6 July 2017

# Background & Motivation

• The share of foreign born in the U.S. has seen a sharp increase from 1970 to 2010

# Background & Motivation

• There is also substantial variation in the diversity and intensity of migrant inflow across counties

• How do these settlement patterns shape local communities and preferences?

• The different consumption and food culture across countries / regions provide an opportunity to test the spillover effects of migrant culture

• Observationally there seem to be a link between the popularity of certain types of cuisine and a city’s main migrant group

# Research question

• What is the effect of migration on local household’s food consumption expenditure?

### Literature review

• A substantial amount of work has been done to study the effect of migration on wages(Card 1990; Borjas 2003; S. P. Kerr and Kerr 2011), but analysis on the broader multidimensional effects of migration is more limited

• Existing work on cultural diversity tends to focus on its economic value measured through outcomes on wages and housing (Ottaviano and Peri 2006; Bellini et al. 2013) or innovation (W. R. Kerr and Lincoln 2010; Hunt and Gauthier-Loiselle 2010)

• Related to Bronnenberg, Dubé, and Gentzkow (2012) who show that the preferences of interstate migrants within the US over consumer packaged goods converge slowly to native preferences

### Key contributions

• This paper extends the literature on the economics of migration and culture by studying a new outcome measure (household food consumption)

• Methodologically, this papwe applies text classification methods on the product attributes field of the Nielsen consumer panel dataset to construct a measure of foreign-linked expenditure

# Model

### Consumer

CES, Dixit and Stiglitz (1977) type model

\begin{align} U_{i} & = C^{\alpha_{i}}_{i}F^{1-\alpha_{i}}_{i} \\ F_{i} &= \left( \sum_{j=1}^{N} \beta_j^{\frac{1}{\sigma}} x_j^{\frac{\sigma -1 }{\sigma}} \right) ^\frac{\sigma}{\sigma - 1} \end{align} Relative demand: $\frac{x_j}{x_k} = \frac{\beta_j}{\beta_k} \left( \frac{p_k}{p_j} \right)^\sigma$

Summing over all products we obtain: $m = \sum_{j=1}^{N} p_j x_j = \left( \sum_{j=1}^{N} \frac{\beta_j}{\beta_k} p_j^{1-\sigma}\right) p_k^\sigma x_k$

Marshallian demand: \begin{align} x_k &= \frac{m p_k^{-\sigma}}{\sum_{j=1}^{N} \frac{\beta_j}{\beta_k}p_j^{1-\sigma}} \\ &= \frac{\beta_k m}{p_k^\sigma} \left( \sum_{j=1}^{N} \beta_j p_j^{1-\sigma} \right)^{-1} \\ &= \frac{\beta_k m}{P} \left( \frac{P}{p_k}\right)^\sigma \end{align} Expenditure share: $s_j = \beta_j \left( \frac{P}{p_j}\right)^{\sigma-1}$
Holding price constant, an increase in $$\beta$$ increases the associated expenditure share.

### Producer

Assume single product firms engaging in Bertrand price competition

Profit function:

$\pi_i = p_i x_i - c_i x_i - F_i$

F.O.C

\begin{align} 0 &= x_i + p_i \frac{\partial x_i}{\partial p_i} - c_i \frac{\partial x_i}{\partial p_i} \\ c_i &= \left( 1 + \frac{1}{\varepsilon_d}\right)p_i \\ &= \left( 1 - \frac{1}{\sigma}\right)p_i \end{align}

Product $$i$$ is imported if profit is higher than fixed cost

$(p_i - c_i)x_i \geq F_i$

Equilibrium (assuming identical consumers):

$\frac{F_i}{p_i -c} = \frac{\beta_i m}{P} \left( \frac{P}{p_i}\right)^\sigma$

Challenges in identifying changes in consumer’s preferences:

• Variety of goods is endogenous

• Hard to distinguish between preference and cost channels

Nevertheless, an increase in migrant flow is theorised to lead to an increase in consumption share

