Change LSR subtitle description to be about deficit

[MaxGhenis]

Budgetary impact is too vague, and in this case essentially incorrect:

This raises the budgetary impact of the reform by $3.5bn (from -$28.5bn to -$31.9bn, a 12% decrease).

Other scorekeepers generally discuss deficit impacts. So in this case I think we could say:

Labor supply responses increase the deficit impact of the reform by $3.5bn, from $28.5bn to $31.9bn (+12%).

Claude suggested these phrasings for the 2x2x2 of cases: positive/negative static estimate positive/negative dynamic estimate static <> dynamic

1a) Positive static estimate, positive dynamic estimate, static < dynamic: The reform has a static deficit impact estimate of $20bn. Considering labor supply responses, the dynamic deficit impact is $25bn, a 25% increase.

1b) Positive static estimate, positive dynamic estimate, static > dynamic: The reform has a static deficit impact estimate of $30bn. Considering labor supply responses, the dynamic deficit impact is $27bn, a 10% decrease.

2a) Negative static estimate, negative dynamic estimate, static < dynamic (absolute value): The reform has a static deficit impact estimate of -$15bn (i.e., a $15bn surplus). Considering labor supply responses, the dynamic deficit impact is -$12bn (i.e., a $12bn surplus), a 20% decrease in the surplus.

2b) Negative static estimate, negative dynamic estimate, static > dynamic (absolute value): The reform has a static deficit impact estimate of -$25bn (i.e., a $25bn surplus). Considering labor supply responses, the dynamic deficit impact is -$28bn (i.e., a $28bn surplus), a 12% increase in the surplus.

3a) Positive static estimate, negative dynamic estimate: The reform has a static deficit impact estimate of $10bn. Considering labor supply responses, the dynamic deficit impact is -$5bn (i.e., a $5bn surplus), a 150% decrease.

3b) Negative static estimate, positive dynamic estimate: The reform has a static deficit impact estimate of -$8bn (i.e., an $8bn surplus). Considering labor supply responses, the dynamic deficit impact is $2bn, a 125% increase.

3c) Positive static estimate, negative dynamic estimate, static < dynamic (absolute value): The reform has a static deficit impact estimate of $5bn. Considering labor supply responses, the dynamic deficit impact is -$10bn (i.e., a $10bn surplus), a 300% decrease. 3d) Negative static estimate, positive dynamic estimate, static < dynamic (absolute value): The reform has a static deficit impact estimate of -$12bn (i.e., a $12bn surplus). Considering labor supply responses, the dynamic deficit impact is $15bn, a 225% increase.

[anth-volk]

Would you like Claude's suggestions implemented in the fix for this, or a more limited implementation akin to:

Labor supply responses increase the deficit impact of the reform by $3.5bn, from $28.5bn to $31.9bn (+12%).

[anth-volk]

@MaxGhenis Hoping to get your input before Wednesday

[MaxGhenis]

Let's remove it entirely for now and return to this later: https://github.com/PolicyEngine/policyengine-app/issues/1829