# (Bloom, Draca, and Van Reenen, 2016)
---
* **Title:** Trade Induced Technical Change? The Impact of Chinese Imports on Innovation, IT and Productivity
* **Authors:** Bloom, N., M. Draca, and J. Van Reenen
* **Journal:** Quarterly Journal of Economics
* **Year, Volume, Issue, Pages:** 2016, 83, 87-117
* **Summary Created On:** April 3, 2020
###### tags: `papersummary` `innovation` `trade` `macro` `readinggroup`
---
## Thoughts
* This paper deals with reallocation of labor--does it comment on training/re-skilling of the workforce?
* How does this paper square with Autor et al's work on the China shock?
* Whether firms upgrade technology must be measured on the sample of surviving firms, right?
* Interestingly, the paper tries to argue that quota abolition came as a surprise to firms because there was uncertainty over whether liberalization would actually occur or not... hmm

## Question
* What is the impact of globalization on the economies of developed countries?
* Focus here is on technology measured as patenting, TFP, and adoption of IT
* Fundamental Schumpeterian force: competition lowers price-cost margins, thus reducing quasi-rents from innovation,
* But reduction in trade costs also increases the size of market, so the effect of competition on innovation incentives is ambiguous

## Results
1. Reduction in trade costs with a low-wage country like China increases innovation: firms expand patenting activity (so the increase in TFP is not just coming from changes in product composition or off-shoring to China)
* Increase in technology also means that employment of low-skilled workers falls, and wage inequality (in favor of skilled workers) increases
* A 10% increase in Chinese import penetration is associated with a 3.2% increase in patenting (without controlling for firm-level employment, which is endogenous). This is not just caused by firms patenting more to protect existing technologies--citation counts increase as well (same threat?)
3. Firms that were more exposed to Chinese imports upgraded their technology more (had more patents, increased R&D spending, raised IT intensity, and management quality improved). This is the *within firm effect*
4. Firms that were low technology had a higher probability of exit. This is the *between firm effect*--> reallocation of factors across firms
* This is the channel that Melitz (2003) and other related papers focus on
6. Both effects were equally important

**General Equilibrium Consequences**
* Eq 4.6 --> a model of aggregate increases in productivity, decomposed between within-firm and between-firm terms
* Between 2000-2007, Chinese imports accounted for 13.9% of the increase in aggregate patenting per worker--> within-firm component accounts for 5.1%, and between-firm for 6.7% and rest due to exit (2.7%)

**Dynamic Selection Bias**
* It may be that firms that knew they were growing technologically would have decided to not exit...I don't see why this is a problem

## Data and Identification
* Countries: 12 European countries (Austria, Denmark, Finland, France, Germany, Ireland, Italy, Norway, Spain, Sweden, Switzerland, UK)
* The entry of China into the World Trade Organization in 2001 and the subsequent erasure of trade barriers within Clothing and Textile industry
* Exposure to Chinese imports is measured as ratio of imports from China as a fraction of imports from all over the world (within 4-dig ind) to a country
* Exclusion restriction: shocks to technology are uncorrelated with changes in trade barriers/ quotas. Quotas are measured as the value-weighted proportion of products in the industry that are covered by a quota against China
* 
* 

* Equation 3.3 is important (not added here)--calculates survival probability of firms, but has an interaction term of Chinese imports with a dummy for high tech firms (tagged as being high-tech at baseline) since the latter are probably less affecte
* Clustering at the country-industry paid level, firm-level time trends (to account for pre-trends in the growth of technology)
* Obviously no threat from simultaneity, and I am not very concerned about common shocks here
* **Datasets:**
* Amadeus (universe of public and private firms in Europe)
* Population of patent applications to the European Patent Office (matching by name...good god)
* UN Comtrade Data: international database of 6-digit product level information on all bilateral imports and exports between any given pair of countries

## Theory (Main Model)

## Facts
* In 2000, 5.7% of all imports to these 12 countries (and the US) originated in China. By 2007, this had doubled to 12.7%
* Industries most affected were toys, furniture, footwear, and textile