Make India grow faster
The “Gujarat Model” was presented as proof that a Modi government could accelerate development—not simply keep India growing at its old pace.
01 Start with the promise
What would count as keeping the promise?
The paper turns Modi’s broad pitch into two things it can test.
Make government work better
“Minimum government, maximum governance” promised cleaner, more capable government and stronger public institutions.
A simple example / two schools
How would you measure the performance of a school?
Imagine you are appointed principal of a school. Its average score is 60 when you arrive. Two terms later, it is 70.
Your school
70
Similar school
80
You improved the school’s score by 10 points. In isolation, you did a good job.
02 Build a fair comparison
You cannot become Vishwa Guru without comparing yourself with the Vishwa
The authors combine small and large shares of several broadly comparable countries until the mix behaves like India did before 2014. The paper calls the result synthetic India. Here, we call it Bhagwan Bharose India.
How Bhagwan Bharose India is put together
The basic rules / set before seeing the answer
Start with comparable countries. Let the numbers choose the mix.
1
Begin in the same broad neighbourhood
The authors start with 14 large developing or emerging economies whose income is similar to, or below, India’s. That starting list is a judgment call.
2
Mix until the old numbers line up
The algorithm adjusts every country’s share until their combined income path and economic basics resemble pre-2014 India. The shares must total 100%.
3
Avoid a country with its own big break
A comparison country should not have its own 2014 turning point—such as regime change, war, famine or economic shock. The authors say none of the final five did.
See the 14 countries the algorithm started with
Bangladesh, Brazil, China, Egypt, Ethiopia, Indonesia, Malaysia, Mexico, Pakistan, the Philippines, South Africa, Thailand, Turkey and Vietnam.
What the algorithm settled on
Five countries become one Bhagwan Bharose India
None of the five has to look like India on its own. Only the combined mix does.
Ethiopia
38% China
28% Bangladesh
25% Pak
7%
38% China
28% Bangladesh
25% Pak
7%
Ethiopia 38 · China 28 · Bangladesh 25 · Pakistan 7 · Philippines 2
About $59The paper’s measure of the typical pre-2014 income mismatch per person
≈ 1%That mismatch as a share of India’s 2013 income level
Check 1 / income before Modi
For 30 years, the income paths mostly line up
The paper shows six checkpoints between 1984 and 2013. At four of them, real India and Bhagwan Bharose India are less than 1% apart. The two largest gaps are about 5% and 3%.
Close enough to continueSwitch to “Every year” to see the full 30-year path in the same space.
Source: Table 3 and Figure 6A, PDF pages 21–22
YearReal IndiaBharose IndiaGap
1986Real India$1,333.15Bharose India$1,400.075.0%
1991Real India$1,597.10Bharose India$1,550.003.0%
1996Real India$1,976.40Bharose India$1,978.07<0.1%
2001Real India$2,422.66Bharose India$2,424.02<0.1%
2006Real India$3,462.68Bharose India$3,442.890.6%
2011Real India$5,113.89Bharose India$5,117.15<0.1%
“Gap” simply shows how far the two numbers are apart, as a share of real India’s income at that checkpoint.
Custom redraw from Penn World Table 11.0 using the paper’s published country shares. The paper’s own calculation uses more precise shares than the rounded numbers it prints.
Investment
Real India0.22
Bharose India0.20
Exports
Real India0.04
Bharose India0.06
Government consumption
Real India0.14
Bharose India0.15
Human capital
Real India1.70
Bharose India1.63
These four pre-2014 comparisons come from Table 3. They are close, but four indicators can never prove that the two Indias are alike in every possible way.
Starting point / all five countries
Original mix
Typical old-path miss≈ $58
Bangladesh share25%
The main calculation uses all five countries, with Ethiopia contributing the largest share.
The paper’s one published removal test
Ethiopia taken out
Typical old-path miss≈ $132
Bangladesh share76%
The match gets worse, but the estimated post-2014 shortfall becomes larger rather than changing direction.
Full replacement mix + chart not publishedThis is the only one-country-out test the paper reports. It tells us the conclusion survives removing Ethiopia; it cannot tell us what would happen after removing every other country in turn.
03 Make your prediction
What do you think happened?
Choose the result you expect. The question is how real India compares with its made-up twin—not how India in 2023 compares with India in 2013.
Your prediction
Pick one square
Your choice will reveal the paper’s answer.
Governance →
Economy →
ImprovesDeclines
Real India grew—but Bhagwan Bharose India grew more
≈ 10% lower
By 2023, the authors estimate that real income per person was roughly $1,000 a year lower than in Bhagwan Bharose India.