# Data

• Nielsen consumer panel dataset
• Detailed information on household shopping trips, purchases and amount spent
• Used 2011 cross-section containing 62,000 households. Focus on white, non-hispanic, expenditure in 8 categories:
• canned seafood, canned vegetables, dry mix, frozen, pasta, ready-to-serve, sauces, spices
• Merged with products attributes dataset
• Recipe corpus
• Scraped 6275 recipes with 25 different geographical region tags from allrecipe.com
• Combined with data from 165 recipe books’ indexes.
• Foreign born data
• Used 5-year 2006-2010 ACS and 1980 census data from NHGIS
• Additional robustness test conducted over choice of dataset

# Data

• 65,600 unique UPC codes in the initial dataset

• Merged UPC codes to type description and product description fields depending on which provide more information to create a vector of search text

• Excluded the following categories: misc. sauces, sandwich spreads meat, extracts, vegetables onion instant, multipack, soup

• Product description fields are quite sparse and contain basic information on product characteristic / flavour e.g. “mahi mahi mango marinaded”, “fajita seasoning”, “beans”

• Search text is subsequently merged with the TF-IDF dataset to create geographical region weights for 47,000 UPC codes

• Search text is merged to two versions of the recipe TF-IDF scores, single word and word pairs

• The word pair score is used if such a merge exists, otherwise the single word match with the highest maximum score is used e.g. “fajita seasoning” would be given the TF-IDF score of “fajita seasoning” if it exists, else it would be assigned the maximum of “fajita” and “seasoning”

• Implemented a bag-of-words model commonly used in document classification problems

• Similar studies using text classification include Antweiler and Frank (2004) and Gentzkow and Shapiro (2010)

• Term frequency - inverse document frequency (TF-IDF) method:
• Use the relative frequency of words/ingredients for a particular region as a signal of importance
• “Normalise” the term frequencies by the inverse document frequency to down-weight common words

$TF\text{-}IDF_{t,d} =\frac{f_{t,d}}{\sum_{t'\in d}f_{t',d}} \cdot log \frac{N}{n_{t}+1}$

• where $$f_{t,d}$$ is the frequency in which term $$t$$ appears in document $$d$$

• $$N$$ is the total number of documents in the corpus and $$n_{t}$$ is the total number of documents where term $$t$$ is found

• The TF-IDF approach is commonly used in document classification problems and usually out performs multinomial naive bayes (Kibriya et al. 2004)

# From TF-IDF scores to weights

• Calculate TF-IDF scores for recipe and book dataset ($$\rho$$=0.77) and take the simple average
• Match the TF-IDF scores to the product attribute dataset and normalise scores across each product
• Use the weights to calculate linked-expenditure shares
• Sample of products

# Empricial Strategy

$Exp\_share_{ij} = \rho FB\_share_{ij} + \mu ln\_dist_{ij} + \lambda_{i} + \gamma_{j} +\epsilon_{ij}$

• Control for unobserved variation at the county($$i$$) and geographical-region($$j$$) level as well as log great circle distance between county and region.

• Omitted variable problem not a main concern. While many factors affect foreign-born share, few, if any are correlated with local food expenditure share

• Potential “peer effects” endogeneity concerns(Angrist 2014) e.g. similar preferences between parent and child of common ancestry

• Solution: restrict the foreign-born share sample to non-European countries

# Empricial Strategy

• Measurement error is still a large source of concern

• Use 1980 foreign-born share as an instrument

• Foreign enclave idea (Card 2001; Ottaviano and Peri 2006). Exploit persistent migrant flow patterns that are plausibly orthogonal to consumption

• Furthermore, the time period also coincides with the liberalisation of U.S. migration policy

• The Immigration and Naturalization Act of 1965 replaced the national origins quota with a category system

• Distribution of the initial wave of migrant plausibly exogenous to consumption

• Policy change resulted in mass migration primarily from Central America and Asia (Chiswick and Hatton 2003; Abramitzky and Boustan 2016)

