Exercise 2: Why the 2013 and 2021 Findings Aren't Actually Contradictory — Possible Solution
==================================================================================================

The two studies would only be genuinely contradictory if they were
measuring the exact same thing and reaching opposite conclusions about
it. They aren't - each one is answering a real, but different,
question, using a different real scope of what counts as an effect.

The real 2013 research specifically measured concentrated REGIONAL and
EMPLOYMENT effects: what happened to workers and local labor markets
in the specific US regions most directly exposed to competition from
Chinese manufacturing imports. Within that narrower, real scope, the
documented harm (higher unemployment, lower participation, reduced
wages) was genuine and real - nothing about the 2021 reassessment
retracts or disputes that this real, concentrated harm actually
happened to those specific people and places.

The real 2021 reassessment instead asked a broader, POPULATION-WIDE
question: accounting for a benefit the original research hadn't fully
incorporated (lower consumer prices from cheaper imported goods
benefiting essentially everyone, not just people in the exposed
regions), what share of the ENTIRE US population ended up net worse
off overall? That's a genuinely different, wider calculation - one
that can find only 6.3% of the total population experienced net harm,
while still leaving the real, concentrated regional damage from the
first study completely intact and true.

Both findings can be simultaneously accurate because "harm was real
and severe in specific exposed regions" and "when spread across the
whole population and offset by consumer-price benefits, most people
came out ahead overall" are answers to two different real questions,
not competing answers to the same one.

ANSWER: The two findings aren't contradictory because they answer
different real questions at different scopes - the 2013 research
measured genuine, concentrated regional/employment harm in specifically
exposed areas, while the 2021 reassessment measured net effects across
the entire US population once consumer-price benefits were included.
Both can be true simultaneously: real, severe local harm coexisting
with most of the wider population coming out ahead overall.

WHY THIS WORKS AS AN ANSWER
------------------------------
This identifies the real difference in scope (concentrated regional
harm vs. population-wide net effect) between the two studies, showing
why they're compatible rather than treating the later study as simply
overturning the earlier one.