VoxDev Development Economics

In 1978, a farmer in rural China could not sell a sack of grain above the state quota, choose an employer, or move to the nearest city without a permit.
In this week's VoxDev Talk, Kaiji Chen and Tao Zha (both Emory University and Atlanta Fed) explain how China dismantled barriers like this one at a time. Gradualist reform first gave farmers the right to sell surplus grain at market prices. Small state firms were privatised 19 years later, while the largest kept state backing. This style of reform powered two distinct growth engines: first, labour moving off the land, then a wave of capital into infrastructure and property. The result: half a century of unprecedented growth -- but it also produced the debt, inequality and trade tension now working against China's growth model.
The research behind this episode:
Chen, Kaiji, and Tao Zha. 2025. "China's Macroeconomic Development: The Role of Gradualist Reforms." Journal of Economic Literature 63 (4): 1331-62.
To cite this episode:
Phillips, Tim, Kaiji Chen, and Tao Zha. 2026. "Fifty years of Chinese growth: The gradualist reform strategy explained." VoxDev Talk (podcast).

About the guests

Kaiji Chen is Professor of Economics at Emory University and a research fellow at the Federal Reserve Bank of Atlanta's Center for Quantitative Economic Research. His research spans financial contracts, business cycles and China's macroeconomy, with recent work on housing policy, credit allocation and household consumption in China.
Tao Zha is the Samuel Candler Dobbs Professor of Economics at Emory University and executive director of the Center for Quantitative Economic Research at the Federal Reserve Bank of Atlanta. He is a research associate at the National Bureau of Economic Research and was elected a Fellow of the Econometric Society in 2017. His research spans macroeconomics, financial economics and econometrics, with a long standing focus on China's economy.
Research cited in this episode

The household responsibility system. Piloted in Sichuan and Anhui from 1978 and adopted nationwide by 1980, this reform kept land collectively owned but contracted it to individual households, who could sell output above a fixed state quota at market prices. It replaced work point pay with a direct link between effort and income, and the productivity gains it released freed the rural labour surplus behind China's first wave of industrialisation.
Township and village enterprises (TVEs). Rural, collectively owned firms that absorbed workers leaving agriculture through the 1980s and 1990s, often with local governments acting as guarantors for bank credit the firms could not secure alone. TVE employment grew from 28 million in 1978 to 135 million by 1997, and TVE output rose from under 6% of GDP to 26% over roughly the same period.
The hukou system. China's household registration system, introduced in 1958, ties access to housing, healthcare, education and grain rations to a person's registered location, rural or urban. It made moving to a city without an urban permit practically impossible. Restrictions eased in stages from the late 1990s, and the formal rural urban distinction was removed nationwide only in 2014.
"Grasp the large, let go of the small." The policy, initiated in 1997, under which China privatised or allowed the bankruptcy of small and medium state owned enterprises while retaining state control of the largest, most capital intensive firms in sectors such as infrastructure, energy and real estate.
WTO accession and permanent Most Favoured Nation status. China joined the World Trade Organization in 2001. From 2002, permanent MFN status with the United States removed the annual threat of tariff spikes on Chinese exports, a stability that Chen and Zha's paper credits with accelerating China's shift from labour intensive exports toward electronics and other capital intensive goods.
The 2009 stimulus and local government financing vehicles. In response to the global financial crisis, China launched a four trillion RMB fiscal package alongside a sharp expansion of bank lending. Much of the resulting infrastructure spending ran through local government financing vehicles, off budget entities set up to borrow for public projects; the debt they built up is now central to China's financial stability risks.
Total social financing. International Monetary Fund. 2026. "People's Republic of China: 2025 Article IV Consultation." IMF Country Report No. 26/044. The IMF's broadest measure of credit in the Chinese economy, the figure Kaiji Chen cites as roughly 315% of GDP in 2025.
More VoxDev Talks episodes

The Four Pests campaign and China's Great Famine, in which Shaoda Wang traces a darker chapter of Chinese economic history, the mass eradication of sparrows during the Great Leap Forward and the millions of deaths that followed.
The rise and fall of China's overseas lending, in which Sebastian Horn explains how China became the developing world's largest bilateral creditor, and why that lending boom has now gone into reverse.
Related reading on VoxDev.org

The Mandarin model of growth, on how China's system of promoting local officials for delivering growth shaped decades of investment led expansion.
How China became the world's factory: Trade, industrial policy, and growth, on the trade liberalisation and industrial policy that took China from export processing to global manufacturing leader.
The bubble dynamics of China's housing boom, Edward Glaeser on the construction surge behind the price rises Chen and Zha describe in this episode.