# Results (Baseline)

 Dependent variable: Exp Share (1) (2) (3) (4) (5) (6) 2010 FB Share 0.1619*** 0.0011 0.3312 0.0791 0.1764*** -0.0490 (0.0558) (0.0174) (0.2501) (0.1480) (0.0593) (0.0365) Sample Full Asia Full Asia Full Asia Geographical region FE X X X X County FE X X X X Observations 796,448 448,002 796,448 448,002 796,448 448,002 Notes: Two-way standard errors clustered by county and geographical region in parentheses. Full sample consists of 18 regions, Asia sample consists of 9 regions. *p<0.1; **p<0.05; ***p<0.01

# Results (IV)

 Dependent variable: Exp Share 2010 FB Share Exp Share 2010 FB Share Exp Share OLS OLS Subset First Stage IV First Stage IV (1) (2) (3) (4) (5) (6) 2010 FB Share 0.1764*** 0.2004*** (0.0593) (0.0426) 1980 FB Share 1.8307*** 2.1091*** (0.1778) (0.2753) Fitted 2010 FB Share 0.2762*** -0.1337** (0.0480) (0.0669) First Stage F-stat 105.99 58.71 Sample Full Full Full Full Asia Asia Geographical region FE X X X X X X County FE X X X X X X Observations 796,448 595,920 595,932 595,920 248,305 248,300 Notes: Two-way standard errors clustered by county and geographical region in parentheses. OLS Full sample consists of 18 regions, OLS Subset and IV Full sample consists of 12 regions, IV Asia sample consists of 5 regions. The decrease in the number of observations in the IV sample is due to data availability limitations of the 1980 NHGIS dataset. *p<0.1; **p<0.05; ***p<0.01

# Results by product category

 Dependent variable: Category Exp Share IV IV IV IV IV IV IV IV (1) (2) (3) (4) (5) (6) (7) (8) Fitted 2010 FB Share 0.3246*** -0.0835*** 0.5860*** -0.2106 0.1438 -0.1101** 0.0029 0.0205*** (0.0559) (0.0322) (0.1748) (0.2660) (0.1364) (0.0482) (0.0204) (0.0026) Sample Full Asia Full Asia Full Asia Full Asia Product category Frozen Frozen Ready-to-serve Ready-to-serve Sauces Sauces Spice Spice Geographical region FE X X X X X X X X County FE X X X X X X X X Observations 564,672 235,280 557,076 232,115 586,908 244,545 539,172 224,655 Notes: Two-way standard errors clustered by county and geographical region in parentheses. IV Full sample consists of 12 regions, IV Asia sample consists of 5 regions. The decrease in the number of observations in the IV sample is due to data availability limitations of the 1980 NHGIS dataset. *p<0.1; **p<0.05; ***p<0.01

# Robustness (measures)

 Dependent variable: Exp Share Exp Share Exp Share IV IV IV IV IV IV (1) (2) (3) (4) (5) (6) Fitted 2010 FB Share 0.2762*** -0.1337** 0.2854*** -0.1165 0.2818*** -0.1457** (0.0480) (0.0669) (0.0484) (0.0884) (0.0527) (0.0593) Sample Full Asia Full Asia Full Asia Expenditure measure Wtd. Avg Wtd. Avg Majority Majority High Majority High Majority Geographical region FE X X X X X X County FE X X X X X X Observations 595,920 248,300 595,920 248,300 595,920 248,300 Notes: Two-way standard errors clustered by county and geographical region in parentheses. IV Full sample consists of 12 regions, IV Asia sample consists of 5 regions. Expenditure measure constructed by weighted average (Wtd. Avg) or with weight=1 for the region with highest weight (Majority) or with weight=1 for the region with highest weight conditional on the initial weight being greater than 0.5 (High Majority) *p<0.1; **p<0.05; ***p<0.01

