### Evolution of thought

* the ideas developed in recent decades
* Boskin commission in 1996
* landmark event for US price measurement:
* Concluded CPI converstated inflation primarily due to:
* Various forms of substitution bias (including outlet substitution)
* New goods bias/quality change

* Laspeyres index (initial prices as weights):
$$ \frac{p_{a1}q_{a2} + p_{e1}q_{e2}}{p_{a1}q_{a1} + p_{e1}q_{e1}}
$$
* Paasche index (final prices as weights):

$$ \frac{p_{a2}q_{a2} + p_{e2}q_{e2}}{p_{a2}q_{a1} + p_{e2}q_{e1}}
$$

### New Goods & Quality Change
* How should we deal with new goods?
* No comparison possible with earlier period since good didn’t exist and so no old price/quantity exist
* Prices of most goods not measured until well after they come into existence (often steep fall to begin with)

#### Case Study : Light
* Progression from open fires to candles to incandescent light blubs to compact fluorescent bukbs and LED lights
* HOW to measure "price" and "quantity" of light?

| | Incandescent | Fluorescent |
| -------- | -------- | -------- |
| P_1 | $10 | -- |
| P_2 | $12 | $24 |
| P_3 | -- | $18 |

Solution:
$$
\frac{P_3}{P_1} = \frac{P_{2, old}}{P_{1, old}} \times \frac{P_{3, new}}{P_{2, new}} = \frac{12}{10} \times \frac{18}{24}
$$

| | Incandescent | Fluorescent |
| -------- | -------- | -------- |
| P_1 | $10 | -- |
| P_2 | $12 | $2.4 |
| P_3 | -- | $1.8 |

### Hedonics
Basic Idea: Directly adjust for quality change between new and old good
• E.g. Nordhaus’ approach of analyzing prices in “per lumen” terms is a hedonic adjustment
• Challenging: Must define which characteristics are important to consumers (and how much they matter)
• e.g. what to include in comparing iPhone 8 and iPhone 14?
• What about Apple iPhone vs old Nokia cell?
• Much more model-dependent (and less transparent) than standard price index formulas
• Also very labor-intensive to specify and estimate hedonic models
• As a consequence, the CPI (and other national inflation statistics) mostly “link in” new goods
• Use hedonics sparingly in some product categories

### East vs West Germany in 1989

* pre-1989 estimates: East German GDP per person 75-85% of West German
* post-1989: East German GDP clearly much smaller relative to West Germany than originally thought

What explains this?

### International Comparisons
* PPP
* Summers and Heston invented
* Taken over by World Bank
* Market comparisons
* Convert each ountry's GDP into US$$
* PPP adjusted GDP imply most developing countries have ***larger*** economies than standard conomies than standard market-based measures suggest.

### This Class:
**Making sense of macro events, e.g.,**
* What caused the Great Depression?
* What happened during the Great Recession?

**What is the effect of macro policy?**
* What is the effect of increasing the money supply?
* What is the effect of higher government spending?

**What role for macro policy?**

### Two Approaches:
1. Structural Approach
* Build an complex Dynamic Stochastic General Quilibrium (DSGE) model of the economy
* Pros: Can tell you quantitatively
* The causal effect of B on A
* the different causes of (say) the Great Depression
* what macro policy should have done
* Cons: Economy very complex and model may be wrong (risk of mis-specification)

2. “Reduced-form” approach
a) Small model to isolate mechanisms and formulate hypotheses, e.g., IS/LM model
b) Careful empirical study to tease out cause and effects
* Pros: Transparent, more robust to model mis-specification
* Cons: Lacks encompassing framework and may give incomplete answers

Difference in group = Average casual effect + selection bias

### OVB
* "selection bias" also called "OVB"
* Get the rid of selection bias
* Randomized trial