The ballot cast by a landless labourer counts the same as the one cast by the landlord whose fields he works. In this way, democracy can give excluded or subordinate groups access to political power.
That's not what has happened, at least, not in large parts of the Global South. Amit Ahuja (UC Santa Barbara) calls those at the bottom of the social order "subaltern groups". Three barriers stand between their votes and any lasting gain. Violence that goes unpunished. The erasure of their shared culture and history by those in power. And they have too little money and organisation to campaign for their rights.
The evidence shows that since 2014 the gains that these groups made in the previous three decades have started to reverse, he warns. The old hierarchies are reasserting themselves.
The research behind this episode:
Ahuja, Amit. 2026. "Subaltern Mobilization in the Global South." Annual Review of Political Science 29: 391-411. Open access.
To cite this episode:
Phillips, Tim, and Amit Ahuja. 2026. "Caste, race, and power: The barriers to political inclusion." VoxDev Talks (podcast).

About the guest

Amit Ahuja is Associate Professor of Political Science at the University of California, Santa Barbara, with research spanning ethnic parties and movements, military organisation, caste, and the politics of inclusion and exclusion in multiethnic societies. His book Mobilizing the Marginalized: Ethnic Parties Without Ethnic Movements won the 2020 Kamaladevi Chattopadhyay NIF Book Prize. He co-edited Internal Security in India: Violence, Order, and the State with Devesh Kapur.
Research cited in this episode

Subaltern. The word means "of inferior rank" and was used by Antonio Gramsci in the Prison Notebooks to describe groups subordinated by a ruling elite. Ahuja gives it three working attributes: a stigmatised identity acquired at birth, disproportionate poverty, and disadvantage that passes from one generation to the next.
Ranked and unranked social orders. Donald Horowitz, in Ethnic Groups in Conflict (1985; University of California Press, 2000 edition), separates societies where groups sit in a vertical hierarchy from those where they sit side by side. The Hausa-Fulani, Igbo and Yoruba in Nigeria, and Maronite Christians, Sunnis and Shia in Lebanon, compete without one being formally subordinate to another. Hierarchy is not a universal feature of diverse societies.
Greed against grievance. Paul Collier and Anke Hoeffler argued in 2004 that people join insurgencies to capture resources such as diamonds or minerals; Ted Gurr's Why Men Rebel (1970) argued instead that rebellion follows the gap between what people have and what they expect. Ahuja adds a third answer drawn from ethnography.
The dignity of insurgency. Elisabeth Wood's study of El Salvador (2003) and Alpa Shah's work on Maoist central India (2013) both find that the poor support armed movements partly because insurgents treat them better than landlords and state officials do. Participation confers dignity in a setting built to deny it.
The identity penalty. Devorah Manekin and Tamar Mitts (2022) find that nonviolent campaigns led by marginalised groups attract less popular participation and fewer elite defections than comparable campaigns led by dominant groups. Nonviolence works, but not equally well for everyone.
Undocumented rights. Afro-descendant communities on Colombia's Pacific coast, the San in the Central Kalahari, and Adivasis in India's forests all hold customary claims that colonial and postcolonial registries never recorded. Without a written record they are reclassified; Adivasis became "encroachers" under British forest law and the conservation regimes that followed.
The Dalit Panthers. Formed in western India in the 1970s, the group combined revolutionary politics with armed self-protection for Dalits facing caste violence, taking its name and some of its style from the Black Panther Party.
Military service as a school for organisers. Black South Africans recruited during the Second World War, and Dalits taken into the British Indian Army during the world wars, both emerged with leadership training and a more assertive political style. Dominant groups are usually reluctant to arm subaltern populations; labour shortages force the question.
The Indian Election Commission. Amit Ahuja and Susan Ostermann (2018) trace how the commission secured polling stations through the 1990s and 2000s, which made it possible for parties to bring marginalised voters to the poll. Institutional autonomy, backed by the courts, can partly offset a coercive environment.
Co-optation in Bolivia. Movimiento al Socialismo absorbed parts of the country's Indigenous movement and blunted its radical edge. Alignment with a party brings resources and patronage positions; it can also cost a movement its original demands.
More VoxDev Talks episodes