# Robustness (datasets)

 Dependent variable: Exp Share IV IV IV IV IV IV (1) (2) (3) (4) (5) (6) Fitted 2010 FB Share 0.5892*** -0.3225** 0.5320* -0.4156*** 0.9673** 0.2032 (0.1372) (0.1418) (0.2774) (0.1196) (0.3940) (0.3453) Sample Full Asia Full Asia Full Asia Dataset used NHGIS 1980 NHGIS 1980 IPUMS 1970 IPUMS 1970 NHGIS 1970 NHGIS 1970 Geographical region FE X X X X X X County FE X X X X X X Observations 595,920 248,300 202,244 101,122 496,550 148,965 Notes: Two-way standard errors clustered by county and geograpical region in parentheses. Baseline NHGIS 1980 IV Full sample consists of 12 regions, IV Asia sample consists of 5 regions. IPUMS 1970 IV Full sample consists of 14 regions, IV Asia sample consists of 7 regions. NHGIS 1970 IV Full sample consists of 10 regions, IV Asia sample consists of 3 regions. *p<0.1; **p<0.05; ***p<0.01

# Heterogeneity (migrant dissimilarity index)

 Dependent variable: Exp Share OLS OLS IV IV (1) (2) (3) (4) 2010 FB Share 0.1756*** -0.0549 (0.0589) (0.0373) Foreign Dis x 2010 FB Share -0.0322 (0.0408) Asia Dis x 2010 FB Share -0.0130 (0.0599) Fitted 2010 FB Share 0.2748*** -0.2170* (0.0492) (0.1179) Fitted Foreign Dis x 2010 FB Share -0.0126 (0.0744) Fitted Asia Dis x 2010 FB Share -0.1494 (0.1499) Sample Full Asia Full Asia Geographical region FE X X X X County FE X X X X Observations 796,256 444,537 595,776 246,375 Notes: Two-way standard errors clustered by county and geographical region in parentheses. OLS Full sample consists of 18 regions, IV Full sample consists of 12 regions, IV Asia sample consists of 5 regions. The decrease in the number of observations in the IV sample is due to data availability limitations of the 1980 NHGIS dataset. *p<0.1; **p<0.05; ***p<0.01

# Heterogeneity (age)

 Dependent variable: Exp Share IV IV IV IV (1) (2) (3) (4) Fitted 2010 FB Share 0.2342*** 0.3086*** -0.1619 -0.1218* (0.0498) (0.0531) (0.1006) (0.0712) Sample Full, Young Full, Old Asia, Young Asia, Old Geographical region FE X X X X County FE X X X X Observations 121,404 474,516 50,585 197,715 Notes: Two-way standard errors clustered by county and geographical region in parentheses. IV Full sample consists of 12 regions, IV Asia sample consists of 5 regions. The decrease in the number of observations in the IV sample is due to data availability limitations of the 1980 NHGIS dataset. Young refers to households where the age of male or female head of household is below 45. Old is the complement group. *p<0.1; **p<0.05; ***p<0.01

# Discussion

• The greater the number of Asian migrants to a particular county, the less likely it is for a native to purchase related supermarket goods

• By contrast, there is a positive relationship between non-Asian migrantion and consumption share

• These results are consistent across OLS and IV specifications and are robust to various methods of constructing expenditure shares and regional aggregations

• The positive relationship between share of foreign born of non-Asian origin and local expenditure may be due to the peer effect endogeneity problem

• Potential violation of the exclusion restriction: the recent migration wave also significantly expanded the availability and quality of outside food options

• Cross-substitution patterns could result in households substituting home-cooked food for outside consumption or perhaps there is a higher standard of what constitutes authentic Asian food and supermarket produce may not make the standard