Real India ≈ 90Bharose India = 100
When the authors test the other countries in the same way, only one ends with a larger gap.
Governance also falls short
96 / 100
Across the 100 year-by-year comparisons, 96 put real India below Bhagwan Bharose India. In 92, none of the same tests on the other countries produces a gap as large.
4 other96 underperform
Across all ten measures, India’s overall gap is larger than every gap produced by testing the other countries in the same way.
Elections
Are elections and political competition healthy?
This measure covers fair elections, voting rights, elected leaders and the freedom to compete for power.
The final gap is shown relative to India’s own 2013 score.
Every year / real India compared with Bhagwan Bharose IndiaAt zero, the two Indias match
Before 2014, the line stays near zero. After 2014, real India falls behind its made-up twin.
Data source: V-Dem, a global project that tracks democracy and governance. Custom redraw using the country shares in Table 1; final gap from Table 2; paper reference: Figure A1.
See the country mix behind this measureArgentina 61.1% + Sri Lanka 14.5% + Brazil 13.4% + Malaysia 7.7% + South Africa 3.3%
04 Check whether the comparison holds up
Could the method be fooling us?
We have already seen the first check: the two income paths match before 2014. The paper then tests whether other countries produce equally large gaps and whether a deliberately wrong start date invents one.
Recap / before 2014
First, the made-up India follows the real one
You saw this when the made-up India was built: across the 30 years before Modi took office, the two income paths stay close.
30-year matchThe algorithm uses 1984–2013 to build the comparison. It does not get to learn from what happened later.
Source: Figure 6A and Table 3, PDF pages 21–22
1 / Follow the two linesOrange is real India; blue dashed is Bhagwan Bharose India.
2 / Check before 2014The paths should run close before the gate.
3 / Then look afterThe split after the gate is the estimated shortfall.
Custom redraw from the paper’s reported income data and published country shares.
What this comparison still cannot prove
Working paper
The study has not yet passed journal peer review. Methods, text and conclusions may change.
Country choices
The authors explain their starting list, but deciding which countries count as fair comparisons remains a judgment call.
Expert-rated governance data
The ten governance measures come from V-Dem and depend partly on expert assessments. The paper cannot rule out a bias specific to how India was rated.
Modi’s decade as one bundle
The calculation covers India under the Modi government as a whole. It cannot tell us which policy or event produced each gap.
What “p-value” means in this paper: the authors run the method on each of the other countries and count how many end with a gap at least as large as India’s. A smaller number means India’s result is harder to reproduce by chance. Here, 0.00 means none of those tests produced a larger gap. It does not mean there is zero chance the paper is wrong. A value of 0.50 puts India around the middle of the group.
★ Bonus level / the political puzzle
Authors’ interpretation · not a causal result from the model
If both missions failed, how did Modi keep winning?
This comparison cannot answer that question. The authors instead point to a durable Hindu-nationalist base and a divided opposition in India’s first-past-the-post electoral system.
Authors’ interpretation · not a causal result from the model
Switch election
31.3% voteBJP wins a single-party Lok Sabha majority while opposition votes are fragmented.
Uttar Pradesh / the clearest example in the paper
Almost the same BJP vote. Very different seats.
In 2014, SP and Congress often competed against each other as well as the BJP.
BJP7142.3% vote
VS
SP + Congress7fragmented
Paper’s reading: a strong base plus opposition vote-splitting converted pluralities into a huge seat advantage.
THE
SHORT
VERSION
SHORT
VERSION
The whole argument in three steps
What the paper is saying
1
India improved after 2014, but improvement alone is not the question. The question is whether India did as well as it reasonably could have.
2
The authors build Bhagwan Bharose India from a mix of other countries that resembles real India before 2014.
3
The paper’s answer is that real India falls short on both income and governance. That is still an estimate from a working paper—not an observed alternate history.
05 Check the original paper
See where each claim comes from
Open the highlighted passages to compare this guide with the authors’ original wording and surrounding context.
PDF page 1The abstract’s claim+
Grier and Grier, 2026, abstract.
PDF page 8How the comparison India is formed+
Grier and Grier, 2026, Section 4A.
PDF page 15The 96-of-100 governance result+
Grier and Grier, 2026, Section 5A.
PDF page 22The final income estimate+
Grier and Grier, 2026, Section 5B.
PDF page 28The false-start test+
Grier and Grier, 2026, Figure 8 discussion.