Can contact between groups reduce prejudice? Matt Lowe on why decades of evidence for the contact hypothesis look weaker under preregistration, and why prejudice does not simply fade with development.
Caste, identity, and worker productivity in India Daniel Keniston on how caste still sorts Indians into occupations, and what the economy loses because of it.
Related reading on VoxDev.org

Something to complain about: How minority representatives overcome ethnic barriers, on a formal complaints system in Bihar that gave Scheduled Caste ward members leverage over village heads who were holding up their projects.
How expanding voting rights shaped politics in India. Franchise expansion in 1935 and 1950 barely moved turnout or competition, and yet it shifted public spending towards the newly enfranchised.
No representation without reservation: The long-term limits of gender quotas in India, on what happens to descriptive representation once the quota is withdrawn.

Do economists treat what men do as normal human behaviour, and then study norms around women as a departure? If so, we need to pay more attention what we mean by masculinity.
That's what Pauline Grosjean (University of New South Wales, CEPR) does. She argues that these masculinity norms carry economic weight of their own. What do men around the world think "being a man" means? The consistent answers: winning against other men, self-reliance, control over women, aggression, and the avoidance of anything coded as weak or homosexual. Grosjean and her co-authors measure adherence to these core dimensions across 70 countries. How do these beliefs vary by age, economic development, and culture? And what do they predict about the way men will behave?
The research behind this episode:
Matavelli, Ieda, Pauline Grosjean, Ralph De Haas, and Victoria Baranov. 2026. "Masculinity Norms and Their Economic Implications." Annual Review of Economics 18:127-55.
To cite this episode:
Phillips, Tim, and Pauline Grosjean. 2026. "Masculinity norms and their economic implications." VoxDev Talks (podcast).

About the guest

Pauline Grosjean is a professor in the School of Economics at the University of New South Wales and a fellow of the Centre for Economic Policy Research. Her research spans culture and institutions, gender roles, violence and conflict, and how historical conditions shape economic behaviour long after those conditions have gone.
Research and concepts cited in this episode

Masculinity norms. Social expectations about how men should behave, or how a "real man" should behave. Grosjean draws the distinction from gender-role norms, which define men relative to women. Masculinity norms are largely about how men rank against other men. The five core dimensions measured in the survey are the importance of winning and status, self-reliance and help avoidance, violence, power over women, and avoidance of homosexuality.
Global Masculinity Survey. The 70-country survey behind the review, built on measurement tools developed by social psychologists and run in partnership with the European Bank for Reconstruction and Development across its countries of operation, then extended to a further 32 countries. See De Haas, Ralph, Victoria Baranov, Ieda Matavelli, and Pauline Grosjean. 2024. "Masculinity Around the World." CEPR Discussion Paper 19493 (gated).
Male-biased sex ratios in colonial Australia. Grosjean's earlier work uses convict transportation, which left parts of Australia with far more men than women, as a natural experiment. Areas that were historically male-biased show stronger masculinity norms today, more violence and male suicide, and weaker support for same-sex marriage; the last of these puzzled economists, because a shortage of women should, on a supply-and-demand reading, make men more relaxed about other men's homosexuality, not less. See Baranov, Victoria, Ralph De Haas, and Pauline Grosjean. 2023. "Men. Male-Biased Sex Ratios and Masculinity Norms: Evidence from Australia's Colonial Past." Journal of Economic Growth 28:339-96.
Ten to Men. An Australian longitudinal study of male health that measures both adherence to masculinity norms and a range of self-reported violent behaviour. Grosjean cites its strong associations between norm adherence and interpersonal aggression, sexual violence and self-directed harm.
Testosterone and status seeking. The link often assumed to run from testosterone to aggression is, on the evidence, better read as a link to status seeking, with aggression only one of its possible expressions. Testosterone itself responds to social context, diet and caregiving, so it sits inside the culture rather than outside it.
Misperceived norms. Men often overestimate how strongly other men endorse restrictive norms, and also misjudge what women want, believing women approve of controlling or emotionally closed behaviour when they do not. Grosjean links this to Marianne Bertrand's work on men declining paternity leave for fear of others' disapproval, and to the gender-segregated online spaces of the manosphere, where the norms are advertised and reinforced.
School intervention in Brazil. An ongoing trial with Ieda Matavelli comparing two approaches in schools, one closer to cognitive behavioural therapy and one that directly tackles beliefs about what it means to be a man. Matavelli's earlier work documents how warped young men's views of others' expectations can be, and tests whether correcting them changes behaviour.
More VoxDev Talks episodes