# Appendix (TF-IDF geographical region correlation matrix)

 Africa Caribbean Central_America China Eastern_Europe France Germany Greece India Italy Japan Korea Middle_East Other_Southeast_Asia Philippines Scandinavia South_America Spain Thailand Vietnam Africa 1 0.903 0.733 0.671 0.725 0.750 0.629 0.813 0.842 0.712 0.639 0.682 0.897 0.761 0.842 0.431 0.910 0.862 0.607 0.587 Caribbean 0.903 1 0.760 0.695 0.688 0.733 0.598 0.827 0.810 0.700 0.700 0.724 0.859 0.803 0.876 0.407 0.912 0.873 0.757 0.706 Central_America 0.733 0.760 1 0.555 0.576 0.638 0.457 0.693 0.734 0.705 0.540 0.558 0.658 0.623 0.692 0.308 0.729 0.655 0.603 0.623 China 0.671 0.695 0.555 1 0.539 0.582 0.496 0.552 0.649 0.539 0.837 0.820 0.663 0.801 0.808 0.358 0.671 0.643 0.752 0.780 Eastern_Europe 0.725 0.688 0.576 0.539 1 0.874 0.920 0.632 0.605 0.579 0.628 0.481 0.695 0.565 0.707 0.861 0.797 0.612 0.469 0.502 France 0.750 0.733 0.638 0.582 0.874 1 0.817 0.719 0.622 0.726 0.583 0.539 0.741 0.623 0.707 0.756 0.822 0.685 0.559 0.577 Germany 0.629 0.598 0.457 0.496 0.920 0.817 1 0.531 0.499 0.456 0.592 0.469 0.605 0.524 0.630 0.930 0.713 0.503 0.425 0.462 Greece 0.813 0.827 0.693 0.552 0.632 0.719 0.531 1 0.638 0.796 0.559 0.572 0.822 0.654 0.733 0.371 0.808 0.829 0.578 0.590 India 0.842 0.810 0.734 0.649 0.605 0.622 0.499 0.638 1 0.623 0.571 0.615 0.791 0.781 0.740 0.329 0.785 0.703 0.646 0.588 Italy 0.712 0.700 0.705 0.539 0.579 0.726 0.456 0.796 0.623 1 0.503 0.535 0.698 0.592 0.668 0.334 0.740 0.684 0.518 0.608 Japan 0.639 0.700 0.540 0.837 0.628 0.583 0.592 0.559 0.571 0.503 1 0.814 0.608 0.769 0.784 0.482 0.639 0.647 0.731 0.736 Korea 0.682 0.724 0.558 0.820 0.481 0.539 0.469 0.572 0.615 0.535 0.814 1 0.653 0.738 0.778 0.331 0.717 0.747 0.701 0.721 Middle_East 0.897 0.859 0.658 0.663 0.695 0.741 0.605 0.822 0.791 0.698 0.608 0.653 1 0.717 0.811 0.415 0.869 0.868 0.606 0.625 Other_Southeast_Asia 0.761 0.803 0.623 0.801 0.565 0.623 0.524 0.654 0.781 0.592 0.769 0.738 0.717 1 0.850 0.371 0.764 0.697 0.809 0.750 Philippines 0.842 0.876 0.692 0.808 0.707 0.707 0.630 0.733 0.740 0.668 0.784 0.778 0.811 0.850 1 0.456 0.865 0.782 0.759 0.738 Scandinavia 0.431 0.407 0.308 0.358 0.861 0.756 0.930 0.371 0.329 0.334 0.482 0.331 0.415 0.371 0.456 1 0.548 0.298 0.304 0.349 South_America 0.910 0.912 0.729 0.671 0.797 0.822 0.713 0.808 0.785 0.740 0.639 0.717 0.869 0.764 0.865 0.548 1 0.840 0.656 0.624 Spain 0.862 0.873 0.655 0.643 0.612 0.685 0.503 0.829 0.703 0.684 0.647 0.747 0.868 0.697 0.782 0.298 0.840 1 0.611 0.574 Thailand 0.607 0.757 0.603 0.752 0.469 0.559 0.425 0.578 0.646 0.518 0.731 0.701 0.606 0.809 0.759 0.304 0.656 0.611 1 0.839 Vietnam 0.587 0.706 0.623 0.780 0.502 0.577 0.462 0.590 0.588 0.608 0.736 0.721 0.625 0.750 0.738 0.349 0.624 0.574 0.839 1

