenth-century worker was primarily a saver, and the modern worker has been required to become an investor.
That sentence described a transition. It did not explain one, and the argument I am making now needs the explanation. So: why did the worker become an investor?
The conventional answers are many, and most of them are correct.
Retirements lengthened, and a longer horizon rewards growth assets, a conclusion lifecycle finance formalized when it showed how a young worker's large stock of human capital should shape the financial remainder. Capital markets deepened and democratized. Tax systems built shelters around long-term saving and attached them to investment accounts. Employers migrated from defined benefit promises to defined contribution accounts, handing the investment function to the individual whether or not the individual wanted it. Household wealth rose, and with it the expectation that retirement would be more than subsistence. Financial innovation made investing administratively trivial. Employment fragmented, so preservation could no longer ride on a single employer's promise.
Every one of those is a real cause. The transition does not reduce to monetary policy or to any single force, and I want that concession on the record before I add anything to the list.
The Fact Hiding in Plain Sight
There is something inside the transition that the conventional list does not reach, and it deserves to be stated as plainly as the rest.
Workers are paid in a medium that the entire apparatus of financial advice, plan design and preservation theory tells them not to hold for the long term.
The worker receives dollars. The preservation system then says, in effect: convert them. Convert them into equities, bonds, real estate, funds, pension claims, annuities, productive capital, other people's promises, anything but the dollars themselves, because dollars held across a long horizon lose purchasing capacity and the alternatives, on the historical record, tend not to.
The instruction is sound. It is the standard advice, it is supported by the long-run return evidence, and no responsible adviser would say otherwise.
That is the plainspoken version, and it is how the situation presents itself to a worker. The precise version, which I will use wherever precision matters, is that the worker is paid in a monetary claim that must either earn interest or be transformed into another instrument to preserve long-duration purchasing capacity. Whether interest alone suffices is a serious question and it gets its own article next week.
What is strange, on either formulation, is not the instruction. It is the structure that makes the instruction necessary.
The medium of compensation and the instruments of preservation are different things, and the worker has to cross from one to the other to keep what they were paid.
Stated as a proposition, because it carries weight through the rest of the series: modern monetary architecture separates the medium of compensation from the primary instruments of long-duration preservation.
That separation is a structural feature of the current arrangement rather than a scandal, and it explains part of the saver-to-investor migration that the conventional causes do not reach.
A Historical Comparison, Handled Honestly
The contrast between a saving past and an investing present is messier than it looks, and I want to draw its limits before anyone borrows it for a purpose it cannot serve.
Nineteenth-century monetary arrangements varied across gold, silver, bimetallic, note and deposit regimes. Banks failed and took savings with them. Short-run deflations and inflations were sharper than anything since. Coin carried an opportunity cost. Financial access was unequal by class and geography. And most workers had little surplus to preserve at all.
Nothing in that record supports the claim that historical workers enjoyed superior financial security or better money. Anyone using this argument to romanticize the past is using it wrongly, and I would rather say so than let the ambiguity work in my favor.
The narrow and defensible version is this. For some historical workers, in some regimes, saving and preservation were institutionally closer to the same act than they are for the modern worker inside a financialized defined contribution architecture. The coin in the savings bank was, however imperfectly and riskily, both the pay and the store.
The modern worker's saving and preserving are different acts, performed in different instruments, and the second requires competence the first did not.
That is all the comparison has to carry. Part of what made the worker an investor is that holding the pay was no longer enough.
The Strongest Objection
The best counterargument is not that the separation is imaginary. It is that the separation is efficient, and it should be stated at full strength before I go anywhere with it.
Money, on this view, is optimized for exchange and liquidity. Universal acceptance, divisibility, low transaction cost, stability over the short horizons in which payments clear. Productive capital is optimized for accumulation across long horizons, and it earns a return because it is scarce, risky and productive. Asking one instrument to perform both functions is asking it to be worse at each.
The separation is therefore a division of labor, and the worker's conversion of pay into capital is not a burden. It is the ordinary act of a saver placing resources where they earn what they should.
That objection has a distinguished lineage in the theory of money, which has often treated the store-of-value function as secondary and derivative. It may be right.
Nothing in my argument requires proving that money ought to preserve purchasing power. The argument needs only a weaker claim, and the weaker claim survives the objection intact.
Whether the separation is efficient or not, it is an architectural choice with consequences for an identifiable person. The worker has to acquire competence, accept market exposure, pay conversion costs, and manage behavior in order to keep what they were paid. Those consequences land inside the compensation relationship whether or not the arrangement that produces them is, on balance, the right one.
That is the distinction the whole paper rests on.
Efficiency is the economist's question, and I do not answer it.
Incidence is the compensation question, and I do.
Where the Boundary Sits
One limit on the incidence claim, drawn immediately, because the version without it would be a land grab.
I am not claiming that compensation professionals own every downstream financial decision a worker makes. That would be both wrong and unwelcome, and the first paper set the boundary deliberately: the employer is accountable for the architecture of its increment, never for the outcome of the whole journey.
The claim is narrower. Where compensation deliberately directs current economic value toward future consumption, the design of the available preservation architecture belongs inside the compensation-design boundary. The menu. The defaults. The ownership and custody terms. The portability. The pooling. The governance. The conversion path.
Individual allocation decisions inside that architecture remain with the worker. The architecture itself was designed by somebody, and that somebody usually works in compensation or benefits.
So here is the shape of the thing. We pay people in a unit. We then tell them, through every default and every piece of education material we produce, that holding the unit is not a plan. We hand them a menu for getting out of it. And we have never treated the properties of the unit itself as part of what we designed, because the unit is the water we swim in.
Next week: if workers have to convert compensation into other instruments to preserve it, then some of the risk they carry is carried to stand still rather than to get ahead. That distinction turns out to be harder to dismiss than it first appears, and a good economist's objection to it is the reason it becomes useful.