Non-elite women in politics. Soledad Artiz Prillaman on how patriarchal norms and social networks shape which women can take part in politics, and why a seat count tells you little about power.
Paid work for women and domestic violence. Deniz Sanin on how the spread of coffee mills in Rwanda gave women paid work and, with it, less violence at home.
Related reading

Masculinity norms and their economic consequences, a VoxEU column by the four authors setting out how these norms shape labour markets, health, education, households and politics, with the country and individual-level survey findings.

This is an episode from VoxDev's new podcast series, Ideas in Development. This series has a separate podcast feed, where you can find every episode of Oliver Hanney’s conversations with leading development thinkers.
YouTube: https://www.youtube.com/watch?v=jgwcv4fyU28 
Apple Podcasts: https://podcasts.apple.com/us/podcast/how-global-supply-chains-are-built-and-how-to/id1866874059?i=1000774798943 
Spotify: https://open.spotify.com/episode/5EECcLVxxEQtkX3gJZcHGH?si=94e15ed7834a4e86 
Audioboom: https://audioboom.com/posts/8922412-how-global-supply-chains-are-built-and-how-to-actually-attract-investment 
Substack: https://ideasindevelopment.substack.com/p/how-global-supply-chains-are-built 
Bill McRaith spent his career building supply chains – from 1980s Britain, to a factory in Panyu, China in 1990, to Ethiopia three decades later. He joins Oliver Hanney to explain what really attracts manufacturing investment to a developing economy, and why the model most countries are targeting no longer exists.
In this wide-ranging conversation we cover why China’s industry grew; why responsive local governments beat tax holidays; the collapse of the wage gap; how Ethiopia was chosen as a beachhead; and why apparel remains the right first-mover industry for countries not yet on the industrialisation road.

More households in low-income countries own a mobile phone than have electricity or clean water.  So why is education one of the last places it has been put to work?
Jenny Aker (Cornell) and Noam Angrist (Oxford) are the academic leads on a J-PAL Policy Insight that reviews 30 studies of mobile phone programmes in education. They tell Tim Phillips that the cheapest option, blasting out SMS messages, is usually the one that fails. But put a real person on the other end of the line, and that's an effective and inexpensive way to improve education. The phone is a platform, Angrist points out, not a pedagogy.
The research behind this episode:
Abdul Latif Jameel Poverty Action Lab (J-PAL). 2026. "Leveraging Mobile Phones for Learning." J-PAL Policy Insight. Academic leads Jenny Aker and Noam Angrist; insight authors Jessica Williams and Demitria Wack.
To cite this episode:
Phillips, Tim, Jenny Aker, and Noam Angrist. 2026. "Using mobile phones to support teachers, parents, and students." VoxDev Talks (podcast).

About the guests

Jenny Aker is the Daniel G. Sisler Professor of Development Economics at Cornell University. She is also a non-resident fellow at the Center for Global Development and co-chair of the Digital Identification and Finance Initiative at J-PAL Africa. Her research spans mobile technology and agricultural markets, adult literacy and numeracy, financial inclusion, and the design of cash transfers and social protection in Africa.
Noam Angrist is Academic Director of the What Works Hub for Global Education and Senior Research Fellow at the University of Oxford, and co-founder of Youth Impact, one of the largest NGOs scaling health and education programmes backed by randomised trial evidence. His research spans distance education, the measurement of implementation, and how programmes that work in a trial can be made to work at national scale.
Research and concepts cited in this episode