# Appendix (household summary statistics)

 N Mean SD Min Q25 Median Q75 Max household_total_exp 50,123 328.91 236.16 0 164.03 277.84 435.17 4,948 Africa 50,120 0.01 0.01 0 0.01 0.01 0.01 0.47 Caribbean 50,120 0.02 0.01 0 0.02 0.02 0.03 0.54 Central_America 50,120 0.09 0.06 0 0.04 0.07 0.11 1 China 50,120 0.02 0.02 0 0.01 0.02 0.03 0.53 Eastern_Europe 50,120 0.02 0.01 0 0.01 0.02 0.02 0.59 France 50,120 0.02 0.01 0 0.01 0.01 0.02 0.60 Germany 50,120 0.02 0.01 0 0.01 0.01 0.02 0.47 Greece 50,120 0.02 0.02 0 0.01 0.02 0.03 0.30 India 50,120 0.02 0.02 0 0.01 0.01 0.02 0.54 Italy 50,120 0.12 0.07 0 0.07 0.10 0.15 0.96 Japan 50,120 0.03 0.03 0 0.01 0.02 0.03 0.74 Korea 50,120 0.02 0.02 0 0.01 0.01 0.02 0.56 Middle_East 50,120 0.02 0.01 0 0.01 0.01 0.02 0.53 Other_Southeast_Asia 50,120 0.01 0.01 0 0.003 0.01 0.01 0.37 Philippines 50,120 0.02 0.01 0 0.01 0.02 0.02 0.34 Scandinavia 50,120 0.02 0.02 0 0.01 0.01 0.02 0.38 South_America 50,120 0.01 0.01 0 0.01 0.01 0.02 0.42 Spain 50,120 0.02 0.02 0 0.01 0.02 0.03 0.50 Thailand 50,120 0.01 0.01 0 0.01 0.01 0.02 0.92 Vietnam 50,120 0.01 0.01 0 0.001 0.003 0.01 0.41

# Appendix (sample of products)

 Top products by weight Africa Caribbean Central_America China Eastern_Europe France Germany Greece India Italy Japan Korea Middle_East Other_Southeast_Asia Philippines Scandinavia South_America Spain Thailand Vietnam Score < = 1 mild moroccan fish navy bean pinto bean spicy kung bowl beef cholent kugel kishka watermelon rind buttery rice rotini tomato basil kidney bean gnocchi potato sushi wrap rice bowl barbecued beef pork bean rice peanut satay sauce beef steak pepper swedish cream peruvian bean spanish style rice thailand fragrant rice dragon roll roasted split turkey breast baked navy bean chile tamale spicy mongolian bowl string bean chestnut puree assorted grape leaves rice pilaf manicotti umeboshi plum braised beef chili pinto bean rice satay sauce beef steak dinner rock roll berry roll vino seco wine pork brains canned fragrant rice dragon sauce spicy moroccan fish aromatic cheesy hashbrown spicy szechuan ramen string bean potato gratin potato sausage hash canned beef burger stew madras lentils gnocchi sushi ocean crab roll miso soybean paste savory bean rice chile pepper supreme sushi piece rutabaga white vino seco wine canary bean buffalo style vietnamese noodle Score < =0.5 yellow rice butterflied shrimp chilies hoisin sauce roast duck sauce herbs chervil flakes macaroni cayenne vermouth wine broiled steak seasoning rice soup powder mughlai kofta rice mild navratan kurma shrimp spring roll raisin gcmgbl medley santa style beef lobster rangoon tiger sauce spicy grass chili rice regular yellow rice island getaway seasoning rotisserie oven roast seasoning miso hoisin sauce smoked bacon spice herbs chervil leaves elbow macaroni chili cayenne powder beef portobello flame broiled cheese beef pork napa cabbage dumpling white beans mild potato spinach rice spring roll raisin crispies mongolian style beef seafood shrimp lobster newberg sauce fillo tiger seasoning grass rice lamb stew shrimp island lime oven chicken glaze rakkyo scallion sour cream chive potato salad herbs chervil shell macaroni mini whole grain pasta roast beef gravy flame broiled fajita chicken seasoned rotisserie great northern white beans mild cstnb spring roll wrap country style dijon mustard style beef lobster cake maine sticky rice chix noodle soup

# References

Abramitzky, Ran, and Leah Platt Boustan. 2016. “Immigration in American Economic History.” Journal of Economic Literature (Forthcoming).

