Why is the social value curve NOT the supply curve shifted downwards?

Abbreviate 'Social Value' as SV. In some other resources such as this, the SV curve is known as the Social Marginal Benefit curve.

The following quotes explain [1.] why the SV curve is the demand curve shifted upwards, but they don't explain [2.] why the SV curve is NOT the supply curve shifted downwards.

Here's my rationale for [2.]: Assume $SV > 0$. Because SV benefits society, the suppliers of this SV can consider the SV supplied, as follows:
Suppliers' true cost $=$ their original costs $-$ SV, $\qquad$ for all quantities supplied.
So the new supply curve must underlie, and be parallel to, the old supply curve.

Source: pp 179-180, Modern Principles of Economics (2 ed, 2011) by Tyler Cowen, Alex Tabarrok

[See Figure 10.3.] The social value curve counts all the benefits of vaccine use, the private value plus the external benefits, so the efficient quantity is found where the social value curve intersects the supply curve.

A final way of illustrating the underuse of vaccines is to notice that if people who got a flu shot did receive all the benefits of vaccination, then their demand curve would shift upward by $20 and would be the same as the social value curve. ... The last way of thinking about the problem of external benefits also suggests one potential solution. If every time someone was vaccinated, they were given a subsidy of$\$20$, the monetary equivalent of the external benefit, they would demand the amount $Q_{Efficient}$. Recall from Chapter 6 that we can analyze a subsidy by shifting up the demand curve by the amount of the subsidy. Thus, in Figure 10.3, notice that a subsidy set equal to the level of the external benefit would shift the demand curve up and increase the market quantity from $Q_{Market}$ to $Q_{Efficient}$.

Footnote: I was reading p 201, Principles of Microeconomics (7 Ed, 2014) by N G Mankiw, when I encountered this problem. I don't quote Mankiw's book because it's brusquer than Cowen's above.

• I think you need to fill in a lot more details before this question is understandable. What is the 'social value' curve? The quote is about how subsidizing consumers shifts the marginal benefit curve, which is the demand curve, upward. You could also think about subsidizing producers to shift the marginal cost curve, or supply, downward. If the marginal subsidy is constant, the new supply curve will be parallel to the old. Maybe this is what you are talking about, but I don't understand your first two paragraphs. – NickJ May 25 '15 at 12:38
• en.wikipedia.org/wiki/Externality ? The question is as unclear as possible, I agree with @NickJ – VicAche May 25 '15 at 16:21
• @NickJ Sorry for the confusion. I edited my OP now. Better? To clarify, I already understand subsidizing producers to shift the marginal cost curve, or supply, downward. If the marginal subsidy is constant, the new supply curve will be parallel to the old. But instead, I'm asking why for positive externalities, the DEMAND curve shifts, and NOT the supply curve. – NNOX Apps May 26 '15 at 2:12
• @LawArea51Proposal-Commit positive externalities can affect both curves, depending on where the externalities appears. If the externality directly affects the producer, then it isn't an externality anymore. – VicAche May 26 '15 at 11:42
• @LawArea51Proposal-Commit which you can sum up as: externalities that shifts the supply curve are not externalities. – VicAche May 26 '15 at 11:43