One Laptop per Child. The plan to put a cheap laptop in the hands of every child in the developing world, launched in 2005 and abandoned as evidence of its effect on learning failed to appear. Angrist raises it as the cautionary tale for anyone introducing a device: the schools could not absorb it, the software was an afterthought, and nobody planned for charging or breakage. Phones avoid that problem because they are already there, so the marginal cost of adding good pedagogy is low. A long-term follow up from rural Peru found no positive effect on educational outcomes.
Mobile coverage. GSMA. 2025. "The State of Mobile Internet Connectivity 2025." Around 95% of the world's population now lives within range of a broadband network. In sub-Saharan Africa coverage is about 80% and adoption about 60%, and what people mostly hold is a simple handset rather than a smartphone.
Phone tutoring in emergencies. Angrist, Noam, Micheal Ainomugisha, Sai Pramod Bathena, Peter Bergman, Colin Crossley, Claire Cullen, Thato Letsomo, Moitshepi Matsheng, Rene Marlon Panti, Shwetlena Sabarwal, and Tim Sullivan. 2023. "Building Resilient Education Systems: Evidence from Large-Scale Randomized Trials in Five Countries." NBER Working Paper 31208, revised 2025. Weekly maths texts combined with live tutoring calls raised test scores by 0.30 to 0.35 standard deviations at an average cost of US$11 per child. SMS on its own moved the needle only where access to learning was already very low.
Scaling up without the drop. The first Botswana phone tutoring trial produced an effect of about 0.12 standard deviations; by the fifth replication, in Uganda, Angrist says the effect was over 0.8, and implementation fidelity data improved in step with the order of the studies. He sets this against the voltage drop described by John List in The Voltage Effect (2022), the tendency of programmes to lose their punch as they scale. Common, Angrist argues, but not inevitable.
A/B testing the tutoring call. Angrist, Noam, Claire Cullen, and Janica Magat. 2025. "Cheaper (and More Effective) by the Dozen: Evidence from 12 Randomised A/B Tests Optimising Tutoring for Scale." What Works Hub for Global Education Working Paper 2025/001. Roughly half the cost of a tutoring programme goes on scheduling rather than teaching. One call of 40 minutes every fortnight, instead of 20 minutes every week, delivered the same dosage for less. Handing the call to a parent partway through doubled to tripled the effect; motivational nudges to parents did nothing.
Monitoring teachers by phone. Aker, Jenny C., and Christopher Ksoll. 2019. "Call Me Educated: Evidence from a Mobile Phone Experiment in Niger." Economics of Education Review 72: 239 to 257. Weekly calls to the teacher, the village chief and two students raised maths and reading scores. The counter-example comes from Côte d'Ivoire, where nudging parents and teachers at the same time left the most motivated teachers turning up less; too much scrutiny, the authors suggest, corroded the motivation the programme depended on.
Platform and pedagogy. Angrist's phrase for the distinction the review keeps returning to. The phone carries an intervention; it is not the intervention. Targeted instruction and early childhood stimulation both travel down the line, but somebody still has to decide how often to call, who picks up, and what happens on the call. The same traffic makes the phone a good instrument for research, since the data is traceable and the units of randomisation are well powered.
Education Cannot Wait. The United Nations fund for education in emergencies and protracted crises. Its 2022 global estimate put 222 million crisis-affected children and adolescents in need of urgent education support, up from about 75 million in 2016. Angrist uses the same figure in the five-country paper, and cites it here to make the point that the humanitarian structures for education in emergencies exist while the evidence base underneath them barely does.
More VoxDev Talks episodes

How AI can put 20 years of development evidence to work. Iqbal Dhaliwal on why One Laptop per Child failed, and what that failure should teach anyone now betting on AI in development.
Why we need to invest in foundational learning. Benjamin Piper on the gap between getting children into school and getting them to read, which is the gap most of these phone programmes are trying to close.
Related reading on VoxDev.org

How simple technologies can improve learning: Evidence from Niger, in which Jenny Aker and Christopher Ksoll show that a weekly call to teachers and students almost doubled the effect of an adult education programme.
Educating children in emergencies: Global evidence from five randomised trials, the five-country evidence on cheap phone tutoring that runs through much of this conversation.
How AI tutors improved learning in Nigeria, on a six-week after-school programme in Edo State that produced large gains in English, and on the guardrails that made it work.