Angrist, Joshua D. 2014. “The Perils of Peer Effects.” Labour Economics 30. Elsevier: 98–108.

Antweiler, Werner, and Murray Z Frank. 2004. “Is All That Talk Just Noise? The Information Content of Internet Stock Message Boards.” The Journal of Finance 59 (3). Wiley Online Library: 1259–94.

Bellini, Elena, Gianmarco IP Ottaviano, Dino Pinelli, and Giovanni Prarolo. 2013. “Cultural Diversity and Economic Performance: Evidence from European Regions.” In Geography, Institutions and Regional Economic Performance, 121–41. Springer.

Borjas, George J. 2003. “The Labor Demand Curve Is Downward Sloping: Reexamining the Impact of Immigration on the Labor Market.” The Quarterly Journal of Economics 118 (4). Oxford University Press: 1335–74.

Bronnenberg, Bart J., Jean-Pierre H. Dubé, and Matthew Gentzkow. 2012. “The Evolution of Brand Preferences: Evidence from Consumer Migration.” The American Economic Review 102 (6). American Economic Association: 2472–2508.

Card, David. 1990. “The Impact of the Mariel Boatlift on the Miami Labor Market.” ILR Review 43 (2). SAGE Publications Sage CA: Los Angeles, CA: 245–57.

———. 2001. “Immigrant Inflows, Native Outflows, and the Local Labor Market Impacts of Higher Immigration.” Journal of Labor Economics 19 (1). The University of Chicago Press: 22–64.

Chiswick, Barry, and Timothy J. Hatton. 2003. “International Migration and the Integration of Labor Markets.” In Globalization in Historical Perspective, edited by Michael D. Bordo, Alan M. Taylor, and Jeffrey G. Williamson, 65–120. Chicago: University of Chicago Press.

Dixit, Avinash K, and Joseph E Stiglitz. 1977. “Monopolistic Competition and Optimum Product Diversity.” The American Economic Review 67 (3). JSTOR: 297–308.

Gentzkow, Matthew, and Jesse M Shapiro. 2010. “What Drives Media Slant? Evidence from Us Daily Newspapers.” Econometrica 78 (1). Wiley Online Library: 35–71.

Hunt, Jennifer, and Marjolaine Gauthier-Loiselle. 2010. “How Much Does Immigration Boost Innovation?” American Economic Journal: Macroeconomics 2 (2). American Economic Association: 31–56.

Kerr, Sari Pekkala, and William R. Kerr. 2011. “Economic Impacts of Immigration: A Survey.” NBER Working Papers 16736. National Bureau of Economic Research, Inc. https://ideas.repec.org/p/nbr/nberwo/16736.html.

Kerr, William R, and William F Lincoln. 2010. “The Supply Side of Innovation: H-1B Visa Reforms and Us Ethnic Invention.” Journal of Labor Economics 28 (3). The University of Chicago Press: 473–508.

Kibriya, Ashraf M, Eibe Frank, Bernhard Pfahringer, and Geoffrey Holmes. 2004. “Multinomial Naive Bayes for Text Categorization Revisited.” In Australian Conference on Artificial Intelligence, 3339:488–99. Springer.

Ottaviano, Gianmarco IP, and Giovanni Peri. 2006. “The Economic Value of Cultural Diversity: Evidence from Us Cities.” Journal of Economic Geography 6 (1). Oxford Univ Press: 9–44.